RIGHTS BEFORE COURTS
Rights Before Courts A Study of Constitutional Courts in Postcommunist States of Central and East...
15 downloads
1321 Views
1MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
RIGHTS BEFORE COURTS
Rights Before Courts A Study of Constitutional Courts in Postcommunist States of Central and Eastern Europe By
Wojciech Sadurski European University Institute Florence, Italy and University of Sydney, Australia
A C.I.P. Catalogue record for this book is available from the Library of Congress.
ISBN 1-4020-3006-1 (HB) ISBN 1-4020-3007-X (e-book)
Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. Sold and distributed in North, Central and South America by Springer, 101 Philip Drive, Norwell, MA 02061, U.S.A. In all other countries, sold and distributed by Springer, P.O. Box 322, 3300 AH Dordrecht, The Netherlands.
Printed on acid-free paper
All Rights Reserved © 2005 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed in the Netherlands.
CONTENTS Acknowledgements
ix
Preface
xi
PART 1 1. THE MODEL OF CONSTITUTIONAL REVIEW IN CENTRAL AND EASTERN EUROPE: AN OVERVIEW 1 1. The Emergence of the Current Model 2. The Powers of Constitutional Courts and Initiators of the Review Process 3. The Tenure and Selection of Judges 4. Constitutional Courts’ Pursuit of a Monopoly over Constitutional Adjudication 5. Conclusions
19 25
2. CONSTITUTIONAL COURTS IN SEARCH OF LEGITIMACY
27
1. The Legitimacy Dilemma 2. Constitutional Courts Between the Judicial and Legislative Branch 3. Why the “Continental” Model of Review: Reasons or Rationalisations? 4. Constitutional Courts as Protectors of Minorities? 5. Conclusions
27 33 40 58 62
3. THE MODEL OF JUDICIAL REVIEW AND ITS IMPLICATIONS
65
1. Abstract Review 2. Ex-Post Review 3. Final Review 4. Conclusions
65 74 79 85
4. CONSTITUTIONAL COURTS AND LEGISLATION
87
1. The Impact of Constitutional Courts on Law-Making 2. Determinants of the “Strength” of Judicial Review 3. Constitutional Court and the Parliamentary Minority 4. The Question of Judicial Activism and Restraint 5. Conclusions
1 5 14
87 90 93 96 104
PART II 5. JUDICIAL REVIEW AND PROTECTION OF CONSTITUTIONAL RIGHTS 107 1. Two Theories about Judicial Review 2. The Fact-Sensitivity of a Theory of Judicial Review
109 114
VI
CONTENTS
3. Rights Protection and Disagreement about Rights 4. Prudence and Judicial Review 5. Conclusions
118 124 125
6. PERSONAL, CIVIL AND POLITICAL RIGHTS AND LIBERTIES 127 1. A Right to Life and Dignity 2. Freedom of Religion 3. The Right to Privacy 4. Freedom of Movement and the Right to Choice of Residence 5. Citizenship and Voting Rights 6. Freedom of Petition, Assembly and Association 7. Freedom of Expression 8. Conclusions
128 135 146 149 151 156 158 169
7. SOCIO-ECONOMIC RIGHTS
171
1. Controversy Around Socio-Economic Rights 2. Constitutional Catalogues of Socio-Economic Rights 3. The Status of Socio-Economic Rights 4. The Drawing of Distinctions Between Different Types of Rights by the Courts: Social Security Cases 5. The Right to Work 6. Rights to Health and Education 7. Conclusions
173 176 178
8. EQUALITY AND MINORITY RIGHTS
195
1. Equality and Constitutional Review 2. Gender and Sexual Orientation Equality 3. Special Case of Affirmative Action 4. Minority Issues in Central and Eastern Europe: An Overview 5. Constitutional Design of Minority Rights: Group or Individual Rights? 6. Linguistic Rights 7. The Special Case of Minority Representation in Public Authorities 8. Conclusions
195 199 203 205 207 212 218 222
9. “DECOMMUNISATION”, “LUSTRATION”, AND CONSTITUTIONAL CONTINUITY
223
1. Main Dilemmas Raised by Decommunisation and Lustration Laws 2. Lustration and Decommunisation in Central and Eastern Europe 3. Retroactive Extensions of Statutes of Limitation 4. Conclusions: Transitional Justice and Constitutional Continuity
224 234 249 258
10. RESTRICTIONS OF RIGHTS
263
1. Constitutional Design of Limits on Rights
263
181 185 189 190
CONTENTS
VII
2. Constitutional Review of Statutory Limits on Rights: Proportionality Scrutiny 3. Other Standards of Rights Restrictions 4. Concluding Remarks on Rights Limitations 5. Postscript on Rights and Duties
266 276 282 284
CONCLUSIONS
289
Endnotes
301
Bibliography
355
Index
367
ACKNOWLEDGEMENTS I have written this book while working at the European University Institute (EUI) in Florence, and have benefited from the remarkable resources of this unique educational and research institution. The intellectual richness of the place derived from the interactions with colleagues as well as with doctoral students is the main asset of the EUI, but I should also note with gratitude the significant financial contribution that the Research Council of the Institute made to the research project which has resulted in this book. I am also very grateful to the Australian Research Council for its support. My main debt is to my colleagues and friends both in the EUI and in places as close to Florence as Pisa or as far as Sydney with whom I had the benefit of discussing both constitutional theories in general and constitutional courts more specifically. Amongst those I wish to thank in particular are Sergio Bartole, Tom Campbell, Adam Czarnota, Bruno De Witte, Venelin Ganev, Leszek Garlicki, Tomasz Gizbert Studnicki, Martin Krygier, Wiktor Osiatynski, Thomas Mertens, Alessandro Pizzorusso, Robert Post, Jiri Priban, Kim Lane Scheppele, Daniel Smilov, Ruti Teitel, Michael Thaler, Michel Troper, Neil Walker, Jeremy Webber, Miroslaw Wyrzykowski, Jan Zielonka and Jacques Ziller. I also wish to express my thanks to Professor Yves Mény, President of the EUI, who has supported the project continuously. I have been lucky to benefit from the help of a number of extremely gifted doctoral students who provided me with their assistance over the past few years, in particular Mehreen Afzal, Navraj Singh Ghaleigh, Victoria Jennett, Cormac MacAmhlaigh, Euan Macdonald, Sejal Parmar and Anna Slinn. My very special thanks go to Marlies Becker whose secretarial support has been invaluable. Earlier versions of some parts of this book have appeared as articles in the Journal of Law and Contemporary Problems (vol. 65 (2002): 101-127), in the Oxford Journal of Legal Studies (vol. 22 (2002): 275-99), and also as chapters in Tom Campbell, Keith Ewing and Adam Tomkins (eds), Sceptical Approaches to Human Rights, Oxford: OUP, 2001 and in Jiri Priban, Pauline Roberts & James Young, eds., Systems of Justice in Transition, Ashgate: Aldershot 2003. I am grateful to the publishers of these journals and books for their kind permission to include the revised versions of these texts here. The book is dedicated to the memory of my mother Anna Sadurska, 1921-2004.
PREFACE On the eve of the collapse of Communist rule in Central and Eastern Europe (CEE), a doyen of international political science, Professor Klaus von Beyme, summarised his conclusions in a comparative study on judicial review in the following manner: “Judicial review works most effectively and intensely in those political systems in which the Rechtsstaat had been established before the advent of democracy, and where federal structures functioned as an obstacle to the majority principle dominating parliamentary democracies”.1 According to this theory, certainly unimpeachable in 1988 when these words were published, judicial review was to have little chances of operating successfully in CEE after 1989: in all of the countries in this region the advent of democracy preceded the establishment of the Rechtsstaat, and only one (Russia) is a federation. And yet, in most of these countries, judicial review by constitutional courts became both “effective” and “intense”, to use von Beyme’s terminology: it has been “intense” in the sense that these courts have not shied away from invalidating often very important pieces of legislation, and “effective” in that other institutional actors in each political system have exhibited a high degree of compliance with the decisions of the constitutional courts. One explanation for the impressive emergence of constitutional courts in all post-communist states of CEE, notwithstanding the absence of the federal or Rechtsstaat-related roots identified as necessary by von Beyme, could be that it was a case of a simple and rather mechanical importation to post-authoritarian states of an institutional design that had proved successful in other countries. Indeed, the post-1989 period was the time of great constitutional borrowings in Central and Eastern Europe: there was remarkably little experimentation in constitutional design, with the aim of the constitution-makers in general being largely to follow the established patterns of Western European (and, to a lesser extent, American) constitutionalism. In the minds of the main constitutional actors in the region, what mattered most was to create a situation in which the constitutional texts of their countries were as similar as possible to those adopted on the Western side of the continent. Thus the transition represented a move towards something safe, to something familiar. The discourse of “return to normalcy”, which prevailed in post1989 Central Europe (“No experiments! We want to build a normal country, a normal economy, a normal constitutional rule-of-law”), hinged upon the perceived standards of real Western states and societies.2 And part of the perceived standards of Western constitutionalism was the centrality of constitutional review: of a system whereby judicial or quasi-judicial bodies can set aside and invalidate the democratically enacted laws on the basis of their alleged inconsistency with constitutional norms. This perception was no doubt largely accurate: the systems of unlimited parliamentary supremacy are now the exception rather the norm in the West; and yet, what was puzzling was that importation of the institutional system of constitutional review was not accompanied by any serious and sincere reflection upon the implications, including the negative ones, of such a system. Consequently, there has been little reason for constitutional courts themselves, or for their doctrinal supporters and commentators, to defend their position in terms of its legitimacy in a
XII
PREFACE
democracy: such legitimacy was assumed rather than contested and debated. The absence of this type of reflection in the post-communist setting becomes all the more striking when one considers the important debates in Western liberal democracies over the limits and justifications of judicial constitutional review. Consequently, the powers of constitutional courts in CEE have been only rarely, if ever, contested.3 This omission is puzzling. An institution has been straightforwardly imported into post-communist systems from the West, but the ideological and theoretical controversy surrounding that institution has not. If there has been a degree of angst caused by the activities of the constitutional courts in the countries that have served as models to new democracies, then there is no reason to refuse to ask similar questions with respect to their progeny. To be sure, the longest and most deeply ingrained habit of challenging active constitutional review has developed in the country that influenced the establishment and development of constitutional courts in CEE to a very modest degree, namely the United States. The very titles of two recent books: by Mark Tushnet, Taking the Constitution Away from the Courts;4 and by Jamin B. Raskin, Overruling Democracy: the Supreme Court vs. the American People5 express well the strength of the scholarly challenge to the principle of judicial supremacy and to the finality of judicial articulation of constitutional norms. Tushnet calls for a “populist” constitutional law in which the courts (including the Supreme Court of the United States) will not occupy any privileged position vis-à-vis the political branches of government, in terms of providing authoritative pronouncements about the meaning of Constitutional provisions. Further, not only academic commentators, but also judges in the United States have long been reminding themselves of the requirements of proper deference to democratically elected bodies, and recognition of the presumption of the constitutionality of statutes. Even if these reminders have often been only rhetorical, serving in many cases as a mere disguise for activism, the very fact that the Justices of the Supreme Court have felt the need to justify themselves in that manner is significant. Often, however, arguments as to the proper respective roles of the legislature and the Court have served as a justification for refusing to invalidate the legislation in question, as, for example, in the decision declining to invalidate legislation prohibiting doctor-assisted suicide. In this judgement, the Supreme Court declared: “By extending constitutional protection to an asserted right or liberty interest, we, to a great extent, place the matter outside the arena of public debate and legislative action. We must therefore ‘exercise the utmost care whenever we are asked to break new ground in this field’….”.6 A similarly sceptical attitude towards activist judicial review has been articulated by many observers of the Canadian Supreme Court, after rights based review was instituted there with the entrenchment of the Charter of Rights and Freedoms in 1982. One legal scholar has noted: “Over the two decades that judges have given concrete meaning to abstract rights, celebration of constitutional rights has turned to scepticism and the institutional legitimacy of the courts has been threatened”.7 This, it should be emphasised, is not the only attitude towards rights review that has been voiced academic or political circles in Canada, and it should be also noted that the public prestige of the Supreme Court among the Canadian population is very high. Interestingly, however, a 1999 survey showed that Canadians were divided evenly
PREFACE
XIII
on the proposition that “the right of the Supreme Court to decide certain controversial issues should be reduced”.8 In contrast to the United States and Canada, European constitutional adjudication has not developed a tradition of self-doubt, agonising over legitimacy, or “exercising the utmost care” whenever “breaking new ground” in constitutional matters. This may be due to a number of factors. For one thing, there is a much stronger tradition and habit of deference to the highest bodies of government (including the judicial branch) in Western Europe than in the United States. More importantly, the power of constitutional adjudication, in its abstract form, is explicitly contained in the constitutional texts of European states, rather than only being recognised as implicit therein by the doctrine developed by the court itself, as was the case in the United States.9 Continental constitutional courts do not, therefore, feel any special reasons for anxiety about their own legitimacy when deciding on the constitutionality of statutes. The general constitutional design of these institutions situates them in a special position, not comparable to that of ordinary courts. As one American observer has correctly observed, “German, French, and other continental constitutional tribunals have neither hesitated nor apologized when issuing wide-ranging decisions on basic constitutional issues, often drawing on unwritten or historical principles and values”.10 It is, however, important not to overstate this point. While the tradition of challenging and criticising the constitutional courts is nowhere near as powerful in Western Europe as it is in the United States, it is not quite absent. In France, for example, the warnings against “le gouvernement des juges” long precede the establishment of the Conseil constitutionnel, going back to a 1921 book by Edouard Lambert who addressed his criticism to the role of the US Supreme Court.11 More recently, Bernard Chantebout has concluded that the power of constitutional interpretation is by its very nature political, and that “it raises the problem of the compatibility of constitutional review with democracy”.12 The “problem” as defined by Chantebout, even if not insurmountable, is real and dramatic; after all, when a court (or a court-like body) sets aside a law on the basis of its inconsistency with constitutional rights, then what it is in fact doing is substituting its own interpretation of those rights – often vaguely defined – for the interpretation, perhaps honest and made in good faith, of the majority of the democratically elected parliament. Electorally accountable bodies presumptively enjoy the paramount legitimate authority to decide on issues of policy about which members of society in general disagree. The judiciary, as we know, is notoriously ill-equipped to evaluate options and choices on some issues, such as socio-economic policies with important financial implications. Placing the protection of certain rights and other political values (such as “social justice”) in the hands of constitutional courts simultaneously removes these spheres from the agenda of the elected bodies, and consequently restricts the capacity of citizens to participate in political decisions that affect the contours of such rights or values. Even if the court is empowered by the constitution to decide on such issues, the wisdom and fairness of the constitutional design should be deliberated openly. Why has the democratic legitimacy of constitutional courts in CEE been taken for granted? Although one should avoid simple (perhaps simplistic) answers formu-
XIV
PREFACE
lated in terms of vested interests and institutional self-aggrandisement, these explanations should not be discounted altogether. After all, the various constitutional discourses have been primarily produced by those who stand to gain the most from the implementation of theories supporting a strong role for constitutional courts: academic constitutional lawyers and constitutional judges themselves (the latter being largely drawn from the former category). Self-congratulatory rhetoric supports the position of both the constitutional judiciary and law professors linked with each other in a symbiotic relationship. Robust constitutional review strengthens the status of academic constitutional lawyers (they get more material to work on – not just on the text of constitutional acts but also on the case law; they may hope to be cited in the judgements; and – the ultimate reward – may find themselves, one day, on the bench), while the supportive doctrines produced by constitutional lawyers elevate the position of constitutional judges vis-à-vis the political branches of governments. The two groups are thus mutually sustaining. This phenomenon is not restricted to the post-communist states of CEE. Martin Shapiro has noted how the emergence and growth of constitutional review in Western Europe has affected favourably the fortunes of academic constitutional lawyers: “European constitutional law teachers went from the bottom of the pecking order of teachers of something like Freshman civics, to near the top of the order as constitutional judicial review came to flourish on the Continent. And just as that particular body of law made more of them, they made more of it”.13 This shift has also been recognised – albeit rarely – by Western European academic constitutional lawyers and judges. Bernhard Schlink, who combines both of these professional roles, has caustically criticised the relationship between the German Federal Constitutional Court (Bundesverfassungsgericht) and the constitutional academia in his country: “Karlsruhe locuta, causa finita – this remark creates an image of this new situation, in which the Bundesverfassungsgericht speaks ex cathedra and representatives of dethroned constitutional scholarship stand at its feet”.14 He further remarked that constitutional scholarship has adapted to the Federal Constitution Court “as a sort of junior partner”, and that many constitutional law professors have behaved “as loyal compilers and systematizers of [the Constitutional Court’s] decisions, even as possible candidates for future positions on the Court… Constitutional scholarship would like to participate in power, and it realizes that the courtiers are rewarded for their service to the royal court by being allowed to influence it”.15 There are, no doubt, also more profound reasons for the largely uncritical approach adopted towards the constitutional courts in the region. The very high social standing of those courts, compared to other public institutions, is a consequence of, and stands in contrast with, the general disenchantment with the political branches of the government. Not unlike in post-war Germany, “politics” was largely discredited after the fall of Communism, and there is a widespread, cynical conviction that politics is a dirty business. Being novel institutions (with the exceptions of Poland and Yugoslavia), constitutional courts have not had to bear the same general opprobrium as those tainted by past complicity in non-democratic practices. Further, in at least some of the post-communist countries (for instance, in Hungary), they were viewed by the then anti-communist, democratic opposition
PREFACE
XV
parties as one of very few effective institutional checks upon the power of less-thanlegitimate parliaments and governments, dominated as they were by the Communists and their successor parties. The two professional groups that together provided the pool of candidates for the position of constitutional court judge – legal academics and the judiciary – were popularly regarded as less compromised than the “political class”. (Even though the latter was largely composed of new people, the pervading cynicism and lack of trust in politicians inherited from the ancien régime affected all politicians during the transition). The fact that legal academics – and most of the justices of the post-communist constitutional courts were law professors, and, in particular, constitutional law professors16 – were among the most conformist in academic circles of communist states has somehow not contaminated their reputation, perhaps thanks to the relatively high prestige enjoyed in CEE by academics generally. As for the judiciary, only a few judges in the Communist era were directly involved in politically sensitive trials, and it would be correct to say that the bench was not in general held in disregard by the public in most of the countries of the region – at least, that is, in comparison with other public figures. There is a more universal reason for the relatively high levels of public support enjoyed by constitutional courts, one that transcends the various state boundaries of CEE. Although these courts are essentially political institutions, engaged in wideranging law-making and policy-setting, they can escape the social criticism endured by other political and legislative institutions thanks to their ability to draw upon the appearance of neutrality enjoyed by courts in general.17 The traditional paradigm of the judicial process is that of a neutral umpire adjudicating between two parties and dispassionately dispensing justice in accordance with apparently objective rules. This is scarcely an apt description of the role of courts when adjudicating in constitutional disputes. When engaged in abstract judicial review, constitutional courts act rather as an additional legislative chamber, which may, often on the basis of vague and eminently controversial constitutional pronouncements, strike down legislation enacted by another body that is also committed to the implementation of the same constitutional norms. (I will return to this point, at greater length, in chapter 2). It is significant that an American legal scholar, very sympathetic to the Constitutional Courts of CEE (and the Hungarian Court, in particular) observes straightforwardly: “Western observers may make the mistake of thinking of these constitutional courts in Eastern Europe as courts in the narrow sense. The better analogy for the Hungarian Constitutional Court might be that it functions like an upper house of Parliament – something like a House of Lords”.18 Indeed it does; however, the paraphernalia of the judicial process – elaborate procedural requirements, the aura of reasoned debate, courtroom symbolism – create an air of majesty and dignity that other political institutions do not possess. The actual character of constitutional courts is substantially obscured by their very practices; the judicial costume lends extra legitimacy and respectability to these essentially political institutions. This is not to say that constitutional courts are indistinguishable from legislative bodies: their role in lawmaking is restricted by a number of constraints (differing from country to country) that do not apply to the parliament. They operate under procedures that borrow greatly from courtroom processes and symbolism. They
XVI
PREFACE
rarely control their own agenda, in the sense that they are not normally selfactivated, although one should not overstate the importance of this factor: in a system in which a large number of important laws are brought before the constitutional court for review, the significance of this distinction between the court and the legislature is greatly diminished.19 Further, they are more of “negative” than “positive” legislators, to use a classical distinction drawn by Hans Kelsen – but then, they sometimes issue affirmative pronouncements as to the specific ways in which defective laws have to be amended, and, in any event, the distinction between “positive” and “negative” in this context is often hard to sustain.20 Most importantly, they are compelled to argue in ways that use the structure of legal syllogism, which means, in effect, that when they wish to invalidate a law of which they disapprove, they must argue that it is inconsistent with their understanding of the constitution, rather than that it is simply unwise or unjust tout court. Certain types of arguments in favour or against a given law that are available to the members of parliament (for example, concerning the degree of societal support for a proposed measure) are normally not available to the justices of constitutional courts. Despite the fact that these courts attempt (often in good faith) to establish themselves as neutral interpreters of the allegedly objective meaning of the constitution, commentators need not take these assurances at face value. The rejection of the judicial paradigm makes it possible for scholars to raise questions concerning the legitimacy of constitutional courts, both in terms of lawmaking and the displacement of the will of other institutions, in particular the parliaments. It does not follow that, at the end of the day, legitimacy is to be doubted on the simple basis of a lack of electoral pedigree. But if the role of the constitutional courts is to be maintained, or even enhanced, their legitimacy has to be argued for, rather than simply assumed. This is the starting point for this book. Its main aim is to consider two main aspects of the constitutional courts in CEE: their democratic legitimacy, particularly when in confrontation with parliamentary claims to legislative supremacy, and their success in protecting constitutional rights. The assumption (which I will elaborate in more detail in Chapter 5) is that these are two different and mutually independent criteria for judging an institution: how legitimate it is in doing what it does, and how effectively it articulates and protects citizens’ rights, as written into the constitution. Both of these criteria are important, and it may well be that an institution that is somewhat deficient in terms of one does very well in the other, and vice versa. Nothing is gained, however, and much is lost in terms of clarity of argument, if these two issues are merged into one. This consideration informs the structure of this book. Part I focuses on the problem of legitimacy. After offering an overview of the structure, the functions and the composition of constitutional courts in CEE (Chapter 1), I discuss the “legitimacy dilemma” of these courts (in Chapter 2), the main implications of the dominant model of constitutional justice for the problems of legitimacy (in Chapter 3), and the resulting relationships – and tensions – between constitutional courts and the legislatures (in Chapter 4). In Part II, I focus on the question of the protection of constitutional rights, beginning with a theoretical discussion (in Chapter 5) of the standards of rights protection that should be applied to the institution of judicial review. In the remainder of this section (Chapters 6-10), I discuss the main
PREFACE
XVII
categories of rights and the relevant questions and issues related to them, as tackled by constitutional courts in CEE. These chapters are largely descriptive: their aim is to give an overview of constitutional design of rights in CEE, and to identify the most important lines of interpretation that constitutional courts have brought to bear on these rights in the countries of this region. In the concluding chapter I attempt to bring the two main themes of the discussion in this book together, reflecting upon a general “score card” that can be given to these novel and in many respects impressive bodies established in CEE after the fall of Communism. A few caveats are in order. Firstly, the activities of constitutional courts are not, of course, limited to the articulation of constitutional rights, so the focus on rights adopted in this book may be seen, quite fairly, as based upon a selective appreciation of the actual role of these courts. As will be described in Chapter 1, they do much more than provide binding interpretations of the rights entrenched in constitutional charters. They adjudicate in conflicts between the supreme organs of the state; they clarify the specific meaning of the principles of the separation of powers and of the electoral system; they monitor the procedures of law-making and the compatibility of proposed laws with the international treaties and conventions ratified by a given state, and so on. Outside the law-making process sensu largo, they also often perform a variety of other functions, such as determining the accuracy of election results, deciding on the outlawing of political parties, impeachment, etc. This brief catalogue shows that the articulation of rights is only one part of the overall process of “constitutional justice”; it is, however, a very important (and, I would claim, the most important) part. Constitutional rights are at the very centre of the self-definition of a polity, and of the construction of the status of an individual vis-à-vis the state. Secondly, in providing an overview of the rights as articulated by constitutional courts in Part II of the book, I have been, of necessity, selective. Some rights are more fundamental or crucial than others, and I have deliberately left certain types of rights outside the scope of this book. One obvious omission is the set of rights related to the integrity of the judicial process; in particular, the rights of defendants in criminal trials. The constitutional courts in at least some of the countries discussed here have made an important contribution to the removal of some highly questionable provisions from the statutes regulating these rights, and the only reason for the exclusion of this issue here is the lack of space. This omission should not be seen as an under-estimation of the value of the role played by constitutional courts in this regard. Thirdly, this book attempts to present a sort of composite picture of constitutional courts in a large and diverse region. However, is it possible to include, in a single analysis, the “output” of some twenty courts operating in different states, affected by different traditions and politico-legal cultures, in countries that differ widely in so many relevant respects? My answer is: yes, it is, as long as one holds that the condition of “post-communism” creates a degree of commonality between all the countries thus described, and, at the same time, one is aware of the important differences between individual countries of the region. This is not the place in which to develop these two themes: about the commonality of some aspects of postcommunism and the divergent conditions that exist in different post-communist states.
XVIII
PREFACE
However, the fact that the examples of the judicial articulation of various rights provided in this book are taken from different countries does not imply that there is a single constitutional space in which courts in Albania, the Czech Republic, Lithuania and Hungary participate together in shaping a uniform regime of rights for a large, CEE polity. This would be nonsense. Rather, the aim is to draw a picture in which similar institutions in different countries of a region, which is still largely unknown to Western constitutional scholars,21 grapple with issues not dissimilar to those that are a staple of constitutional review in the West. Finally, I should perhaps disclose a hidden agenda that I had in writing this book. One of the fascinating issues in the period of the accession of eight CEE countries to the European Union (and on the eve of the accession of a number of other states of the region to the EU), is how the constitutional texts, structures of thinking and practices – in a word, the constitutional cultures – of new members states will fit with the common constitutional traditions of the older EU members. This notion of “common constitutional traditions” is not merely an academic and intellectual construct; it is also a term of art in the constitutional language of EU law.22 One crucial issue related to the eastward enlargement of the EU is whether this concept will retain any significance now that the EU has embraced a number of new members. What constitutional traditions, patterns of conduct and structures of thinking will they bring to bear upon the issues of constitutional rights for the citizens of the EU? Will they be partners in good standing in the sharing, understanding, and implementation of those rights that were agreed upon by the older member states, in the form of the EU Charter of Fundamental Rights? Elsewhere, I have sketched an answer to this question by showing that the constitutional texts in the then candidate states disclose very similar patterns of thinking about constitutional rights to those revealed by the EU Charter.23 However, this was only the beginning of an answer; a constitutional culture of rights may well begin with constitutional texts, but is certainly not confined to them. Much of a constitutional culture of rights can be illuminated by a study of the ways in which constitutional courts understand, interpret, articulate and develop (or limit) constitutional rights as outlined in the constitutional texts. This book attempts to describe this phase of the creation of a constitutional culture of rights: the phase of the articulation of those rights by constitutional courts. It may, therefore, be seen as a resource that helps to answer the question of whether, and if so, to what degree, the constitutional structures of CEE states can be aligned with the “common constitutional traditions” of the member states of the EU insofar as the constitutional rights of citizens are concerned.
PART I
CHAPTER 1 THE MODEL OF CONSTITUTIONAL REVIEW IN CENTRAL AND EASTERN EUROPE: AN OVERVIEW Before the fall of Communism, there existed only two constitutional courts in Central and Eastern Europe (CEE): in Yugoslavia from 1963 and in Poland from 1985.1 Although these were not exactly sham institutions, their position was far from strong enough to allow for the exercise of robust constitutional review. Quite apart from the legal definitions of their competence, the genuine powers of both were inevitably subject to the restrictions stemming from Communist Party rule. The position today, on the other hand, is that all of the post-Communist countries of CEE have constitutional courts and, while the effectiveness of these tribunals varies, they have made a strong mark on the process of constitutional transition. I will begin the analysis of this phenomenon with an overview, provided in this chapter, of the emergence of a (reasonably homogenous) model of constitutional review in CEE, and with a brief “anatomy” of that model: the powers of constitutional courts, the modes of activation of constitutional review, and the system of selection and tenure of constitutional judges. In the last section, I will describe the relationship between the constitutional courts and the ordinary courts, and in particular the Supreme Courts, characterised as it is by tensions and struggle for monopoly over privileged and final access to “constitutional wisdom”. All in all, the purpose of this chapter is to prepare the ground for a discussion of the legitimacy dilemma engendered by the emergence of robust systems of constitutional review in the post-Communist states of CEE. 1. THE EMERGENCE OF THE CURRENT MODEL The pre-World War II traditions of constitutional review in CEE were, as everywhere else in Europe, weak, but they were not quite non-existent; in some countries at least, the constitutional designers after the fall of Communism could have appealed to their own past in making a case for constitutional review. One country in which a Constitutional Court actually existed, drafted according to the Austrian “Kelsenian” model, was Czechoslovakia. The Court was set up by the Czechoslovak Constitution of 1920, and actually established in November 1921; it only had the power of abstract review of the constitutionality of laws. Interestingly, this power existed alongside the competence of ordinary courts to review the constitutionality of the laws they applied, which shows that the whole system “of judicial review was inspired simultaneously by the Austrian and American models”.2 The Constitutional Court’s jurisprudence was, however, extremely sparse. During the whole period of its existence, from 1921 to 1939, in addition to reviewing a number of parliamentary resolutions, it examined the constitutionality of only two laws (and these cases came as late as 1938), without even completing its proceedings
2
CHAPTER 1
in either case. As one contemporary Slovak scholar has commented, “the dominant practices of strong party government and majority rule limited the institutionalization of judicial review”3 in pre-War Czechoslovakia; “limited”, but did not eliminate altogether, thus providing at least a symbolic point of reference in for future constitutional designers in Czechoslovakia, and eventually in the Czech Republic and Slovakia. There were no other examples of abstract constitutional review in the countries of the region before World War II but judicial review in the form of concrete review by courts in the process of specific cases was not unheard of. In Romania, the power of judicial review of laws by the courts that had to apply them was recognised from the beginning of the 20th Century, despite the lack of a textual basis for this in the 1866 Constitution; a Bucharest court successfully claimed this power in the 1912 Bucharest Tram Company decision. Both inter-war Romanian constitutions (of 1923 and 1938) adopted the principle of concrete review exercisable by the Court of Cassation and Justice (the equivalent of the Supreme Court). However, any finding of unconstitutionality resulted only in the non-application of the challenged law to the particular case under consideration, and not in a general invalidation of the law.4 This power was exercised only occasionally but the very presence of such a legal possibility had a symbolic significance even outside Romania’s borders. Inspired by the Romanian example, a prominent constitutional scholar in Bulgaria, Professor Stefan Balamazov, put forward a constitutional draft in 1936 that would have had the Supreme Court of Cassation exercise a US-style judicial review. He could refer to his predecessors; a great Bulgarian legal scholar of the beginning of the 20th Century, Professor Stefan Kirov, had advocated allowing the Supreme Court to undertake judicial review in the process of hearing concrete cases. The legacy of Balamazov was so strong that, after the fall of Communism, some constitutional scholars looked sympathetically at his ideas, although it was ultimately the Kelsenian model that was adopted.5 Under Communism there was a universal rejection of the very principle of constitutional review, which was seen by official doctrine as inconsistent with the principle of the “supremacy of the national representative body … according to which the parliament is the supreme organ of state authority”.6 While the doctrinal rejection of constitutional review appealed to the principle of sovereignty of parliament, in reality judicial review was, of course, incompatible with the total political control exercised by the Communist Party. As announced by Professor Stefan Rozmaryn, a leading ideologue of constitutional law in Poland during the Stalinist period, constitutional control is “a reactionary and not a progressive institution”.7 In time, however, the severity of this hostility declined, and the first cautious proposals for a form of extra-parliamentary control of legislation were raised. This was consistent with the general calls for some institutional guarantees of legality in the “thaw” that followed the Stalinist period. It is significant, however, that any such calls were confined to procedural and formal control of legislation, and did not include demands for substantive control of laws in terms of their consistency with constitutional values.8 Obviously, such demands would have had a subversive potential, as even the Stalinist constitutions (which in many cases persisted into the post-Stalinist era) contained many impressive-sounding proclamations of civil and
CONSTITUTIONAL REVIEW
3
individual rights. Those in power could not risk a situation in which less-than-fully controllable judges would replace the official authoritative interpretations of these rights with their own. As a Polish student of the history of constitutional justice in Central Europe has noted, although the very idea of judicial control of laws ceased being viewed as a heresy at some time in the 1980s, it was nevertheless seen as a way of insuring the “quality” of statutes rather than as a fully-fledged constitutional review.9 The example of Yugoslavia, where the Constitutional Court had existed from 1963 onwards without undermining the doctrine of “the unity of state power”, was a model that could have been accepted by Communist states without any threat to themselves. The Federal Constitutional Court (mirrored by the constitutional courts in all of the republics making up the Socialist Federal Republic of Yugoslavia) was seen “as a guarantee of the proper functioning of the institutions but not as a mechanism of a protection of individual rights or relationships [between the component units] in the Federation”.10 In any event, any head-on clash between the constitutional courts and the ruling elite controlling the Yugoslav “assemblies” was, of course, unthinkable; as one Slovenian legal scholar has observed: “the functioning of these courts remained within the confines of the communist system and ideology”.11 Apart from Yugoslavia, the only other country in which the establishment of a constitutional court preceded the fall of Communism (with the exception of the abortive attempt in Czechoslovakia in 1968, when the Prague Spring reformers designed a strong Constitutional Court as part of their overall plans for constitutional reform)12 was Poland, where the constitutional amendment of 1982 (during the period of martial law!) announced the establishment of the Constitutional Tribunal. It was a somewhat premature announcement, though, as the statute on the Constitutional Tribunal – a prerequisite for the setting up of that body – was not enacted until April 1985. All other states of the region introduced constitutional courts only after the political transitions of 1989, beginning with Hungary, where a constitutional amendment of November 1989 was quickly followed by the relevant statutory regulation. In Czechoslovakia, a Constitutional Court was set up as a result of the constitutional amendment of late 1990 and a statute of February 1991. However, the only important case that this Court had time to consider (before the dissolution of Czechoslovakia) concerned the “lustration” law of October 1991, i.e. the law intended to prevent all former employees of the police, security forces and higher functionaries of the Communist Party from being employed in any high government posts.13 In Russia, even before the new constitution was adopted, the law on the Constitutional Court was passed in July 1991. The first Constitutional Court was set up in October 1991 and, contrary to widespread expectations that it would be a powerless body, it became a very active institution under the chairmanship of Valery D. Zorkin, heavily embroiled in the political controversies of the day. Legally, it acquired very wide powers, including the competence to initiate legislation and also a duty to exercise this power in the case of a breach of the Constitution due to a failure by parliament to legislate. In the same year, constitutional courts were established in Albania (by statute) and in Bulgaria and Romania (by new
4
CHAPTER 1
constitutions), and in the following years, in all of the remaining post-Communist states of CEE. In contrast to those early cases, the birth of constitutional courts in some other countries was quite protracted, as illustrated by the efforts in this regard in the Ukraine. A provision concerning the Constitutional Court was inserted into an amended Soviet-era constitution as early as October 1990, and a fully-fledged statute on the Constitutional Court was adopted in June 1992. Despite this, however, and despite the fact a President of the Court (Leonid Yuzhkov) was appointed in the same year, the actual setting up of the Court had to wait until after the adoption of a new law on the Constitutional Court in October 1996, due to the repeated failure of the Supreme Council to appoint the necessary judges.14 Equally tardy was the establishment of the Constitutional Court in Latvia where the appropriate constitutional amendment and a statute on the constitutional court were passed only in June 1996, and the Court met for its first session in December of that year.15 An altogether special case is the Constitutional Court of Bosnia and Herzegovina (BiH), established by the Constitution that in fact forms part (Annex 4) of the Dayton Accord of November 1995 (which entered into force on the 14 December of that year) on peace in Bosnia and Herzegovina. The Accord established an unusual, mixed “local-international” Court, with six judges from BiH (four elected by the House of Representatives of the Federation and two by the National Assembly of the Republika Srpska) and three external judges, with the proviso that the latter must not be citizens of BiH or from any neighbouring state, and must be appointed by the President of the European Court of Human Rights after consultation with the Presidency of BiH.16 While the typical trend in CEE countries has been towards the establishment and then gradual strengthening of the constitutional courts, one country has been an exception, and has gone from having a relatively independent (even though ineffectual) body to a “court” only in name, with grotesquely reduced powers and no genuine independence. In Belarus, the Constitutional Court set up under the 1994 Constitution had, by November 1996, ruled almost twenty acts of President Lukashenko unconstitutional; immediately after this date, however, the position of the Court was drastically weakened as a result of constitutional amendments (brought about by a referendum that the Court itself had found unconstitutional, but with no effect).17 While under the old Constitution all of the judges were elected by the parliament, under the terms of the 1996 amendments six judges are appointed by the President, and six by the Council of the Republic (the upper chamber of parliament). However, even the latter “election” remains a fiction because the candidates for the position of judge can only be proposed by the chairman of the Constitutional Court, himself appointed by the President. In practice, then, all of the justices of the Court are either directly or indirectly selected and approved by the President. Furthermore, the rules of standing to bring a challenge have been restricted, with the result that, in practice, only the President has initiated challenges to legislation. The President also has power to dismiss any judge, while the Court has been stripped of the power to declare that the president has violated the constitution. Put simply, the Constitutional Court of Belarus has been turned into a sham institution, which accurately reflects the current sorry state of democracy and the rule of law in that Republic.
CONSTITUTIONAL REVIEW
5
2. THE POWERS OF CONSTITUTIONAL COURTS AND INITIATORS OF THE REVIEW PROCESS Although there are certain local variations, one may attempt a description of the common model of constitutional tribunals in the region. The general model adopted is that of a ‘concentrated’ or ‘centralised’ constitutional review, conducted by a court composed of judges appointed for limited tenure by the political branches of government, exercising abstract, ex post and final review of the constitutionality of statutes and other infra-constitutional acts. There are, however, also some departures from the dominant model, which I will now briefly note. Centralised and concentrated review is understood here as an arrangement according to which only one institution in a given country has the right to authoritatively scrutinise laws in terms of their constitutionality. The task of constitutional review is conferred upon a special body established outside of the regular judicial system. The only, and minor, exception is Estonia, where the constitutional court is known as the “Constitutional Review Chamber” and is structurally a part of the National Court (the equivalent of the Supreme Court). This Chamber is elevated above the other chambers of the National Court (criminal, administrative and civil) in that its chairman must always be the chief justice of the entire Court. This special design, however, does not importantly affect the position of the Constitutional Review Chamber in the overall constitutional system and, for all practical purposes, the Estonian Chamber can be viewed as a separate constitutional court, like any other in the region. The most important power of constitutional courts, for the purposes of the discussion in this book, is their exercise of abstract judicial review, that is, the method of considering a statutory rule not in the actual context of a specific case but rather in abstracto. It is the textual dimension of the rule rather than its operationalisation in application to real people and real legal controversies that is assessed by judges. Most of these courts, like their counterparts in Germany, Italy or Spain, in addition exercise the power of concrete review initiated by other courts, which, when faced with doubt as to the constitutionality of a law that they are about to apply in the case before them, are obliged to suspend the proceedings and address their constitutional query to the Constitutional Court.18 The list of those subjects who are authorised to formally initiate abstract review varies from country to country. In all countries of the region, the president has such a power,19 and almost all grant similar competence to the government or prime minister.20 Many states also allow for the review process to be initiated by groups of parliamentarians, which may be defined either as a particular number (e.g. a minimum of 20 members of the Lithuanian Parliament) or a fraction (e.g. one-fifth of the members of the National Assembly of Bulgaria or one-third in Slovenia).21 Furthermore, a large number of CEE countries allocate this power to the Prosecutor General,22 Ombudsman (or an equivalent body),23 a Supreme Court,24 and/or central audit bodies.25 There are also a number of other assorted bodies that, in various constitutional systems, can initiate abstract judicial review, such as the executive and
6
CHAPTER 1
legislative bodies of constituent units of a federation,26 representative bodies of local government,27 and trade unions.28 This suggests that there is a spectrum of patterns of accessibility to judicial review. On one hand, there are very restrictive systems such as those of Estonia (where not even members of parliament can challenge the laws) or Lithuania (where only the government and a group of parliamentarians can initiate abstract review of statutes). On the other hand there are the systems that assure virtually unlimited access to this process in the form of actio popularis, that is, the right of each individual citizen to initiate abstract review regardless of their specific legal interest in the case in question. The greatest opening-up of such access has occurred in Hungary, where even non-citizens have the right to launch an actio popularis and where this possibility has been effectively used to invalidate a number of important laws. For example, the death penalty,29 the official use of personal identification numbers,30 and penalties for criticism of public officials31 have all been abolished as a result of constitutional review initiated through actio popularis in Hungary; the statistics show that, also in quantitative terms, actio popularis has been by far and away the most frequent way to initiate review in that country.32 It is for this reason that a (former) Chief Justice of the Hungarian Constitutional Court, László Sólyom, could recently observe, with only a little exaggeration, that the “actio popularis became a substitute for direct democracy”.33 Likewise, another Hungarian lawyer has noted that this method of triggering review “has made the citizens participants in the process of the transformation of the old legal system”.34 The most important difference between a “popular action” and a constitutional complaint is that the former presupposes no restrictive rules of standing. Indeed, the very idea of standing is inappropriate to an actio popularis because the individual who is launching such an action is seen as “a trustee of the public good” rather than as someone with a particular grievance who approaches the Constitutional Court as an instance of last resort. Restrictions on the right to introduce actio popularis (bringing it, in effect, closer to a classic constitutional complaint) have been discussed in Hungary for some time now, as a means of coping with what has become an almost unmanageable workload for the Court.35 The distinction between an actio popularis and a constitutional complaint (a complaint made by a citizen who claims that her constitutional rights have been violated by a particular rule, either directly or through an authoritative decision taken on the basis of that rule) is reasonably clear in theory but, in practice, may be somewhat blurred. According to one justice of the Constitutional Court of the Czech Republic, an actio popularis-type complaint can be smuggled in through the back door of a constitutional complaint when a petitioner complains about a particular decision and additionally claims that the law behind the decision is unconstitutional.36 When the process of constitutional complaint is available only against the legal rule in general and not against a specific decision in the application of the rule (as is the case in Poland), the only thing that distinguishes the complaint from actio popularis is the operation of the rules of standing. In Poland, the current position is that a complaint can be made against a specific statute (or other act) that formed the basis of a particular decision that has directly affected the petitioner. In other words, to prevent the transformation of the complaint into an actio popularis, there is a
CONSTITUTIONAL REVIEW
7
relatively stringent principle of standing: one cannot petition the court regarding a statutory provision “that only indirectly affects the legal position of a complaining person but has not been the basis for a final determination in a given case”.37 Further, in Poland a complaint cannot be made against a legislative omission or a law that is no longer valid.38 In addition, the complaint must be based on a violation of one’s constitutional rights and liberties; it cannot be founded exclusively on a violation of other constitutional provisions. In particular, constitutional complaints cannot be based on those articles that concern state policy goals (even if, formally speaking, they are in the Polish Constitution’s chapter on constitutional rights, e.g. art. 71: protecting the interests of the family), or on general constitutional clauses such as art. 2 (the principles of a democratic state based on law, and of social justice). This latter article can be cited in conjunction with a claim of violation of other rights but is not in itself considered to be an independent source of any constitutional rights or liberties.39 The important thing is that in Poland – in contrast to, for example, the Czech Republic – a constitutional complaint cannot be made against an alleged violation of constitutional rights in the process of the application of the law, but must rather be based on the alleged unconstitutionality of the law itself.40 This has increasingly been seen by the justices of the Constitutional Tribunal as a weakness of the current model of constitutional complaint; current Chief Justice Marek Safjan stated recently that the complaint procedure “does not meet the citizens’ expectations” because it does not provide any remedies for “a defective application of the law”.41 The pressure for change has, however, been resisted by the Supreme Court which fears that such an extension of the ambit of constitutional complaint would allow the Constitutional Tribunal to police its own decisions.42 The Hungarian position on individual constitutional complaint is similar in many ways. The target of the complaint cannot be the decision in application of a subconstitutional rule, but only that rule itself. Given that virtually the same result (invalidation of the provision in question) can be achieved in a simpler fashion by an actio popularis, it is no wonder that constitutional complaint has not been a particularly popular device there.43 In fact, the Constitutional Court attempted to enhance the complaint (“to breathe life into the lifeless institution”, as one commentator put it)44 by allowing to use a specific complaint to overturn a judicial decision which, as it held, had applied a regulation in an unconstitutional manner. However, in a pattern resembling the similar institutional conflict in Poland, it ran into conflict with the Supreme Court which protested that the Constitution did not provide that such a course of action should be available to the Constitutional Court, and that it was thus infringing the competencies of the regular courts.45 In contrast in Slovakia, the Czech Republic and Russia constitutional complaints are available against statutes “as applied”. This immediately raises the issue of a law that is about to be applied, and that the citizen fears is likely to violate his or her rights. For example in Russia, although the rules of standing applicable to such cases are rather unclear, there is the possibility of complaint by a citizen against a statute even before its actual application, as long as the statute is likely to be applied against that citizen. To ascertain this likelihood, the Constitutional Court requires some documentary evidence, “something in writing . . . such as a letter from an official”, that supports the proposition that the statute is about to be applied against the
8
CHAPTER 1
individual in question.46 The constitutional complaint procedure was introduced in Russia as early as 1991, by the Law on the Constitutional Court, and was further expanded by a 1994 law that stripped the Court of the discretion to refuse to hear citizens’ claims). While there have been some important decisions taken as a result of citizens’ complaints (for instance, the decision of 2 February 1999 suspending the death penalty throughout the Russian Federation on the basis that it violated the right of capital defenders to trial by jury, as very few regions in Russia had jury trials), nevertheless the popularity of this device is not great. As one legal scholar suggests, this relative unpopularity “may be explained by the weakness of the law enforcement system as well as by the general dissatisfaction (widespread among the Russian population) with the work of the Russian Government”.47 In either form, whether applying to the rule in abstracto or to its specific application, the possibility of constitutional complaint initiated by citizens exists in only about a half of the CEE states.48 There is a clear correlation between the existence of an activist, powerful constitutional court and the availability to citizens of a direct constitutional complaint procedure, which, of course, is not surprising. However, it must be said that in those constitutional systems that fail to offer their citizens this legal instrument, the views of lawyers and constitutional judges (including those who are certainly committed to the protection of constitutional rights) are not unanimously in favour of legal reform aimed at the introduction of such a procedure. The reasons behind this uneasiness are varied. For example, in Romania the (then) President of the Constitutional Court stated, in an interview, that the availability of constitutional complaint would clash with the principle that the Constitutional Court should be activated only when there is a “hot case”, as evidenced by the existence of a legal case before a regular court.49 In Bulgaria, one constitutional judge argued that, for practical reasons, it would not be a good idea for the Court to hear individual complaints: “We are only twelve judges. Practically, if you knew Bulgarian mentality, every case before a regular court would be appealed before the Constitutional Court to challenge the constitutionality of such and such an article. We have neither the money nor the apparatus to bear such responsibility”.50 However, even when such complaint procedures are not available, there may be some limited ways open to citizens to attempt to initiate judicial consideration of the constitutionality of a law. In Ukraine, for example, citizens may apply to the Constitutional Court for an official interpretation of the Constitution or of a statute; the application must state that a particular interpretation of the rule in question may lead to the infringement of that citizen’s rights. In practice, the Constitutional Court will consider such applications only if a citizen can show that, in reality, the application of a given statute is not uniform throughout the Ukraine.51 In Bulgaria, when the ordinary Bulgarian courts are confronted with the argument that a particular norm is unconstitutional, they may refer the issue to the Supreme Court of Cassation, which, in turn, may hold the proceedings in abeyance and file a petition to the Constitutional Court. To date, however, this procedure has not often been used. There is, finally, the possibility for some constitutional courts of the region to act sua sponte, that is, to initiate constitutional review themselves.52 As with many other
CONSTITUTIONAL REVIEW
9
aspects of constitutional courts’ activism, the Hungarian Court is the leading example here. Its self-initiated procedures, in particular with regard to the unconstitutionality of legislative omissions, have led it to make some important decisions in which the Court set deadlines for the Parliament to fill the relevant legal lacunae. For example, in a 1992 decision, it established that the Parliament had failed in its constitutional duty to enact a law regulating the broadcast media;53 in response, though with a considerable delay, the Parliament passed the Media Law in 1996, satisfying the Court’s requirements. In the early years of its existence, the Hungarian Court had made use of type of review quite frequently but, with time, its popularity among the judges diminished.54 In other countries, the right of selfinitiation of judicial review has had a more tentative status. In Russia, it existed in the “First Court” (1991-1993) but this power was removed by the 1994 statute on the Constitutional Court. Similarly in Poland, such a possibility existed under the old statute on the Constitutional Tribunal of 1985, but the new law on the Constitutional Tribunal (of 1997) extinguished this procedure, and in addition established explicitly (in Article 66) that the Tribunal is bound by the limits of a petition. The rejection of this power to initiate review, in the legislative process leading to the adoption of the 1997 law, was argued on the basis that the Tribunal should not be “politically activist”, and that the Tribunal would be vulnerable to an accusation of bias if its decision was not bound by the substantive parameters of a petition.55 However, it should be added that, even under the pre-1997 rules, self-initiation was considered an exceptional power, to be used only with regard to particularly important constitutional violations.56 As already mentioned, in Poland the Constitutional Tribunal is restricted by the limits of the original petition; in some other countries, however, the constitutional courts have the power to go beyond a petition and invalidate even those provisions of a statute under review that had not been challenged (this power may be seen, naturally, as a weak version of the sua sponte initiation of review by the court). For example in Slovenia, Article 30 of the Constitutional Court Act explicitly states that the Court “shall not be bound to the proposal given in a request or initiative [to evaluate a statute]”, and the Court has actually taken the opportunity to exercise this power. For example, when reviewing the 1995 law on associations which was challenged (successfully) on the grounds that it restricted the rights of minors by requiring their parents’ permission to join an association, the Court looked at the other provisions of the same statute and, on its own initiative, found the requirement of a minimum of ten members to form an association unconstitutional (it decided then that three was a sufficient minimum).57 Elsewhere, constitutional courts have assigned to themselves the power of going beyond the bounds of a petition; for instance, in Estonia – just as in Poland – the statute on the Chamber of Constitutional Review explicitly restricts the Chamber’s deliberations to the limits of the original petition.58 Nevertheless, the Chamber has occasionally noted the unconstitutionality of provisions other than those under challenge. For example, in a decision that struck down the 1996 law on non-profit associations because it restricted the rights of children (as the petitioner, the President of the Republic, claimed), the Chamber “considered it necessary” to observe that there was also another provision in the same act that was constitutionally faulty; namely the
10
CHAPTER 1
requirement of parliamentary permission to form associations possessing weapons or having a paramilitary character.59 Perhaps the fact that the Chamber invalidated the law as a whole, not just the defective provisions, renders less relevant the question of whether it confined itself only to the provision actually challenged or whether it did indeed go beyond the bounds of the petition. In the countries where the possibility of self-initiation is absent, the views on the lack of this capacity are mixed. In Bulgaria, one Constitutional Court justice, when asked about the idea of self-initiation, stated: “I think it could be quite risky here. … It could tempt political justices to participate in the political game”.60 For a similar reason, a Russian constitutional judge has also rejected the idea: “It would make us the judges in our own case”.61 However, in Romania, according to some observers this is the greatest weakness of the Constitutional Court. As one expert puts it, as a result of the inability to act sua sponte, the Court is absent from the major debates involving constitutional issues whenever the government and the opposition parties have no interest in raising a particular issue.62 Thus, Professor Parvulescu argues that “this is an essential power for a strong [Constitutional] Court”.63 However, the predominant doctrinal opinion seems to be to the contrary. According to Professor Mihai Constantinescu (himself a former Judge of the Romanian Constitutional Court), the availability of self-initiation undermines the legitimacy of the Court which should only be able to act when there is “a social conflict” as confirmed by a legal challenge by one of the parties to the conflict.64 “Have we abolished the power of kings only in order to replace it with the power of judges?” he asks in this context.65 He strongly criticises the evolution of the Constitutional Court which became, in his opinion, “a co-legislator” and not merely a “negative legislator”.66 Similarly, the (then) President of the Court, Lucian Mihai, considers the absence of the power of self-initiation of review to be the “correct solution”, which assures the Court’s “neutrality” and prevents the situation in which the Court could “select the issues taking into consideration the political situation”.67 According to him, “it is not for the Court to decide which issues are important”, as this is a task for the political branches of the state.68 In fact, the Romanian Constitutional Court did once attempt bravely, even if unsuccessfully, to establish for itself the authority to initiate review, by pointing at an unconstitutional omission and trying to set a deadline for the government to remedy this. In 1993, a panel of three judges: Justices Mihai Constantinescu (the same who was later to strongly criticise the idea of self-initiation), Antonie Iorgovan and Viorel Ciobanu found that, under Art. 135 of the Constitution (the right to property), the higher criminal penalties for crimes against public (as opposed to private) property were unconstitutional. They gave the Parliament six months to amend the Criminal Code. However, according to one report, the (then) Chairman of the lower chamber of Parliament, Andrea Nastase, convinced the remaining six judges to overturn the decision at a plenary session of the Constitutional Court, which thus, in effect, overturned it by a six to three majority.69 The plenary decision stated that the Court had no authority to tell the Parliament what to do. According to one observer, this was “a decisive moment in the struggle for the authority of the Constitutional Court vis-à-vis the Parliament and the government, and it was lost [by the Court]”.70
CONSTITUTIONAL REVIEW
11
For a similar reason – namely, concern over the intrusion of courts into the domain of legislators – the constitutional courts are not seen as being entitled to pronounce upon failures by the legislature to enact a law in most CEE countries (although, as we have seen, the Hungarian Court is an exception in this regard). Significantly, the (then) judge of the Polish Constitutional Tribunal explained, in a law review article, this inadmissibility of establishing unconstitutional omissions by arguing that, in principle, to make omissions subject to review “would require making political judgements”.71 However, the same Court did establish a narrow window of opportunity for itself in this field, by holding that, if a given issue has been regulated by statute, an objection of unconstitutionality “may apply both to what the legislators did in a given statute and to what they failed to do even though, in accordance with the Constitution, they should have regulated”.72 In Bulgaria, the Constitutional Court has occasionally declared “the lack of law” to be unconstitutional. One example was the 1995 budget law adopted by the (post-Communist) BSP-dominated parliament, which did not allocate any moneys for the Supreme Judicial Council, dominated then by the rival (liberal) UDF party.73 As a justice who participated in this decision notes: “We declared this omission unconstitutional arguing that it is the obligation of the Parliament to assure money for all constitutional organs to function normally and to fulfil their constitutional obligations”.74 The parliament duly complied with the Court’s decision and amended the budget law accordingly. In addition, some of the constitutional courts in the region also have the competence to establish a binding interpretation of the Constitution75 and/or of statutes.76 The power to declare a binding interpretation of the Constitution, at the request of authorised bodies,77 is vested in these courts quite apart from their function of examining the constitutionality statutes, and is seen as a natural complement to their overall guardianship of the Constitution; it is, incidentally, not an unusual function for some Western European courts to exercise.78 Some of the CEE courts have attempted to restrict this function somewhat, by establishing the requirement that there must be an existing dispute that has led to the uncertainties as to the correct interpretation of the constitution. For example, the Hungarian Court refused, in 1990, to consider a request by the Minister of Finance to interpret the scope of the constitutional right to social protection. This refusal was based upon the fact that the request had been motivated by the minister’s intention to use the Court’s opinion in the legislative process related to the bill on social protection, rather than the existence of a real controversy.79 Notwithstanding this limitation, the Hungarian Court has issued some important decisions in the process of abstract interpretation of the Constitution; for example, establishing that the Constitution cannot be amended by a referendum,80 as one of the political parties (the Independent Smallholders Party) had proposed in 1995. It seems natural that this function of interpreting the Constitution played a relatively important role in Hungary, where the choice was taken to heavily amend the old Constitution instead of creating a new one, thus perhaps giving rise to many doubts and uncertainties in the initial posttransition period, particularly in relation to the organisation of government, the legislative process, etc. At the same time, the request for an interpretation of the constitution can be used by the executive as a form of preventive review. As one
12
CHAPTER 1
observer notes, a number of requests for constitutional interpretation were made because “the executive wanted the constitutionality of specific regulatory concepts clarified before a draft law was introduced”81. In such cases, this process comes dangerously close to an ex ante clearing by the Court of a draft law. It is in this context that the restriction (noted above) on the admissibility of such requests was introduced by the Court itself. Those other courts that possess this competence have more rarely been asked to exercise it, even although there have been occasional important decisions issued under this procedure in Russia,82 the Ukraine83 and Bulgaria.84 The legal doctrine in these countries accepts, albeit often grudgingly, that the power of establishing a binding interpretation of the constitution amounts to a form of law-making by the court or of ex ante constitutional review (because as a rule, petitions for such interpretation are filed in the course of legislative work on concrete statutes),85 or to a precedent-creating role for the court,86 though this characterisation is sometimes mitigated by the proviso that these decisions are only “secondary” sources of constitutional law.87 This is not universally applauded; according to one Bulgarian constitutional scholar (and former Constitutional Court judge), the persistence of this power is “a legacy of Communism” when a “collective head of state” (the Presidium of the National Assembly) had the competence to provide a binding interpretation of the Constitution”.88 In turn, in those countries in which the constitutional courts do not possess this power, there are occasional attempts by other bodies to treat the court as the authoritative interpreter of the Constitution, even outside of the formal review process. In Romania, for example, some officials regularly ask the Constitutional Court to provide them with the “correct” interpretation of this or that constitutional provision.89 As an expert on the Romanian Court, Renate Weber, observes: “on each such occasion the Court . . . answered starting [sic] by saying that they were not allowed to make any comment . . . but ended by giving their opinion on the issue discussed!”.90 This, Weber explains, is due to the sense of selfimportance of the judges in question, who, regardless of the institutional restrictions in place, want to show that they are “the only ones who can give a professional authorised opinion” as to the correct meaning of the Constitution.91 One rather unusual power that the Constitutional Tribunal in Poland possessed between 1989 and 1997 was the binding interpretation of statutes. There were actually a number of important decisions taken by the Tribunal in this way, including on the interpretation of the election law, which effectively made the Senate elections possible by means a creative application of the Sejm (the lower chamber of Parliament) election law.92 In the process of preparing the Constitution of 1997 there was a degree of controversy over whether this procedure should be maintained; its opponents claimed that such decisions of the Tribunal become, improperly, an “extra-statutory source of law”.93 Those who supported the retention of this power, who included the then Chief Justice of the Constitutional Tribunal Andrzej Zoll, warned that its removal would mean the end of constitutional justice in Poland. In the end, the arguments of the opponents prevailed, and in particular the argument that the principle of the independence of the (regular) judiciary required that ordinary courts be able to provide their own interpretation of a statute when applying it, subject only to a formal appeals procedure.94
CONSTITUTIONAL REVIEW
13
These two types of prerogatives – binding interpretations of the constitution and of statutes – can be seen as offshoots of the central power of the constitutional courts which is to review (and, if necessary, invalidate) the laws it finds to be inconsistent with the constitution. However, for the sake of completeness, it should be added that the constitutional courts in the region, like many of their Western European counterparts, perform a number of other functions that are marginal from the point of view of the central focus of this book, but that nevertheless add to the overall authority and power that they enjoy. For one thing, most of these courts review international treaties prior to their ratification,95 although this can be still seen as deriving from the general power to review the constitutionality of laws. Secondly, some of the constitutional courts of the region have a formal role in the process of making or amending the constitution.96 Thirdly, in a number of CEE states, just as in Germany and Austria, constitutional courts have the power to decide on jurisdictional disputes between the highest institutions of the state, as well as between central and local institutions.97 Fourthly, and again similarly to the Austrian and German institutional design, some constitutional courts in CEE play a role in decisions on the constitutional liability of top officials, in particular the president,98 by ruling on impeachment.99 Fifthly, similarly to the German and Portuguese models, some of the CEE constitutional courts can outlaw a political party on the basis of inconsistency of the aims, statutes and/or the activity of the party with the constitution.100 Finally, some of the courts of the region have some specific functions with regard to the control of elections and referenda; for example, the Romanian Constitutional Court, following the example of the French Conseil constitutionnel, controls the procedures for the election of the President and certifies the result of the election, and performs similar functions with regard to referenda (but not with regard to parliamentary elections). A similar role is played by the constitutional courts in Slovakia, the Czech Republic, Lithuania and Bulgaria. As already noted, the dominant model of constitutional review in CEE is that of ex post review, that is, of laws already enacted, although there are some exceptions. In Romania, abstract review can be undertaken only in respect of statutes that have been adopted by the Parliament but not yet promulgated (although there is also the possibility of concrete review, initiated by regular courts, which by its very nature can only be ex post). This resembles the position of the French Conseil constitutionnel which can also review parliamentary acts only before promulgation. Furthermore, some other constitutional courts in the region (in Poland, Hungary and Estonia), in addition to their more routine, ex post review process, can be asked by the presidents to conduct an ex ante review of an act just passed by parliament. The Hungarian Court can even be asked to issue an advisory opinion on a bill not yet voted on by Parliament. There is, however, a marked tendency to view such ex ante reviews and advisory opinions as an exception rather than the rule.101 With the exception of Romania, decisions on the unconstitutionality of statutes in all CEE countries are final; there is no way of reversing the verdict other than by constitutional amendment. In Romania, verdicts of the Constitutional Court resulting from abstract review, conducted prior to promulgation, can be overridden by a twothirds majority of both chambers. In Poland, a similar possibility existed until the Constitution of 1997 introduced the concept of the finality of all Constitutional
14
CHAPTER 1
Court decisions.102 3. THE TENURE AND SELECTION OF JUDGES Judges of constitutional courts are appointed for a limited tenure, usually for nine years.103 The prevailing pattern is that they can be reappointed once, although in a number of countries there is an explicit ban on reappointment.104 The exception is the Constitutional Review Chamber in Estonia the members of which are appointed until retirement (at 68) – a reflection of the position of this body as one of the chambers of the Supreme Court. Another exception is the Constitutional Court of Bosnia and Herzegovina, where the first judges appointed were to serve a 5 year term, with all subsequently appointed ones to serve until the age of seventy. With very few exceptions, constitutional justices tend to be either legal scholars (with a marked preponderance of constitutional law professors) or senior members of the “regular” judiciary. The length of tenure (and the possibility of re-appointment, or lack thereof) is related to the question of judicial independence; it is frequently asserted that life tenure (perhaps only subject to a compulsory retirement age) guarantees the highest degree of independence, while appointment for a set period with the possibility of re-election has the opposite effect (because, it is claimed, “[I]f judges can run for a second term, their independence in the first term is severely hampered and they will want to be popular among their nominators”).105 The system of set (preferably, reasonably long) terms without the possibility of re-election is seen as falling in between.106 As virtually none of the CEE constitutional court systems opted for life tenure,107 which may be seen as a natural consequence of seeing these courts as de facto “third chambers” (or “second chambers”, as the case may be), it is tempting to assert that long set periods without the possibility of re-appointment produce better circumstances for ensuring the independence of judges than those systems that do offer the chance of renewal. This, in any event, is the view that is often expressed by the specialists. A Polish scholar (and former judge of the Constitutional Tribunal) has claimed that the prohibition against re-appointment is an important guarantee of independence “because it prevents [the judges] from soliciting favours of the Parliament and politicians”.108 Does the possibility of reappointment (as exists, for example, in the Czech Republic) indeed create incentives for such “solicitation of favours” from politicians? Speculation of this kind cannot be simply ignored; indeed, it is actually supported by some knowledgeable observers of the Czech Constitutional Court.109 The following anecdote from the Czech Republic may be an illuminating illustration of this: at a recent conferral ceremony for new judges of the Constitutional Court, lead by President Havel, only five of the current Constitutional Court judges attended. This was interpreted by a leading Czech politician as a possible sign of opportunism by those judges who chose not to attend; the attitude being “Who knows who the new President of the Republic will be, and how he will view my presence at ceremonies chaired by President Havel?”110 It is generally considered a matter of constitutional error in the Czech Republic that the possibility of reappoint-
CONSTITUTIONAL REVIEW
15
ment has not been excluded explicitly.111 What of judges who do not have enjoy the possibility of reappointment, and yet who attempt nonetheless to secure a good position for themselves after their term at the court? Professor Leszek Garlicki concedes that the prospect of a post-court career “may. . . encourage them to seek connections with politics before their leaving the Constitutional Court” but quickly adds that “this is rather a question of personal integrity and character”.112 However, it is hard to see how this is any different from the situation of judges who can be reappointed; after all, is the question of whether they will try to win favours from the powers that be in order to secure a second term in office not also a matter of “personal integrity and character”? In fact, judges who come to the end of their term at the constitutional court and yet are well before retirement age can be quite conscious of the fact that their future may be shaped by politicians, and this may contribute to their political dependence. This may be particularly visible with regard to those judges who do not come on secondment from universities, and who therefore do not have the natural and guaranteed prospect of going back to an educational and research career. This is, for instance, the case in Bulgaria where a relatively low percentage of constitutional judges (in comparison with other CEE countries) are academics.113 As one of the Constitutional Court judges there told me, “this [phenomenon of trying to please political forces in order to assure good post-tenure prospects] happens continuously. . . It depends on the self-esteem of a given judge, and also on his plans for his future career… but most of the judges here think this way. It is understandable”.114 He added that he himself was not “thinking this way” because he had a guaranteed position as a university law professor waiting for him upon his retirement from the Court. Another Bulgarian constitutional expert has argued, for the similar reasons, that the judges of the Constitutional Court should not be young; younger judges tend more to “think about their future and to maintain contacts with the political parties that nominated them to the Court”.115 Similarly, in the Ukraine (where the tenure system is the same: 9 years, non-renewable) a Constitutional Court judge openly admitted that, although such a tenure system is “quite good for those who will retire” soon after serving at the Court, “our younger judges think about what they will be doing after” and “must adjust themselves to the political forces”.116 For these reasons, he believes that the system of tenure until retirement (say, at 65) would be optimal from the point of view of guaranteeing judicial independence. Such a system once existed in Russia and, even though life tenure (until retirement) has been abolished, a very long set tenure (15 years) helps attain a similar result; even judges who are relatively young at the point of selection will be close to retirement age at the end of their tenure, and so the incentive to please politicians in order to assure a post-judicial career is minimised. The appointment process of judges to constitutional courts in CEE is thoroughly political although “high legal qualifications” (or an equivalent description) are usually listed as one of the criteria of eligibility. In some countries of the region, constitutional judges are appointed through a process that requires the participation of both the legislative and executive branches. This is often referred to as a “collaborative” system of appointment (similar to that used in the US) in which the parliament (or one of its chambers) elects the judges from a group of candidates
16
CHAPTER 1
chosen by the president,117 or the president appoints the judges from a list nominated by the parliament.118 The second model is a ‘split’ appointment system, in which different bodies, usually the president, the parliament and the body representing the judiciary, each have their own “quota” of positions to fill on the constitutional court; this system is based upon that adopted in Italy, where each of these bodies elects one-third of the Corte costituzionale.119 The third possibility is that of the exclusive competence of the parliament to elect the judges of the constitutional court.120 Again, an altogether different system was adopted in Estonia, where all members of the Constitutional Review Chamber are professional judges and are appointed in exactly the same manner as all the other judges of the National Court, i.e. by the Parliament at the proposal of the President of the National Court (who is obliged to first consider the opinion of the National Court en banc concerning a candidate). Each of these systems of selecting constitutional court judges creates different dysfunctionalities and incentives for politicians to shape the composition of the court to suit their own political needs. Perhaps the most obvious risk occurs when the parliament has the full power of appointment, as in Poland, Hungary and Croatia. As a Polish constitutional scholar has observed, this generates “a risk of excessive politicisation of these [appointment] decisions” and the “temptation of a system of “spoils” may become too irresistible for parliamentary majorities”.121 Indeed, the practice of the selection of constitutional judges in Poland shows that the process has been overwhelmingly controlled by the parliamentary majority of the day. Out of twenty-nine judges elected to the Tribunal in 1989-2002, in twenty-seven cases the elected candidates were those who had been supported by the then parliamentary majority (even though in most cases the opposition also nominated candidates).122 This shows that “the majority parties, regardless of their political provenance, believed that they did not have to share decisions concerning appointments with the Opposition”; as a result, “it became increasingly apparent that each change of political configuration of the Sejm will be reflected in future appointments”.123 In Hungary this effect was somewhat mollified. The fact that the nominating committee is composed of an equal number of representatives of each party eliminated the dominance of the stronger parties at the crucial early stage of the process, and the requirement that judges are elected by a two-thirds majority vote of the parliament (rather than by an absolute majority, as in Poland) meant that a degree of consensus was required.124 In the third country that adopted the system of exclusively parliamentary appointment, Croatia, the fact of concentrating the selection process in one body has made it possible for politicians to control the process. As one Croatian expert observes with regard to the election of the eight judges who took up their posts on the 7 December 1999: “The election procedure was in fact politically predetermined, allowing bloc vote for all the judges without examination of and discussion on their individual professional competences. Needless to say, [the] majority of them were publicly very well known [sic] members of political parties”.125 In fact, they were not all members, or sympathisers, of one party. Even though, in the opinion of Croatian experts, the then ruling party (the Croatian Democratic Union (CDU)) could have filled all eight vacancies at that time, they decided to compromise and “to leave some [positions on the Court] to other parties”,126 in particular to the Social Democratic Party and to the Croatian
CONSTITUTIONAL REVIEW
17
Social Liberal Party, largely as part of their strategy to be on good terms with the opposition. However, this does not deny the fact that the CDU had full control over the process from the beginning to the end, and that the deal between the CDU politicians and the opposition leaders took the form of a shrewd give-and-take within the respective party leaderships.127 In turn, when the appointment process is managed by different bodies acting separately from one another, one danger is that each of the bodies will elect “their own” judges who will then be held under an obligation to be loyal to their appointing body. This can be seen, for example, in the Ukraine, where appointments are made by the President, the Parliament and the Council of Judges who each have six positions to fill. In effect, those who were appointed by the President are likely to be seen as “his” judges and, according to one expert, are largely “loyal to the President” and “support him”.128 According to a judge of the Ukrainian Constitutional Court, this is not the case but at the same time he believes that a collaborative appointment process would be a better system (e.g., elections by the Parliament from a list proposed by the President).129 Still another expert, and a former judge of the Ukrainian Constitutional Court, believes that the distinctions between the three groups of appointees are not in terms of their institutional loyalty but in terms of their professional background and qualifications; the President tends to appoint law professors, the College of Judiciary chooses judges, and the Parliament tends to opt for politicians.130 According to this expert, the real split is between those judges who are “competent” and those who are “incompetent” (the implication being that the Parliament-appointed judges belong to the latter category). Another example is that of Bulgaria where the party-political colours of the judges are also evident. As one of the Justices reports, in the last term of the Court there were 3 or 4 justices who always voted against any legal challenges made by the BSP (then in opposition) to laws sponsored by the UDF Government, except when the unconstitutionality of the laws was absolutely obvious.131 As Professor Nenovsky, another (ex-)Justice of the same Constitutional Court states, “we do not have a situation like in Germany when different parties discuss with each other about the new justices in order to maintain a proportional representation [of parties] on the Court. In Bulgaria it is different. If the President and the parliamentary majority represent the same parties, as it is now [May 2001], they can select very politicised judges to the Court”.132 This is what happened, according to him, in Autumn 2000 when four vacancies were filled with judges characterised as “reasonably good jurists but very strongly connected to political parties”.133 This is said to have caused “a very big surprise in the [legal] community”. Professor Nenovsky’s preferred method of preventing the recurrence of this situation is the adoption of a rule similar to that currently applied in Austria, to the effect that whoever has performed political functions in the four years preceding the elections of judges should be ineligible for admission to the Court.134 Furthermore, it must be added that at least one-third of the Court (the justices elected by the two Supreme Courts) can be, and actually have been, largely apolitical. In addition, those judges appointed by successive Bulgarian presidents have thus far been selected not on the basis of their party-political affiliations but rather because they enjoy a good personal rapport with the President, thus creating what one observer has called “a
18
CHAPTER 1
presidential power base” within the Court.135 In fact, the allocation of a “quota” of judges to different branches of the government may be largely illusory if one branch has strong control over the others; if, for example, the president dominates the parliament, and the constitution does not (through, for example, a requirement of qualified majority) compel the parliament to obtain opposition consent to the nominations. This is the case in Romania where the President and the two chambers nominally have the power of appointing three judges each; in fact, however, the perception is that the central role in appointing all of the judges is played by the President, and, more specifically, by his legal and constitutional advisors.136 This de facto domination by the President over the entire composition of the Court has been pushed to an extreme (which cannot be described other than as an aberration) in Belarus. Here, the President appoints one half of the Constitutional Court, and has the exclusive power to appoint its President; the other half is appointed by the Senate, which is, however, also dominated by the President, who nominates one third of its members! Finally, US-style systems in which different bodies control different stages in the process of admission may well turn out, de facto, to be a process controlled almost exclusively by one of the branches, usually the executive. This is the case in Russia, where the President has almost full control of the candidates (the Duma “may” also propose candidates but the President has the final say over the composition of the list). Parallel to this, the Parliament has a negative power in that the Federation Council must give final approval for appointments; a power that it has exercised at times, for example in rejecting many of President Yeltsin’s candidates, thereby “creating a long and arduous appointment process”.137 As one of the current judges of the Constitutional Court of the Russian Federation, Mr Vladimir Yaroslavtsev, describes, his own appointment to the Court on 25 October 1994 was preceded by a lengthy process managed by the presidential administration.138 In the end, President Yeltsin’s decision was announced to Mr. Yaroslavtsev in a manner not unlike the announcement of victory to the winning candidate in a competition for a valuable prize.139 Significantly, much of Mr Yaroslavtsev's narrative, developed during an interview, creates the perception that the President's administration was aiming to engender a feeling of gratitude in him.140 It has to be added, however, that the Federation Council voted against those of Mr Yeltsin’s candidates who were perceived as being too close associates of his. At the selection round during which Mr Yaroslavtsev was appointed, President Yeltsin proposed six candidates for the six vacancies at the Court; only three were approved at that time, and the remaining three were appointed one-by-one during the next four months. The picture emerging from this account is of a certain game between the President who fully controls the pool of candidates and wants to have as many loyal members appointed as possible, and the Federation Council which fully controls the actual appointment stage and will oppose those whom they see as being too close to the President. Neither party can achieve the goal of appointing a judge without the co-operation of the other. In consequence, in order to avoid a stalemate (which serves neither of the institutional players) they must be, in the end, “reasonable”. The President must choose candidates that may be seen as relatively “neutral”141 and the Council must not undermine the President’s desire to have judges whom he trusts.
CONSTITUTIONAL REVIEW
19
4. CONSTITUTIONAL COURTS’ PURSUIT OF A MONOPOLY OVER CONSTITUTIONAL ADJUDICATION All CEE countries have established a power of judicial review to be exercised by the constitutional courts, and all of the constitutional courts have been granted the authority to conduct such review in abstracto, irrespective of the concrete cases to which the law might have, or actually was, applied. Most of the constitutional courts of the region also have the task of considering requests from ordinary courts of law to examine the constitutionality of laws that the (ordinary) courts are obliged to apply in the consideration of concrete cases. It should, though, be added that concrete control, initiated by regular courts, is still very rarely practised in CEE; overall, the courts are the least frequent initiators of constitutional review of statutes.142 There are a few exceptions to the adoption of this German-type model (characterised as it is by a combination of the powers of abstract and concrete review).143 These exceptional cases apart, the dominant model indicates a neat division of labour: ordinary courts merely apply the law, and if the law seems questionable all they can do is to ask the constitutional court for its advice as to the constitutionality or otherwise of the law in question. This division is, however, neat in theory only, and the reduction of the “ordinary courts” (including the supreme courts) to the role of docile supplicants who should defer to constitutional courts whenever they have any doubts as to the constitutionality of a law immediately raises problems of the mutual relationships between these two categories of court. As an American constitutional theorist has noted: “Systems that divide legal authority between a constitutional court and a supreme court face co-ordination problems when allocating jurisdiction and resolving inconsistencies in rulings”.144 Such a division also raises problems of professional pride, the sense of dignity and, last but not least, institutional competence. After all, it is the ordinary courts that are at the front line, so to speak, of the application of the law, and whether a given law is constitutional can be best detected from the perspective of its application in specific cases. Specific cases are the business of ordinary courts, and it would thus seem that they should be entrusted with the power to decide on the constitutionality (or otherwise) of the laws that they are obliged to apply. This, in brief, is the standard rationale for granting the power of concrete judicial review to the ordinary courts. I will discuss the abstract/concrete review dilemma more specifically in Chapter 3(2); here, however, my purpose is to give a brief account of the institutional conflict (between constitutional and ordinary courts of the region) generated by the constitutional courts’ monopoly over constitutional adjudication. This monopoly is not a necessary consequence of their existence; one could reconcile the role of constitutional courts in the conduct of abstract review with the role of ordinary (including supreme) courts in setting aside laws that they find unconstitutional in the course of the consideration of concrete cases. As Gábor Halmai proposed with respect to the Hungarian practices of judicial decisionmaking:
20
CHAPTER 1 When a judge, identifying a conflict between the fundamental law and the legal instruments he is supposed to apply, lets himself be guided by the former, he is only applying section of Article 77 of the [Hungarian] Constitution . . . which describes the fundamental law as a system of norms which is binding on everyone.145
In practice, however, the attribution of the power of abstract review has created an incentive for constitutional courts to insist that they alone have the rights to review laws in terms of their constitutionality. The reverse side of the coin has been a (more or less successful) denial of the power of ordinary courts to set aside laws that they find to be unconstitutional, even though they are supposed to apply the constitution directly. How can one reconcile the duty of the courts (of all courts) to apply the constitution directly with the constitutional court’s monopoly over declarations of unconstitutionality? Direct application of the constitution, in conjunction with the principle of the supremacy of the constitution over ordinary laws, necessarily implies the right (indeed the duty) of an institution to refuse to apply any law found by it to be unconstitutional. Whether this refusal must be followed by a direct application of the constitution by that very court, or should rather lead to a referral to a constitutional court that has the last word whenever a doubt as to the constitutionality of a law arises, became the central issue in the interinstitutional conflicts that emerged in CEE in connection with constitutional review. In these conflicts a familiar pattern occurred in almost all the countries of the region.146 Ordinary courts (led by the supreme courts) fought for the right to set aside laws they found to be unconstitutional, while the constitutional courts claimed a monopoly over the power to make such findings. Russia provides a good example of such a conflict; it resulted in a victory for the Constitutional Court which fiercely fought against granting regular courts the power to make their own declarations as to the unconstitutionality of statutes. It maintained that the only avenue open to courts in such cases was to stay the proceedings and address the Constitutional Court in the form of “concrete review”. This doctrine, developed by the Constitutional Court, was a reaction to the “rasjasnienie” [clarification] by the Supreme Court of the 31st October 1995,147 addressed to all the courts of the land. The Supreme Court urged all judges that if they were to come across, in the process of taking a decision in a specific case, a statute that they find to be unconstitutional, they should disregard it and apply the Constitution directly instead. The Supreme Court explicitly stated that “the court, in determining the matter under review, will apply the Constitution directly . . . whenever a judge becomes convinced that a federal law . . . is in conflict with the Constitution”. The option of seeking the opinion of the Constitutional Court on the question of constitutionality was recommended in cases “of imprecision relating to the question of conformity, or lack thereof, with the Constitution of a law which has been, or is to be, applied to an actual matter”.148 The clear implication of this clarification was that the process of concrete review by the Constitutional Court should be activated only when an ordinary court has doubts. In contrast, if the court was convinced about the unconstitutionality of a statute, it could simply disregard the statute and apply the Constitution directly. In other words, it could give effect to its own understanding of the true meaning of a constitutional provision as applied to a concrete case before it.
CONSTITUTIONAL REVIEW
21
However, the Constitutional Court reacted by asserting its monopoly over judicial review. In a decision of 16th June 1998, taking the form of a binding interpretation of the Constitution, it reminded the courts that it was itself the only body competent to decide upon issues of constitutionality, either in abstracto or in concrete cases.149 Also, acting extra-curially, some Constitutional Court Justices strongly reacted against the Supreme Court’s view and reasserted, in no uncertain terms, that even when an ordinary court is convinced about the unconstitutionality of a statute it has a duty to petition the Constitutional Court rather than to refuse to apply the law in question.150 It is only with regard to sub-statutory acts that the regular courts may directly apply the Constitution rather than the act.151 A justice of the Constitutional Court in Russia explained the reasons for the strong rejection of the principle that ordinary courts could apply the Constitution directly and, in the process, declare statutes unconstitutional in the following manner: “We would reach the situation that, in one place, a court would declare the statute invalid and, in another place, the same statute would be found valid. … We would have disorder. … In such a big country we need to have a centralised system and this role is played by the Constitutional Court”.152 However, not all Russian Constitutional Court justices think in this way. It is worth noting the remarkable opinion expressed by Justice Gadzhiev in his dissent to the June 1998 decision: Decisions of the ordinary courts which have identified a conflict between a law and the Constitution, and have declined to apply the law, without repealing them [sic], represent the birth of a judicial law, the development of which is particularly indispensable to the Russian legal system in its search to avoid positivist approaches.153
It should be added that such a pattern of conflict is not specific to CEE courts only. One may observe a more general trend that whenever constitutional courts have been established in post-authoritarian contexts, a pattern of conflict between these courts on the one hand, and the supreme courts (plus other ordinary courts) on the other, has emerged. This was the case, for example, in Spain soon after the establishment of the Constitutional Court,154 and also in Italy and Portugal. Such conflicts are understandable. A new, powerful and popular institution, endowed with a lot of political prestige and eager to make use of self-righteous rhetoric, is seen with distrust and perhaps envy by the less glamorous, more established institutions, including the Supreme Court. On the other hand, the supreme courts and the rest of judiciary are viewed by constitutional courts, composed largely of academics (often former dissidents or opposition activists in the authoritarian era), as the vestiges of the ancien régime, unwilling and incapable to behave in the new spirit of democratic and liberal values prevailing after the transformation. Hence, a territorial conflict develops, in which both institutions try to enlarge their own powers and to restrict those of their rival. Often, this conflict leads to a real “war of the courts”, in which the supreme courts seek to simultaneously establish their own right to declare statutes unconstitutional (and also to establish such a right for lower courts) and to limit the authority of the judgements of the constitutional court vis-à-vis the judiciary. The latter aspect became particularly pronounced in the Czech war of the courts.
22
CHAPTER 1
The “war”, apart from its legal and constitutional implications, also had a clearly political dimension. The Constitutional Court, perhaps uniquely among the constitutional courts in CEE, was, since its inception, homogeneously anticommunist in its political make-up. All of the judges were appointed almost at the same time, between July 1993 (when the first twelve Judges were selected) and January 1994 (by which time the remaining three had been appointed), by President Vaclav Havel, with the consent of the Parliament in which the Civic Democratic Party (of the then Prime Minister Vaclav Klaus) enjoyed a majority.155 There was therefore no need to try to strike a compromise between the forces of the ancien régime and the new post-transition elite (as in Hungary). Neither was there a gradual, staggered cycle of appointments resulting in a politically heterogeneous court (as was the case, for example, in Poland). The judges on the Czech Constitutional Court felt that they belonged to a relatively cohesive democratic bloc, and they considered the Supreme Court as belonging to the old apparatus of the authoritarian regime. Conversations with the judges of the Constitutional Court and with its observers clearly confirm the prevalence of this self-perception. For example, one of the justices described the judges of the Supreme Court (at least, at the time of the “war of the courts”) as belonging to the “old structures”, by which he meant that they were rooted in the previous, communist establishment. 156 The legal dimension of the conflict in the Czech Republic was related to the binding power of Constitutional Court decisions taken in the process of hearing constitutional complaints from individual citizens. The Supreme Court maintained that such decisions did not bind the ordinary courts (including the Supreme Court) in their future decision-making, and further that ordinary courts were not included in the category of “public authorities” that, according to the constitution, are bound by the decisions of the Constitutional Court.157 In turn, the Constitutional Court held that it had the authority, within its constitutional-complaint jurisdiction, to review the final decisions of all authorities, including the courts, insofar as they may breach the constitutional rights of Czech citizens. The immediate confrontation arose with respect to a decision on conscientious objection, in which the Supreme Court refused to consider the argument used by the Constitutional Court in arriving at its judgment (as opposed to a mere statement of the verdict) as being binding upon it.158 To be more precise, the controversy between the two courts centred on three inter-related issues. First, whether the Constitutional Court could review the decisions of courts (this being an element of the debate over whether courts are “public authorities” in the constitutional sense). Second, whether decisions of the Constitutional Court are binding upon the future decisions of all courts, and not only in the specific case decided. Finally, whether the decisions are binding only insofar as the statement of the verdict is concerned (which is a statement of unconstitutionality only a few sentences long) or whether the reasoning (ratio decidendi) is binding as well, as the Constitutional Court maintained. The controversy was pushed to the point that the Czech Constitutional Court, in one of its decisions, accused the Supreme Court of violating the Constitution by refusing to decide a specific matter in conformity with the law as enunciated by Constitutional Court jurisprudence.159 The Constitutional Court eventually prevailed on all three of these issues. However, at least according to some academics, there was no clear legal basis for
CONSTITUTIONAL REVIEW
23
granting it the authority to establish norms binding upon the future decisions of all courts. According to these critics, the ratio decidendi should properly have a binding effect only upon the Constitutional Court itself, and a merely “persuasive” effect upon all other courts.160 Against this point of view, a justice of the Constitutional Court has argued that the confinement of the binding power to the statement of verdict only would lead to a paradoxical result in those cases in which the Court upholds a statute but at the same time prescribes a constitutionally required interpretation. It would render such decisions superfluous, because the constitutionally prescribed interpretation is contained in the justification of the decision, not in the statement of the verdict.161 This, according to Justice Holländer, “would also force the Constitutional Court into a course of action which, in its consequences, would appear absurd and unsustainable: not to rely on the possibility of a constitutionally conforming interpretation, to abandon the principle of judicial restraint, and to nullify a contested enactment whenever there is the least possibility of it being interpreted in a constitutionally non-conforming manner”.162 Regardless of this, the level of compliance by the ordinary courts with Constitutional Court precedents is less than perfect; as Justice Holländer concedes, “In its practice, the Constitutional Court is frequently confronted with cases in which ordinary courts refuse to decide in conformity with the proposition of law declared by the Constitutional Court”.163 Similarly in Poland, the Constitutional Tribunal has repeatedly rejected the proposition that the duty of regular courts to apply the Constitution directly justifies those courts’ prerogative “to disregard statutory regulations” deemed unconstitutional. It has urged that the only avenue open to courts in such cases is to address a question to the Tribunal.164 As it declared in one of its early decisions: “A judge, while being subordinate to the Constitution, is not thereby relieved of subordination to a statute”.165 The combination of the principle that only the Tribunal can declare laws unconstitutional with the principle of the presumption of the constitutionality of statutes, effectively denies the regular courts any role in deciding to disregard a statute due to its unconstitutionality. As a senior Constitutional Tribunal judge (indeed, its Vice-President) recently declared in a scholarly article, to allow the “regular” judges to review the constitutionality of statutes would lead to situations in which a judge could, for example, impose a 40-year jail sentence or establish taxation rates on the basis of his own understanding of the constitutional principle of social justice.166 While this example is quite misleading, the fact that such a “parade of horribles” was given as an illustration of the perceived consequences of universalised judicial review indicates how strenuously the principle is rejected by constitutional judges, and how strongly they defend their exclusive authority to review the constitutionality of laws. This doctrine of the exclusive competence of the Constitutional Tribunal on matters constitutional is not universally accepted in Poland.167 As one justice of the Tribunal has admitted, “the problem of a ‘division of powers’ between the Constitutional Tribunal and the other parts of judiciary [has] acquired now a very serious character”.168 Clearly, the Supreme Court of Poland believes that all ordinary courts are authorised to declare whether a statutory provision is inconsistent with the Constitution, and to refuse to apply it. This view is seen as resulting from the combination of two constitutional rules: that of the direct applicability of the
24
CHAPTER 1
Constitution (art. 8 (2)), and the principle of judicial independence and, more relevantly, their “submission only to the Constitution and statutes” (art. 178 (1)). The only limit upon this power is the binding nature of the decisions of the Constitutional Tribunal. If the Tribunal has already declared a given statutory provision constitutional, an ordinary court (including the Supreme Court) cannot declare it otherwise. It is significant, however, that, even in this latter case, the subordination of ordinary judges to the Tribunal was questioned by the regular judiciary. The Supreme Court recently had to deal with a “legal question” addressed to it by a judge who inquired whether the courts are bound by “affirmative” decisions of the Constitutional Tribunal, on the basis that only the invalidating decisions can be seen to be generally binding decisions in the constitutional sense.169 The Supreme Court declined to go down this path, and confirmed the equally binding nature upon all courts of “affirmative” decisions of the Tribunal. Nevertheless, the very fact that such a doubt was expressed by a judge indicates the persisting unease within the judiciary relating to the monopolistic role of the Constitutional Tribunal in terms of the articulation of constitutional norms. However, in the absence of a specific decision by the Constitutional Tribunal, the power of judicial review of the constitutionality of statutes by all judges is considered by the “regular” judiciary to be a natural corollary of the direct binding effect of constitutional provisions.170 As the Chief Justice of the Polish Supreme Court recently observed, ordinary courts do, after all, routinely conduct a “positive control of constitutionality of laws”, namely, when they decide not to lodge a constitutional question to the Tribunal, “and yet, such decisions may also in practice lead to inconsistency between the case law of the Supreme Court and of the [Constitutional] Tribunal”.171 Quite apart from the doctrinal controversies, it is a fact that courts of lower instance do occasionally refuse to apply laws that they consider unconstitutional, on the basis of the constitutional principle of the direct binding force of the Constitution, proclaimed in Article 8(2). The same applies to the Supreme Administrative Court; as its President recently wrote, his Court has on a number of occasions refused to apply a statute (for example, some articles of the Customs Code) on the basis of its unconstitutionality, upon ascertaining that the Constitutional Tribunal had previously invalidated similar statutory provisions.172 In these cases, according to the President of the Administrative Court, it was not necessary to lodge a “legal question” to the Tribunal. At a certain point in the Polish “war of the courts” the Supreme Court achieved a victory in the battle. Under the pre-1997 constitutional rules, the Constitutional Tribunal had the power to provide a binding interpretation of the law (not just of the Constitution but also of statutes). The Supreme Court considered this to be an anomaly and thought, in particular, that it conflicted with its own power to interpret the meaning of a law as applicable in a given case. In the end the arguments of the Supreme Court prevailed; in the 1997 Constitution the power in question was removed from the Constitutional Tribunal. As the Chief Justice of the Supreme Court (appointed to this position only after the 1997 constitutional change) remarked on the background to this controversy: “Many judges of the Supreme Court had not only expressed doubts as to the merits of certain interpretative decisions [by the Constitutional Tribunal] (particularly if they differed from the jurisprudence of the
CONSTITUTIONAL REVIEW
25
Supreme Court), but they had also questioned the binding force of such interpretations upon the courts, including naturally upon the Supreme Court”.173 However, the tension did not end there. Ever since the loss of this particular power, the leading justices of the Constitutional Tribunal (including its Chief Justice) have continuously complained about this “defect”, demanding the restoration of their competence to provide binding interpretations of all laws.174 At the same time, the justices of the Supreme Court have warned that giving in to such a demand would inevitably lead to “a new conflict between the Constitutional Tribunal and the Supreme Court”.175 Similarly, the constraints upon the individual constitutional complaint procedure, which in Poland is restricted to the abstract review of a law that has formed the basis for a particular, rights-affecting decision (rather than the actual application of the law in question), have been a source of tension between the two courts. The Supreme Court fears that removal of this restriction would place the Constitutional Tribunal in the position of a super-appellate court, with the competence to control, and overturn, “final” decisions of the Supreme Court.176 The Chief Justice of the Polish Supreme Court recently warned that to allow the Constitutional Tribunal to rule on potentially unconstitutional applications of laws would imply “the entry of the [Constitutional] Tribunal into the field of concrete decision making”. He noticed caustically that it would mean that Poland would have “the Supreme Court and also the ‘Truly’ Supreme Court”.177 Indeed, the choice of a model of constitutional complaint restricted to control of the unconstitutionality of laws (and not of the application of those laws, as is the case in some countries, such as the Czech Republic) was partly a result of the “fears of representatives of the [Supreme Court] that the [Constitutional Tribunal] may become a judicial body empowered to control the decisions of the [Supreme Court]”.178 5. CONCLUSIONS Notwithstanding the virtual absence of any tradition of judicial review in their preCommunist histories (not to mention the abhorrence with which Communist constitutional theory viewed judicial review), all post-communist states of CEE have established institutions and processes of constitutional review of considerable homogeneity: they have all opted for a system of centralised, ex post and abstract review of statutes, with no possibility of appeal from the constitutional court’s decision. As noted, there have been some departures from this paradigm: for example, almost all countries of the region provide for the possibility of a “concrete” review of a given provision, in conjunction with a specific court case (always, however, in addition to, rather than replacing, abstract review); some constitutional courts can review bills before they enter into force; and there have been occasional departures from the principle of the finality of the decisions of these courts, etc. Nevertheless, the exceptions to the norm have been rare, and the similarities clearly outweigh the differences. The powers of the different courts to control the constitutionality of statues vary somewhat in terms of the authority to identify constitutional omissions, to act sua
26
CHAPTER 1
sponte (and to go beyond the bounds of the original petition), or to review the constitutionality of the application of statutes as opposed to the abstract content of the statutes themselves – but again, these are differences pertaining to the margins, not the core, of the process of judicial review. More significant divergences exist in relation to the definition of who has standing to bring a challenge before the court, and in particular whether individual citizens have the right to lodge a constitutional complaint; there is, however, quasi-uniformity on the conferral of the right of initiation of proceedings upon top executive officials and groups of MPs. The tenure of judges is limited but relatively long (although there is a lack of uniformity regarding re-appointment), and the systems through which they are appointed are thoroughly political: such appointments are either made exclusively by parliaments, through a process of collaboration between the executive and the parliament, or by allocating a quota of seats on the court to be filled by different bodies (normally the president, the parliament and a body representing the judiciary). This institutional design has led to a concentration of the power to interpret, articulate and apply the meaning of constitutional norms in a prominent body of high public visibility, setting it quite evidently apart from the judicial system. This, naturally, created tension in terms of the interaction between and interdependence of the judicial system and the new institutions for judicial review; a tension fuelled by the apparent monopoly granted to constitutional courts to articulate with finality the “correct” meaning of the relevant constitution. This, in turn, led to the “wars of the courts”, which occurred, along very similar lines, almost everywhere in the region. This “war” was not only about institutional self-aggrandisement; it also concerned the much more fundamental principle of the direct application of the constitution by all courts, and consequently their power to set aside laws that they consider unconstitutional, making in the process direct appeal to their own understandings of the constitution. Such an approach has, however, been strongly resisted both by the constitutional courts and by the predominant part of scholarly commentary on this issue: the power of the “ordinary” courts to conduct their own binding constitutional interpretation has been seen as inconsistent with the logic of the system of judicial review as adopted in CEE. But can the monopoly claimed by constitutional courts, to articulate the true meaning of the constitution, be reconciled with their claims to legitimacy as judicial (or quasi-judicial) bodies? Is this zeal in asserting a privileged insight into constitutional wisdom not a good reason to reconsider the grounds for the adoption of the concentrated/abstract model of judicial review in the first place? These two questions will be pursued further in the next chapter.
CHAPTER 2 CONSTITUTIONAL COURTS IN SEARCH OF LEGITIMACY
As Robert A. Dahl has observed, There is necessarily an inverse ratio between the authority of the quasi guardians1 and the authority of the demos and its representatives. . . . Even if the authority of the guardians were restricted solely to certain questions of fundamental rights and interests, on these matters the demos would necessarily alienate its control. . . . The broader the scope of rights and interests subject to final decision by the quasi guardians, the narrower must be the scope of the democratic process.2
This commonsensical observation illustrates immediately the nature of the fundamental dilemma related to the legitimacy of “quasi guardians”, i.e., constitutional courts, when exercising the power to invalidate democratically enacted laws on the basis of their own understanding of constitutional rights. The nature of this dilemma, and various different attempts to address and resolve it, will be discussed in some detail later in this chapter. I will begin by outlining the contours of the problem, focusing on what is at stake in the controversy over the legitimacy of judicial review, namely, the perception of the objectivity of ascertaining the “true” meaning of constitutional norms, and the decision as to the best possible institutional devices in terms of gaining access to that objectively valid meaning. I will then examine a certain paradox faced by constitutional courts, namely, that their best means of defending their legitimacy to articulate the meaning of constitutional norms lies in conceiving of themselves as quasi-legislative institutions, a characterisation that the courts themselves strenuously resist. I then revisit the reasons normally provided in support of the introduction of an abstract/concentrated (“Kelsenian”) system of judicial review in the postCommunist states of CEE, and trace the legitimacy dilemma to the insufficiency of these grounds to supply convincing arguments in favour of such a system. One rationale, however, stands apart from the others, and thus deserves more serious consideration: the suggestion that constitutional courts as set up in CEE are protectors of minorities and minority rights. The status of this argument will be considered at the end of this chapter. 1. THE LEGITIMACY DILEMMA The standard defence of the legitimacy of constitutional judicial review proceeds on the basis of the very nature of constitutionalism: it is seen as a deliberately counter-
28
CHAPTER 2
majoritarian device, i.e., as a constitution-based constraint upon the majority rule exercised by parliament or, as the case may be, through mechanisms of direct democracy. From this perspective, an external, extra-majoritarian institution is required to make sure that the behaviour of the elected branches of government is in conformity with constitutional constraints. With respect to constitutional rights, the argument appeals to the need to observe rights as a democratic demand in itself, no less important than the need to give effect to majority preferences. The reconciliation of majoritarian politics with respect for rights is achieved by the counter-balancing of parliamentary rule with the power of the constitutional court to invalidate legislation. So the standard argument goes. An obvious response to the standard argument as a justification for judicial review is to argue that the very meaning of rights, as applied in specific circumstances, is a matter of deep controversy within society, and that this controversy may be replicated, rather than resolved, in cases of disagreement between a majority of parliament and a majority of the constitutional court: why privilege the latter by giving it the last word on the meaning of rights? Moral disagreement among reasonable persons of good faith about the correct articulation of rights seems to indicate that there is no “canonical” meaning of any particular rights, and that they are merely shorthand ways of referring to a bundle of entitlements that correspond to certain values. As people disagree about the proper balancing of those values, they will also disagree as to the “correct” meaning of any particular right, even though they may all agree about the worth of a “right” when stated in its abstract, and necessarily vague, constitutional form. From this perspective, any decision to empower constitutional courts to invalidate statutes under the rights provisions of the constitution is seen, at best, as a pragmatic institutional arrangement, but which is prima facie questionable because it needs to defeat the arguments that privilege the legitimacy of parliaments to issue the laws for the societies that they represent. This is not to say that judicial review cannot be defended on the basis of a nonobjectivist theory of rights (that is, a theory that denies the existence of a canonically correct articulation of rights); only that it must proceed in a pragmatic rather than a principled fashion, that is, it must appeal to the institutional qualities of the relevant bodies entrusted with the power to pronounce the “last word” on the articulation of a constitutional right. In addition, the argument for judicial review from this theoretical perspective must overcome particularly high argumentative hurdles resulting from a general presumption in favour of the paramount authority of parliaments. Judicial review is, however, usually supported by an objectivist theory according to which the correct meaning of rights is objectively discernible by human reason, with the correct institutional incentives optimising the circumstances in which the ascertainment of the right meaning is likely. As an example, consider the following description of a “moral realist” judge offered by Michael Moore: When a moral realist judge today invalidates the expression of majority will that a statute presumptively represents, he does so in the name of something beyond his power to change and beyond the power of a societal consensus to change. . . His justification for judicial review is straightforward, and so is his mode of practising it: he will seek to
IN SEARCH OF LEGITIMACY
29
discover the true nature of the rights referred to by building the best theory he can muster about the nature of equality, the nature of liberty, etc.3
There is an air of circularity in Moore’s reasoning: he opts for a “moral realist” account as one which “can make sense of some of our adjudicatory practices”4 in a way that competing theories cannot. Consequently, some of our practices (including the practice of judicial review) are used to argue for moral realism (because in the absence of moral realism they would not be defensible); but, on the other hand, it is moral realism that supports those very practices because, if the rights can be discerned at the level of objective moral reality,5 then judicial review acquires its much needed legitimacy. In other words, if rights exist independently of the mind, and at the same time they form a part of a valid (legitimate) constitutional system, then the body that can best access this reality is eo ipso legitimate. If, on the other hand, we adopt a position that may be called “constructivist”, that is, that there is no objective articulation of general rights available to human reason but rather that such articulations are constructed in political practice, then the argument for strong judicial review is much more difficult to sustain. From this standpoint, it is not appropriate to show that courts are in a better position to grasp the “correct” meaning of constitutional rights, precisely because this very meaning is constructed through social practice and validated in a discourse on fundamental values for which rights themselves are simply a shorthand formulation. A constructivist who also favours judicial review needs to show why, in such a discourse, the courts should have the last word (or nearly the last word, subject to the possibility of constitutional amendment). By contrast, the moral realist has an easier task: she will discharge the burden of argument that the courts are legitimate in invalidating legislative choices as inconsistent with constitutional provisions of rights if she can make a compelling case that the institutional circumstances under which constitutional courts operate lend themselves better to the correct discernment of the objective meaning of rights. As these rights, in their abstract, constitutional formulation, are by definition legitimate (they are legitimate by virtue of the legitimacy of the constitution), so are those specific articulations of rights that are more likely to approximate the “correct” meaning in any given case. As moral realists believe in the existence of such objective correctness, it is clear that moral realism (even if only implicitly presupposed) is of great help in formulating an argument in favour of judicial review. The dominant discourse about constitutional rights in CEE reveals an implicit adoption of a moral realist position, whereby the constitutionality or otherwise of a particular piece of legislation can be shown with a high degree of confidence, supported by an “objective” reality of rights. Such an understanding is presupposed, at least implicitly, by those who argue for strong powers of judicial review for constitutional courts. Consider, as partial evidence, the debate in Poland surrounding the power of the parliament to override, by qualified majority, the invalidating decisions of the Constitutional Tribunal; the power that existed until the 1997 Constitution put an end to it.6 Not surprisingly, the Constitutional Tribunal (and the supportive academic writers) was openly hostile to the override power of the Parliament. At the stage of constitutional drafting, the justices of the Tribunal often
30
CHAPTER 2
demanded the abolition of this legislative capacity, and then warmly applauded its eventual discontinuation. What is significant here is the type of contention that was put forward in demanding the abolition of the parliamentary override: the dominant argument was that such a practice rendered it possible for “unconstitutional” laws to be reaffirmed by the legislature as valid. The argument did not concern the comparative institutional advantages gained by allowing the court to resolve disagreement over the proper meaning of constitutional rights; rather, it took for granted that the very fact that a law had been declared unconstitutional by a Court was conclusive evidence of its “objective” unconstitutionality. In a statement typical of that debate, the Chief Justice of the Polish Constitutional Tribunal, Professor Marek Safjan, declared (after the constitutional possibility of a legislative override has been already terminated) that it was ironic that, in a state that recognised the supremacy of the Constitution and its binding character upon all state institutions, “the parliament [could] affirm the validity the laws which are inconsistent with the Constitution”; he thus applauded the discontinuation of the override possibility as “a final victory of the Constitution over politics and recognition that nothing can justify keeping unconstitutional legal provisions within the legal system”.7 This view can be held only if we assume that the meaning of constitutional provisions (including rights provisions) is discernible in an objective fashion,8 so that we can declare with certainty what is and what is not unconstitutional. This in turn presupposes an approach based upon moral realism. Incidentally, we may note that the recourse to objectivism (that is, to the notion that there is a stable meaning of constitutional provisions that is accessible through legal interpretation and that transcends actual moral disagreement over conceptions of the good) is a universal argumentative device used by courts all over the world, and that it is a very important ingredient of the self-legitimating rhetoric employed by the judiciary. The German Federal Constitutional Court has long declared that constitutional guarantees of the rights contained therein rely upon “eine objektive Wertordnung”: an objective order of values.9 Courts – and in particular constitutional courts – make frequent appeal to such notions in order to distinguish themselves from political institutions, the latter embroiled as they are in moral and political battles, and to gain support for the authority of their decisions. In a recent comprehensive study on social support for the highest courts in different countries, James Gibson and his collaborators establish beyond any doubt that the popularity of courts increases as public knowledge of them grows.10 The explanation that Gibson et al. provide is that, while “ordinary people who know little about courts have few reasons to believe that judges make decisions differently from any other politicians”, in contrast, those who are attentive to courts adopt a different view, which is not, however, the view of legal realists: “Greater awareness is associated with the perceptions that judges are different, that they rely on law not values in making decisions, that they are ‘objective’”.11 This view, which of course contributes significantly to the legitimacy of the courts, is largely learnt from the courts themselves: those who support the judiciary do so largely because they are “exposed to a series of legitimizing messages focused on the symbols of justice, judicial objectivity, and impartiality”. 12 Therefore, at the level of public opinion, as at the
IN SEARCH OF LEGITIMACY
31
level of argumentative structure, there is a strong link between an objectivist account of the judicial articulation of the constitution and the legitimacy of judicial review. However, to say that the argument for strong powers of judicial review is made easier by the acceptance of a (controversial and questionable) position of moral realism does not mean that such an argument is compelling. For even if we accept (for the sake of argument) that the correct meaning of abstract constitutional rights is discernible in an objective fashion, we still need to show what it is about constitutional courts that makes them more likely to discern the true meaning when they disagree with the parliament (or, more precisely, when a majority of the court disagrees with a majority of the parliament). There are two principal types of arguments that attempt to do this: a negative argument based on a distrust of legislatures, and a positive argument based on the deliberative nature of constitutional courts. These two arguments are independent of each other: the argument from distrust does not hinge upon the deliberative ideal (we can distrust an institution for reasons other than that it is non-deliberative), and, on the other hand, the expectation of deliberation is not necessarily based on the trust that perverse incentives will not affect a given institution. I will return to the deliberation argument in Chapter 5 and here will address only the first argument. It is straightforward: it claims that we cannot expect our democratically accountable representatives (and those directly dependent on them) to produce a fair articulation of constitutional rights; and that it was precisely this distrust that provided the grounds for “constitutionalising” rights in the first place. The actual reasons for this distrust may have to do with various incentives that act upon the democratically accountable politicians and that are not conducive to the best articulation of vague constitutional rights. In particular, those incentives may favour the oppression of the minority by the majority, because there are not enough votes to be gained in supporting minority causes; and it is precisely the protection of minority against majoritarian oppression that constitutes one of the main rationales for constitutionalising human rights.13 (Note that, contrary to some simplistic interpretations, the argument from distrust is not a version of the “nemo iudex in res sua” precept, which is sometimes presented in the form that those who made the law should not sit in judgement on constitutionality thereof. The invocation of this principle in the context of scrutinising laws in abstracto in terms of constitutional rights is a mistake, for the reasons so convincingly identified by Jeremy Waldron).14 First of all, it needs to be noted that what matters is how trustworthy one institution (or one set of institutions) is compared to another institution (or another set of institutions) in its actual operations.15 It is no good to compare a realistic, unwholesome account of a legislature with an idealised model of a constitutional court. Whether or not we can trust that one particular institution more than another will strive to articulate human rights, rather than pursue the self-interest of its members, depends on a great variety of factors. Most of them (though not all)16 are of an institutional character; that is, they are related to the formalised patterns of screening, selection, accountability, length of term, revocation etc. of those who people the institutions. For example, while a limited term in office with no possibility for reappointment may promote self-serving behaviour consisting of adjusting one's action to post-term career, a limited term with the possibility of re-
32
CHAPTER 2
appointment may promote the self-serving behaviour of trying to ingratiate oneself with those political agents (or citizens) who have the greatest influence on renomination and re-appointment. Similarly, life tenure may promote a disregard for changing social values and perceptions regarding the articulation of a particular right; specific professional or competence-related conditions for appointment may promote various types of déformation professionnelle; whereas transparency of official proceedings leading to an authoritative articulations of rights may increase the importance of good reputation (the avoidance of public shame) as a motive for behaviour and thus an impediment for self-serving conduct (but may also, under less favourable circumstances, engender demagogy and quest for popularity); and so on. There is a long list of institutional variables that produce different types of incentives, each of which may produce dishonesty, self-serving conduct, myopia or irrationality. Different constellations of these institutional variables – different institutional designs – and their corresponding incentives may affect differently our judgment concerning the comparative “trustworthiness” of one institution vis-à-vis another; and there is no universal reason to believe that representative legislative institutions are necessarily affected by perverse incentive-creating factors to a higher degree than any extra-political institutions, such as constitutional courts. In this context, it is useful to recall Philip Pettit’s distinction between two different strategies in institutional design: the deviant-centred strategy and the complier-centred strategy. The former presupposes that people are likely to cheat whenever they can do so with impunity, and so the institutional design is focused on the elimination of pathologies; however, in the process, it fails to provide optimal incentives for the “non-knaves”.17 The complier-oriented design, on the other hand, presupposes a more optimistic view of human nature; namely, that most people are not knaves. Therefore, it tries to maximise the opportunities for valuable action, and, although it also provides for some sanctions against knaves, it does not focus all of its attention on the prevention and punishment of knavish action. These two strategies correspond to two very different sets of specific “screens” and “sanctions” (to use another useful distinction by Pettit) and, of course, both have their advantages and disadvantages. It may be the case that within one and the same system, the relative proportion of deviant- v. complier-centred strategies varies from one institution to another, but these proportions will also vary from country to country. For example, election laws in different countries may reflect different approaches towards deviant- v. complier-centred strategies. As a result, in some countries we will have stronger reasons to suspect members of political institutions of behaving in a self-serving way, and weaker reasons for harbouring such suspicions in others. In any event, distrust is a more contingent basis for strong judicial review than many of its proponents seem to presuppose. For consider: if we thought that the majority was inherently unable to respect and honour the legitimate interests of minorities and individuals (the interests that we consider as important enough to warrant their elevation to the position of constitutional rights), and that this is the reason why we need a counter-majoritarian body to ensure the legislative respect for constitutional rights, then we would be incapable of understanding how constitution-making (including the adoption of a bill of rights) is possible at all. After all, it is the majority that ultimately adopts the constitution – a
IN SEARCH OF LEGITIMACY
33
qualified majority, and a majority acting in a special way, but a majority nonetheless.18 And if we never trusted the majority to be able to consider, in good faith, the legitimate interests of the minority, then we could never have a genuine bill of rights in the first place. If, on the other hand, there are some circumstances in which we can trust the majority to conduct proper rights determination (partly because we have no other choice), then this opens the way to trusting the majority in other circumstances as well – as long as these circumstances resemble to a significant degree the circumstances which supported the trust in the first place (i.e., the circumstances of constitution-making). It is true that there are important differences between constitution-making and ordinary lawmaking, but the differences are of degree rather than of kind. To draw a sharp contrast between the majority deliberating on the constitution and the majority deliberating on the statutes (including those which would impact upon constitutional rules) would be particularly suspect in those constitutional systems that have a recently adopted constitution, and therefore where the authors of the constitution are largely identical with the parliamentary law-makers. We may, of course, retain a healthy scepticism as to the elected lawmakers’ motivations when they vote in favour or against a particular rights-implicating statute. We may suspect, with a high dose of realism, that they do it not only, or not principally, because they genuinely believe that their vote is guided by the best interpretation of the constitutional right in question, but rather because they want to pander to their electorate (or comply with their party leaders who, in turn, want to pander to the party’s electorate) in order to secure re-election. But this, in itself, is not the sort of distrust that should persuade us to look elsewhere for an institution (such as a constitutional court) that would be less susceptible to such perverse incentives, because what really matters is what type of motivations lie behind those electorates to whom the parliamentarians will “pander”: are they guided by their interests or by the values which underlie their understanding of rights? The answer is, both; but the motivations will vary from one area to another: voters may be more guided by their self-interest when it comes to tax law, and more by values when it comes to the law on abortion. In the latter case, the “pandering” to the electorate by the legislators assures an indirect impact of values upon the legislation; there is therefore no neat correlation between interest representation and the parliament on the one hand, and value-articulation and the court on the other.19 2. CONSTITUTIONAL COURTS BETWEEN THE JUDICIAL AND LEGISLATIVE BRANCH As indicated in the Preface to this book, it is indisputable that the constitutional courts in the region discussed here enjoy a high level of social acceptance, despite occasional disagreements with and criticisms of particular decisions. They do not, therefore, have a problem with “legitimacy” in the sense of a general public acceptance of their authority to do what they are doing – including, the invalidation of statutes.20 Further, these courts do not have a problem of legitimacy in the formal and institutional sense of the term, which may be understood as compliance with the
34
CHAPTER 2
constitutionally recognized limits and working under constitutionally defined standards. They do not, as a matter of routine, exceed the powers granted to them by the respective constitutions, by the statutes on constitutional courts or by other relevant laws of their respective jurisdictions. Even if one disagrees, on the merits, with this or that decision, one must be careful not to frame the criticism in terms of a charge that the court acted ultra vires. The charge that a court decides on the (allegedly improper) grounds of the political or moral preferences of its judges, as opposed to the (allegedly proper) grounds of inconsistency with the constitution, is a statement that reflects, rather than stands outside of, the substantive disagreement as to the wisdom or otherwise of a particular decision. Whether the court’s decisions are genuinely based on constitutional principles rather than the judges’ own policies and moral values is in itself a controversial matter, and the level of this controversy is no different from the controversy of the wisdom (or otherwise) of any other political decisions. To be sure, even if a constitution or a statute that governs a particular institution confers upon that institution very broad powers, it does not necessarily follow that the institution is always acting wisely by exploiting those powers to the limit – and note that I am not suggesting at this point that the CEE courts behave in such a manner. The range of powers conferred by a constitution (or a statute) is in itself a matter for interpretation, and that interpretation may be narrow or expansive. But even the expansive interpretation of one’s own limits does not run an institution into a serious legitimacy problem, in the institutional sense of legitimacy. It is a matter of wisdom, or prudence, but not of legitimacy. This is what Bruce Ackerman had in mind when, commenting on Justice Sólyom’s perception of the role of his Court as a “free-floating problem-solver”,21 he observed: I cannot say that Solyom’s conception is inconsistent with the enormous authority granted his court in its governing statute. The question is whether it is prudent to make use of power that will lead to political catastrophe. Surely, it is well within the court’s capacity to construe its statutory jurisdiction narrowly.22
And yet, from the mere fact that the court remains intra vires and does not violate the formal, institutional rules concerning the legitimacy of its decisions, it does not follow that the court’s actions are unproblematic from the point of view of legitimacy in a broader, critical sense of the word. The question then becomes not: “Is the court authorized to take these types of decisions” but rather, “Should the court be authorized to take them?” Should the only precaution against the exercise of an “enormous authority” (to use Ackerman’s words) be our faith in the judges’ prudence? The question of the democratic legitimacy of an institution is not exhausted by the fact that the institution acts within the constitutionally established limits, and that the constitution itself has been enacted democratically; there is no contradiction in terms when one claims that a constitutionally and democratically established device is undemocratic.23 It is a commonplace that a democratic procedure for establishing an institution does not necessarily confer a democratic character on the institution itself. A democratically established constitutional convention proceeding in a democratic manner may decide to establish a nondemocratic, or imperfectly democratic, institution. The degree of democracy that the
IN SEARCH OF LEGITIMACY
35
constitutional convention wishes to infuse into the institutions that it is about to set in motion is in itself a matter of free choice, if the convention is to be truly democratic. Of course, the distinction is not always as neat and as clear as the statement just made implies; for one thing, it builds upon a clear dichotomy between the constitution-making process and the political (or legislative) process.24 Such dualism is not always easy to sustain in practice, as long as we understand the constitution not merely as a formal document reached at a particular point in time, but as a process whereby the true meaning of constitutional norms emerges from the practices, conduct and behaviour of various actors in a given constitutional polity. However, to the extent to which there is some dualism between constitutional and political rules, i.e. to the extent to which we can distinguish between the decisions made by various political actors and the rules by which those decision-making actors are bound, we can also talk about the distinction between the democratic pedigree of a rule (or system) and the democratic (or otherwise) content of such a rule (or system). And this is not just a matter of theorizing: it is obvious to any observer that a community may, as the result of democratically constituted debate, agree about the need to undertake undemocratic measures in some circumstances. The constitutional rules concerning various states of emergency are just one example of a structure which is non-democratic in character (and self-evidently so) but which may have been established democratically. Just as there is no necessary connection between a democratic procedure for setting up an institution and the democratic character of that institution, so there is no necessary connection between the undemocratic nature of an institution and its illegitimacy. A central bank, a civil aviation authority, the army or a national opera company are not “democratic” institutions (and this is not merely in the sense of internal decision-making process, but, more importantly, in that their specific acts, or sometimes even whole sequences of acts, do not track the actual distribution of social preferences); this, however, does not render them illegitimate. More relevantly for our purposes here, ordinary courts are not, and are not meant to be, democratic institutions; and yet this fact, in itself, does not adversely affect their legitimacy. The main source of their legitimacy, as Martin Shapiro famously argued in his classic study on courts, derives from the “triadic” model in which two persons decide to call upon a third, neutral umpire, in order to resolve their disagreement.25 Shapiro argued further that “the substitution of law and office for consent” which distinguishes courts par excellence from go-betweens, mediators and arbitrators, creates an important tension between the social logic of a triad (which is a source of legitimacy of a court) and the actual operations of particular courts.26 In particular, Shapiro argues that the courts’ involvement in public law, their exercise of social control and their lawmaking functions significantly weaken their triadic, legitimising structure. And yet, it is Shapiro’s thesis that courts, as we know them, are not qualitatively different from more obviously triadic institutions (such as mediators); they “are simply at one end of the spectrum rather than constituting an absolutely distinct entity.”27 The need to elicit some traces of consent (illustrated, for instance, by courts’ reluctance to decide in the absence of one of the parties), their frequent pursuit of a compromise, and many other mediating components in judging, render them simply one species of a broader family of triadic institutions.
36
CHAPTER 2
It is important not to overstate Shapiro’s point: much of his argument is devoted to showing that the traditional “prototype of courts” is not reflected in the actual operations of judicial bodies. And yet, it is important to retain his general conclusion that it is precisely the departure from the triadic structure that is a source of possible weakness of judicial legitimacy. “[F]rom [the triad’s] overwhelming appeal to common sense stems the basic political legitimacy of courts everywhere,”28 asserts Shapiro, but then “[c]ontemporary courts are involved in a permanent crisis because they have moved very far along the routes of law and office from the basic consensual triad that provides their essential social logic.”29 This tension between courts’ claim to legitimacy and their non-triadic patterns of operation is further magnified when the procedure abandons all pretences of adjudicating between conflicting interests of two parties, and focuses instead on an abstract scrutiny of a legal text. If the scrutiny is unrelated to any particular conflict between two parties, the “triadic” sources of legitimacy of courts disappear altogether. This is the predicament faced by those constitutional courts whose functions include abstract judicial review. One could perhaps try to argue that some remnants of the triadic structure are still present: there is a complainant (usually, the representatives of the outvoted parliamentary minority, or of the President), a respondent (the representatives of the parliamentary majority, or of the government), and a neutral umpire, namely the judges of the constitutional court. This analogy is, however, inappropriate. The “triad” that underpins the prototype of courts is not constituted by two parties disagreeing about what social norms should be properly turned into law, and a third party who resolves their dispute, which is precisely the case of adjudication by a constitutional court. The conflict which is the stuff of a triadic judicial resolution revolves not around some abstract ideas concerning rights and wrongs, but rather concerns the claim that one party’s interests have been impermissibly (under the existing, valid rules) violated by another. A better analogy to the conflict that lies at the heart of abstract judicial review is that of disagreement between the majority and the opposition over what law or policy is best for their society, subject to general and underspecified constitutional provisions. Indeed, this is precisely what is at stake in the discourse pertaining to the abstract constitutional review of legislation; and in this discourse, the constitutional court is unable to rely on the argument that all it is doing is applying the existing law because it is precisely the “rightness” (in terms of general constitutional standards) of the law that forms the subject of the controversy. As Jürgen Habermas has observed, “[t]he legitimating reasons available from the constitution are given to the Constitutional Court in advance from the perspective of the application of law – and not from the perspective of a legislation that elaborates and develops the system of rights in the pursuit of policies.”30 While the court-based (“triadic”) legitimacy seems hardly applicable to abstract judicial review, one can think of some different types of democratic legitimacy that might support the authority of courts to invalidate statutes. If we cannot derive the constitutional court’s legitimacy from the idea of an “impartial umpire” (because an abstract consideration of laws and policies does not lend itself to such a conceptualisation), then we should look for a more representative type of legitimacy, derived not from impartiality but from the democratic pedigree of the judges. It is
IN SEARCH OF LEGITIMACY
37
not unthinkable, and certainly not patently absurd, that a sort of “third chamber” (or a second chamber, in unicameral parliamentary systems) endowed with the task of re-examining a bill, this time from the narrower perspective of its compliance or otherwise with constitutional values, could be justified in terms of general principles of democratic legitimacy. A combination of long tenure, immunisation from direct societal pressures and from temptations connected with seeking re-election on one hand, and a degree of electoral pedigree (after all, judges of constitutional courts are almost always appointed by democratically accountable bodies) on the other, may provide just the right combination of a good democratic mandate with the institutional incentives necessary for the serious, principle-based review required of a “negative legislator”. If what worries us (as it should) is the non-existence of a democratic mandate of the negative legislator, then this concern may be (partly, at least) assuaged by the fact that appointments to constitutional courts are much more democratically based than those of ordinary judges; they are either made solely by the parliament31 (in which case the link between the judges and the democratic decision of the voters is reasonably direct), or at least with the participation of parliament in the recruitment process.32 In addition, political sympathies and/or legal and constitutional views of the judges are known (or at least, are knowable) prior to the selection to the court; and the system of limited tenure makes them relatively sensitive to the views of the general population – more so, in any case, than where judges have life tenure, as in the United States. For these reasons, one can claim that a constitutional court is an indirectly elected democratic (or near-democratic) “chamber of reflection”, the purpose of which is to reconsider the bill in a more dispassionate manner, removed one step further from specific political controversies. (Incidentally, such an approach permits us to look differently, and more leniently, upon the “politicisation” of the system of the appointment of judges to constitutional courts: the “politicisation” which is often depicted as an aberration of the system, turns out to be a desirable feature that endows the third chamber with legitimacy based upon its indirectly representative character; the members of this chamber are thus meant to represent the range of views within the community as to the meaning of broad constitutional provisions).33 This immunisation from the passions of the moment need not necessarily deprive the constitutional courts of their representative character; one may, for example, charge the court with the task of identifying (and giving effect to) whatever consensus can be found on a given issue (which has a bearing on constitutional interpretation) in the light of (rather than in isolation from) the actual, prevailing moral and political views in the community. One can even appeal to the Rawlsian idea of “overlapping consensus” as the proper device upon which a constitutional court should base its representative function.34 It is not my claim that such an argument is compelling. As a matter of fact, I do not believe that it is. To illustrate why not, consider this typical statement from a proponent of the idea of the representative functions of the United States Supreme Court: Without surrendering its prerogatives of judgement or compromising its obligation to uphold constitutional values in the face of political opposition, the Court, in specifying the meaning of constitutional principles, must be accountable at least in part to
38
CHAPTER 2 manifestations of reasonable moral and political commitments displayed by the citizenry, both nationally and locally.35
For one thing, there is an apparent possibility of tension between the obligations proclaimed in the first and in the second parts of the sentence. What if “citizens’ commitments” clash with “constitutional values” as understood by the justices of the Court? Secondly, the proviso that the only commitments that the Court must respect are the “reasonable” ones opens the gate to a number of “filtering devices”, which will transform the actual conventional morality into something hardly recognisable by the citizenry as its own moral commitments.36 Finally, the idea that the Court must be accountable to “commitments” rather than to the citizens themselves, strikes me as fanciful. Accountability presupposes the possibility that the principal may censure the agent: how can “commitments” do this? And yet the choice of words is not incidental, because, naturally, there is no way in which the justices of the US Supreme Court (or of any other court, for that matter) can be “accountable” to the citizens in the ordinary sense of the word. Furthermore, an “overlapping consensus”-based rationale would generate a number of more practical questions: if we need a “negative legislator” whose task would be to test bills from the point of view of constitutional values, should it be composed in exactly the same way as the actually-existing constitutional courts? Why should its composition be limited to lawyers only – given that, after all, legal skills are not decisive (nor are they the only relevant skills) in approaching the question of how best to articulate the specific meaning of broad, value-based constitutional pronouncements? As Burt Neuborne correctly observed, “When substantive-review judges identify values and totally insulate them from majority will, the troublesome question of why judges are better than other officials in identifying and weighing fundamental values cannot be avoided”.37 (This had been recognised in the design of at least one constitutional court outside CEE, namely the French Conseil constitutionnel; the members of this tribunal do not have to have, and some of them do not have, any formal legal qualifications). These are important questions but they will not be pursued here. The only point being made here is that to construe constitutional courts as belonging to an institutional branch of law making is not incoherent, and does not seem to raise insurmountable problems regarding their democratic legitimacy. Certainly the prospect of finding legitimating arguments for abstract review in terms of traditional representative democracy seems to be more promising than in terms of judicial function. The paradox is that the constitutional courts themselves, and their most fervent academic supporters, usually strenuously resist the characterisation of their position in the political system as a second or third legislative chamber, and construct their own self-perception as “courts”, albeit somewhat differently than the “ordinary” courts. There has been a discussion among constitutional lawyers of CEE as to whether constitutional courts should be classified as belonging to the judicial branch or as sui generis bodies. Furthermore, the actual location of the provisions pertaining to the constitutional courts in the structure of the respective constitutions varies somewhat from country to country. For example, in Slovakia the Constitutional Court is regulated in the part of the Constitution devoted to “[t]he judicial power”38
IN SEARCH OF LEGITIMACY
39
and is characterised inter alia as “an independent judicial authority”.39 Similarly, the constitutional courts in Russia and in the Czech Republic are regulated in separate chapters, while in Poland the Constitutional Tribunal is regulated in the chapter generally entitled “Courts and Tribunals”, but within its own subchapter. (This being said, the Polish statute on the Constitutional Tribunal explicitly states, in its first article, that the Tribunal is a judicial body). By contrast, several other constitutions include provisions on constitutional courts in separate chapters or parts altogether, without including them in any broader subdivisions. For example, in the Croatian Constitution the chapter on the Constitutional Court comes between the chapters on “judicial power” and local administration, in Lithuania between the chapters on the Government and the Courts, and in Hungary between the chapters on the President and the Ombudsman. The approaches in these countries range between pigeonholing constitutional courts in the “judicial” branch (which seems to be the dominant practice)40 and characterising them as sui generis institutions – which is arguably simply an avoidance of characterisation. For example, Professor Janusz TrzciĔski, in a chapter on the Constitutional Tribunal in the fundamental treatise on the Polish Constitution (written when he was himself a presiding judge on the Constitutional Tribunal), concluded that “the functioning of the CT [Constitutional Tribunal], as determined by the Constitution and by the Law on the CT, does not fit the accepted classifications [of branches of government into legislative, executive and judicial].”41 To my knowledge, there have not been any strongly expressed views, within the mainstream constitutional doctrine in the region (and certainly not by any of the constitutional courts concerned) that constitutional courts, when exercising abstract judicial review, belong to the legislative branch of the state. The self-perception of those courts as part of the judiciary, broadly speaking, has been also endorsed by some friendly commentators from outside the region. Owen Fiss has stated that, “In the new democracies of the East . . . the judiciary . . . must give life and force to the idea of a constitutional court. Judges on these courts must convince their fellow citizens that law is distinct from politics, and that they are entitled to decide what the law is”.42 The characterisation of constitutional courts qua courts is implicit in Ruti Teitel’s view that the power held by the Polish Parliament to override the decisions of the Constitutional Tribunal (before the adoption of the 1997 Constitution) was an example of the conflation of judicial and legislative powers, and evidence that “the understanding of separation of powers is far from entrenched in the region”.43 The image is of a legislative body (the Parliament) intruding upon the functions of a judicial body (the Tribunal). It may seem ironic that the conception that would offer perhaps the most promising path of legitimating constitutional courts in their exercise of abstract constitutional review is most decisively resisted by the constitutional courts themselves, while the doctrine that is patently unsuited to provide such legitimacy is the one most zealously defended by those courts and their apologists. But the paradox is of course, illusory. If one adopts a “third chamber” perspective on the exercise of abstract constitutional review, there is no justification whatsoever to stick to the current composition of the courts consisting, as they do, of lawyers only. Decisions about the death penalty, abortion, defamation of public officials, etc. may be dressed
40
CHAPTER 2
up in legal garb but they ultimately hinge upon fundamental value-choices, and legal qualifications have no bearing whatsoever on how these choices are made. The fact that it is the constitution rather than a non-textual moral or political theory which forms the direct basis for the scrutiny of a given law is no good reason to restrict the range of scrutinisers to lawyers. After all, what constitutional review in such cases is about is not the detection of the “true” legal meaning of such constitutional concepts as the right to life, privacy or freedom of speech but rather a decision about what cluster of values is preferable to others in the articulation of a vague constitutional formula with reference to a specific problem, which is underspecified, and thus left indeterminate, by a constitutional text. It is precisely because the issue is a choice of a cluster of values rather than an exegesis of the legal concept that the scrutinisers must be called upon to play a representative, not merely deductive, role. It is, however, for this very reason that no necessary connection exists between the legal qualifications of scrutinisers and the nature of the scrutiny; that is why the democratic legitimacy of constitutional courts, as they are currently constituted, is continually called into question. Moreover, it is insufficient to attempt to legitimise the existing constitutional courts by pointing out the less-than-perfect legitimacy of parliaments. “The conventional concern of the absence of democratic accountability posed by judicial lawmaking seems less apt in periods of political flux. In such periods, the transitional legislature frequently is not freely elected and, further, lacks the experience and legitimacy of the legislature operating in ordinary times”.44 Ruti Teitel makes this observation as a response to the charge of the lack of legitimacy of constitutional courts, with particular reference to the post-Communist transition. However, the observation about the legislatures not being freely elected applied to some of the legislatures in the region only (for example, to the post-Round Table election in Poland in June 1989), and even this was usually limited to the first term of legislatures after the transition. This observation has, therefore, now only a historical value. In those countries where the freedom and fairness of the election of legislatures is questionable (Belarus), the problem of “activist” constitutional courts does not arise in the first place. In fact, the most activist constitutional courts operate alongside the fully mature, freely elected legislatures. The “experience” of these parliaments may be called into question (as may be the experience of the new constitutional courts); but the remark about their “lack of legitimacy” is question begging. Anyway, even if the legitimacy of the parliaments is less than perfect, surely the remedy is not to transfer part of their power to bodies that have even less legitimacy to create law and determine policies. 3. WHY THE “CONTINENTAL” MODEL OF REVIEW: REASONS OR RATIONALISATIONS? To ask why the CEE countries have adopted, without exception, a “European” model of abstract judicial review, concentrated in a specialised constitutional-review body, may seem odd. After all, they are European countries, they do belong to a “continental” legal and constitutional tradition, and those same factors that
IN SEARCH OF LEGITIMACY
41
determined the victory of Kelsenian45 judicial review (as opposed to the US model) on the continent arguably must have informed the emergence of this very type of court in the Central and Eastern part of the continent, when the circumstances for democratic development finally became ripe. At the end of the day, this may be a correct answer, and yet I do not think that consideration of this question is superfluous. Asking the simplest and the most naïve questions can sometimes be illuminating; I believe that this is the case here. For one thing, not all Western European countries have adopted a system of judicial constitutional review at all, and of those Western European countries that have adopted a Kelsenian approach, at least one (Greece) comes close to a dispersed, USstyle model.46 As Allan-Randolph Brewer-Carias argued at length in his classic book on judicial review, there is no necessary connection between the way in which constitutional review is designed (that is, whether it is centralised or diffuse) and the family of legal systems to which a given nation belongs (that is, whether it is a civil or common law system).47 Secondly, even in some of the most emblematic systems of abstract and centralised review, such as Spain, there had been proposals made to establish a decentralised, American-style model, in which all courts would be authorised to review the compatibility of statutes with constitutional rights claims.48 As a matter of fact, the constant tension between the constitutional courts and the supreme courts in a number of CEE countries, based on the aspiration for the quasiterritorial monopoly of constitutional adjudication on the part of constitutional courts versus a demand by the supreme courts (and often other “regular” courts) to be able to set aside the rules they deem unconstitutional and thus make sense of the idea of the direct applicability of the Constitution, bears witness to the fact that the idea of a dispersed, decentralised and concrete review is very much alive, and not just the fantastical view of an academic commentator.49 Thirdly, we should be wary of explanatory determinism; after all, the emergence of the Kelsenian model in CEE may be under-determined by the factors usually referred to in this context. If this is the case (as I indeed believe with regard to a number of explanations discussed below), then the emergence of such a model may be seen to be historically contingent, and a belief in the plausibility of an alternative scenario (under which the American-style model would have been chosen) may not be as absurd as it seems at first blush. Furthermore, if that is the case, the usual explanations for the emergence of the current system may be better characterised as justifications for the maintenance of that particular system. They can therefore be seen more as legitimating the status quo than analysing it dispassionately. After all, the post-1989 constitutional and political scene in CEE was, partly at least, something of an experimental laboratory, in which many of the decisionmakers may have thought that they were making a “fresh start”. Of course, no start is ever fresh; nonetheless, the post-1989 period was a mixture of the embeddedness in the old traditions and experimentation with the new. There were many options on the menu, and an American-style solution to many issues of constitutional design was not out of the question. For example, in the works on the process of constitution-making in Poland in the second half of the 1990s, some of the leading experts advocated the adoption of the US-style review in which all courts would have the power to set aside the laws which they deem unconstitutional – and this
42
CHAPTER 2
plea was made on the grounds that such a dispersion of this power would be more conducive to the entrenchment of the constitutional culture within the judiciary in general. As one of those experts later reported, even the first informal remarks aimed at placing that proposal on the agenda “brought about a hostile response from the Constitutional Tribunal”50 which in Poland had existed since 1985 and which was not prepared to share its power with the “ordinary” courts. In addition, there was no shortage of American experts around, including constitutional experts, to provide advice and advocate the right solution, and it just so happened that the solutions proffered by these American constitutional experts, more often than not, corresponded to the liberal (in the American sense of the word) reading of US constitutionalism – which included an activist, US-style judicial review. Some of those American liberals explicitly urged the new activist constitutional courts (in particular, the Hungarian Court) to abandon abstract review altogether and, hence, to follow the US path.51 If we reconsider the question of why the CEE countries adopted the centralised and abstract model of review, and if as a result of this reconsideration we conclude that the usual explanations fail to fully account for the choice of the model (hence, they “under-determine” the reasons for adopting the model), then we can gain two things from such an exercise. First, we can help to re-open the debate surrounding the relative merits of the US style review and its future prospects in the region. (This, of course, is relevant only if we find that the decentralised model has some advantages over the centralised and abstract model, something that I will discuss in Chapter 3). Second, we can debunk the usual explanations by showing that, to some degree (that is, to the degree that there is under-determination), they are ex-post facto rationalisations, and thus must be seen as legitimating ideologies rather than dispassionate accounts. Let us dispense first with what are arguably the two weakest explanations of the phenomenon of constitutional courts in CEE. The first one appeals to the willingness of the countries of the region to match the expected criteria which would facilitate their admission to the European Union, and those criteria are said to include a “European” rather than a US style of judicial review. It has been also said that the EU, generally, expected the candidate countries to set up a system of constitutional courts that would be in a very strong position vis-à-vis the legislatures: [w]hile parliaments and presidents will predictably resist judicial interventions, they are painfully aware that highly visible confrontations with their domestic constitutional courts will gravely threaten prospects for early entry into the European Union, which is already looking for excuses to defer the heavy economic costs that admission of the East entails.52
This is sheer speculation, and improbable at that. I know of no evidence to suggest that the accession to the EU figured on constitution-makers’ minds when deciding on which system of constitutional review should be adopted in CEE, and I do not know why it should. After all, the preparations for accession to the EU, even in the cases of those countries long considered to be the most obvious candidates, began well after the establishment of the constitutional courts. Moreover, I know of no evidence that shows that the EU made it a part of its set of criteria for candidate states that they establish a system of constitutional review that granted a strong
IN SEARCH OF LEGITIMACY
43
position to the courts vis-à-vis the legislature. In the first important decision of the EU, which can be seen as setting the conditions of membership for post-Communist European states, the European Council established in Copenhagen in December 1993 that the candidate countries, in order to be successful, must display (among other things) “stability of institutions guaranteeing democracy, the rule of law, human rights and respect for and protection of minorities”; however, no specific institutional forms of attaining these conditions were ever established. Apart from everything else, it would be hypocritical for the EU to expect, let alone to demand, that constitutional courts be established: there are members of the EU whose democratic credentials are unimpeachable, yet who have no French- or German-style constitutional review.53 To be sure, in the process of monitoring the progress of CEE candidate states before their accession to the EU, the European Commission sometimes noted, with approval, the existence and the actions of the constitutional courts but nothing in its reports suggests that anything of importance hinged on it, and normally it was nothing more than part of the recital of a number of institutions in the sections of those reports on “Democracy and the Rule of Law”. None of these reports imply that the existence and the actions of the Constitutional Court was a litmus test for the strength of democracy and the rule of law in the countries under scrutiny. The second explanation which seems to me also quite weak is that there is a correlation between the fact that a country has just emerged from a period of authoritarian rule and the fact that it has established a “Kelsenian”, rather than the US, model of constitutional review. One can see a certain logic in the question by Louis Favoreu: How could an American system function in the Federal Republic of Germany, Italy, Spain, or Portugal, with judges from the preceding period of dictatorship named to the courts? Adopting judicial review in these countries would require “purification” on a massive scale of a corps of magistrates, while one could immediately find a dozen or so constitutional judges with no prior culpability during those periods, capable of carrying out their duties without mental reservations.54
The argument about a generalised distrust of the judiciary when a state emerges from a period of authoritarian rule is then extrapolated to the CEE post-Communist countries;55 and yet, the reality of post-Communist regimes defies the simple dichotomy noted by Favoreu. Neither were the judges of constitutional courts in the region quite “purified” of their old habits and ideologies, nor were the ordinary judges as hopelessly immersed mentally in the “preceding period of dictatorship” as to offer no likelihood that they would dispense justice in accordance with the new axiology of the law. Perhaps the clearest case of a “purified” supreme court is provided by Poland where all of the judges of the top judicial body were appointed anew after the transition of 1989. It follows that their “moral mandate” to interpret the Constitution in accordance with a democratic system of values is as good as that of the Constitutional Court. And this, I should add, is not merely an interpretation by an external observer but part of the actual self-perception of the judges themselves; at least some of those sitting on the Polish Supreme Court resent the implication that their authority and competence in articulating and applying the democratic constitution is inferior to that of their colleagues from the Constitutional Court.56
44
CHAPTER 2
One must not protest too much. Ruti Teitel certainly has a point when she observes that “as new forums specially created in the transformation, [the] very establishment [of the new constitutional courts] defines a break from past political arrangements”.57 Indeed, a “concrete” system of review would most probably have to rely on the old judiciary and so the symbolic effect of novelty would be lost. The explanatory power of this observation is, however, limited. Even leaving aside the counter-examples of Poland and ex-Yugoslavia (where the establishment of constitutional courts did not coincide with the transformation), some “old” institutions (such as the Presidency in the Czech Republic or Poland) quickly acquired much more symbolic power as vehicles of transformative politics than the “new” constitutional courts. There is no doubt that Vaclav Havel or Lech WaáĊsa were more potent emblems of the new, even though they occupied “old” offices, than the largely nameless and faceless judges of the constitutional courts in Warsaw and in Prague. Perhaps the most significant explanation can be found in the attachment of lawyers and constitution-makers in the region to the traditional “European” tradition of separation of powers in which the role of ordinary judges is strictly confined to the application, as opposed to the making, of the law. The adoption of the Kelsenian system seemed to disturb this tripartite structure of government to a lesser extent than allowing all of the regular courts to examine laws in terms of their constitutionality in the course of ordinary adjudication. The point made about the Western European systems, namely that the Kelsenian model “could be easily attached to the parliamentary based architecture of the state”,58 applies to the CEE countries as well. Certainly this has been a frequent argument within the doctrine of these countries: that to authorise regular judges to declare the laws unconstitutional would place them above the legislature, and would thus be inconsistent with the tripartite division of powers.59 These are plausible explanations, as far as the compatibility of any form of judicial review of constitutionality with “old constitutionalism”60 is concerned. However, much the same arguments that are being produced against the US-style judicial review, in terms of the traditional tripartite separation of powers, can be used to attack abstract and centralised judicial review, as long as it remains a judicial rather than a legislative function. These criticisms apply yet more forcefully to concrete judicial review by constitutional courts when they exercise it alongside their power of abstract review. If a single ordinary court can initiate a review of the constitutionality of a statutory provision by (what is seen to be) a court, albeit a special type of court, what then is left of the traditional European separation of powers, and the related dogmas of the sovereignty of parliaments and the limitation of to applying, as opposed to making, the law? Perhaps a more relevant point is the formal absence of a doctrine of “stare decisis” in the continental legal tradition. In the “decentralised” model of judicial review, such as in the US, a strong precedent doctrine provides for a degree of consistency within the overall judicial system. When all the courts have to follow the rationes decidendi of the Supreme Court and of the relevant higher appellate courts in their respective jurisdictions, the dangers of arbitrariness, uncertainty and lack of uniformity are minimised. However, where there is no stare decisis (so the
IN SEARCH OF LEGITIMACY
45
argument goes), a concrete-decentralised model threatens the unity of a legal system, and one can envisage an unwieldy situation in which some courts could find a particular law unconstitutional while others might uphold it. Indeed, this worrying prospect of lack of uniformity was the main argument given to me by a leading Russian constitutionalist in explaining the inapplicability of decentralised review to Russian conditions,61 and in this he was certainly echoing a dominant opinion in the constitutional doctrine in CEE. This distinction, however, is one of a degree rather than of kind, and it cannot make all that much difference to the debate. The decentralised system yields a degree of uncertainty and inconsistency, regardless of the stare decisis doctrine. In the United States, unless and until the Supreme Court has pronounced on a given issue (which, under a certiorari system and due to the control by the Court of its own agenda, need not be the case on every contentious constitutional issue tackled by lower appellate courts), there may exist a situation in which the Courts of Appeals for different circuits will come up with different solutions to one and the same constitutional controversy.62 On the other hand, it is simply not the case that a system of judicial precedent does not in fact operate in the legal orders of continental Europe. In that system, consistent decisions of the courts – especially of the highest courts – are in practice treated as unquestionable sources of law. This is so even if the official doctrine explicitly rejects the idea of precedent as a binding source of law. Consider this exposition of the French approach by two leading French constitutional theorists: The courts very rarely cite precedents and must not base their decisions on them, because the only legitimate source of law consists of statutes. On the contrary, if one looks at the material that is in fact used, one realizes that precedents are the most important. . . . Precedents, without being formally binding, may have force if created by a court superior to that where the case is pending. This simply reflects the hierarchical structure of the courts.63
Similarly, in the CEE countries it has long been accepted that, for instance, judgements of the Supreme Courts have the character of binding precedent for all other courts, at least to the degree to which the written laws do not provide a determinate solution to a particular controversy.64 The upshot is that neither the general “architecture” of the system of separation of powers, nor the significance (or otherwise) of precedent, provide sufficiently strong reasons for opting for a Kelsenian as opposed to a US-style model of constitutional review; indeed, they may be viewed more as excuses than as convincing justifications. In this, the establishment of abstract/centralised review after the fall of Communism resembles the establishment of abstract/centralised review in Western Europe where, as Alec Stone notes, “a majority of political élites remained hostile to sharing policy-making authorities with judiciaries”, and where the opponents of decentralised concrete review saw in such a scenario “the spectre of the dreaded ‘government of judges’”.65 It may well be that the same fears also weighed on the minds of the constitutional decision-makers in the CEE countries when they refused to consider the decentralised, US-style of constitutional review. But let us note a strange inconsistency between such an explanation and another conventional reason given against the adoption of US-style judicial review in
46
CHAPTER 2
Europe, namely, the low status, prestige and skills of continental judges as compared to the US. If indeed (as is largely the case) “the judiciaries of these new nations [sic] have very little institutional capital”,66 then the fear that these judiciaries will attempt to augment their power, amounting to a “government by judges”, seems ill founded. Perhaps the example of Japan, which operates a concrete/decentralised model within the context of a relatively low-status judiciary67 shows that the fear of “government by judges”, if the decentralised model of constitutional review is instituted, is groundless. Upon reflection, the reasons for this are clear: decentralised review carries with itself a whole set of doctrines of judicial restraint that are simply inapplicable to abstract constitutional review.68 But (it may be claimed) the reason for the preference for abstract and centralised review, and for hostility towards the concrete and decentralised one, is deeper than that. The United States – an emblematic case of a decentralised/concrete review, has a tradition of a free-market, anti-statist approach to law, officials and the state. In contrast, the CEE countries share with their Western continental counterparts a tradition of statist and centralised approach to the state in general, not only to the judiciary. The stronger the role of the state in the society and the economy, the more tendency there is towards state-controlled review of constitutionality. Such an argument has been recently made by John C. Reitz, who describes a close correlation between the forms of review adopted and the general approach to the role of the state.69 In the US, where a market-centred approach prevails, only concrete review is available, with some residual aspects of abstract review (that cannot, however, be set in motion by political actors). At the other extreme of the spectrum, in the most statist-centrist tradition (France), only abstract review initiated by political actors is allowed. In the mixed systems (e.g. Germany), we find a combination of abstract review, concrete review and constitutional complaint. There seems to be an undeniable logic in the asserted fit between abstract review and statism because various forms of concrete review (which normally have to be initiated by subjects not directly controlled by politicians, such as ombudsmen, the judiciary, or, in the case of constitutional complaint, individuals) imply a partial loss of control by the state over the initiation of constitutional review. The correlation seems to be supported by the other European cases not considered by Reitz, namely Italy and Spain that can be located, on the spectrum ranging from statism to marketcenteredness, half-way between the US and France, and where the judiciary (in the case of Italy) and the judiciary, Ombudsman or the individuals concerned (in the case of Spain) can initiate the process of concrete review. But is there really such a correlation? Was the Italy of 1948 (when the rules of judicial review that are still in force today were adopted) so much less statist-oriented than the France of 1958 as to account for the difference between the presence and the absence of concrete review? And similarly, was the post-Franco Spain of 1978 infused with non-statist, corporatist elements to the same extent as Germany in the early 1950s, so that the presence of concrete review which can be initiated by individuals can be explained by a particular understanding of the role of the state? Perhaps. I am unable to pursue such an analysis in the framework of this chapter. However, three observations are in order at this point. First, one should be careful not to take the very availability of concrete review as a possible symptom of the less
IN SEARCH OF LEGITIMACY
47
statist approach to the role of the state (as public lawyers would probably tend to do), as the explanatory role of the state factor would be then nil. One and the same factor cannot at the same time be a result of and the evidence for a proposed causal factor. Second, if the argument is that the general approach to the role of the state can explain the nature of judicial review in Central and Eastern Europe (something that Reitz is not claiming, I should add) then we have a clear case of underdetermination here. After the transition, the question of the proper role of the state in society (and towards the economy in particular) has been and remains one of the most contentious unresolved issues in the region. Third, if “statism” yields abstract review because concrete review involves the loss of control by the state over the initiation of the review process, then this argument is unavailable to those systems in which abstract review is complemented by concrete review and/or constitutional complaint; thus, France would remain the sole instance of abstract review which can be explained by statism. However, a more interesting question is whether or not such a “fit” is present also at a deeper level, as Reitz suggests that there is a connection between the fundamental values underlying the model of review and those behind the model of the state. According to Reitz, the principal value that underpins abstract review is “legal certainty”. This is because an authoritative decision about the validity (constitutionality) of a new statute is taken even before (or soon after) the statute enters into force, and there is no period of uncertainty between the enactment and the review. In turn, such a period of uncertainty is necessarily produced by a form of review which is conditioned by a specific legal “case or controversy”. So much is probably non-contentious; legal certainty may be indeed higher in the system of abstract as opposed to concrete review. I say “may” because, as soon as the system allows concrete review alongside abstract review, as do all of the Western European and CEE systems of judicial review (with the exception of France and Ukraine, respectively), the benefit of legal certainty related to purely abstract review is lost. Indeed, the effect of legal certainty is assured only when a review is solely ex ante, so that once the law is ratified there is no possibility of ever declaring it unconstitutional (as is the case in France). An ex post abstract review, on the other hand, introduces an element of uncertainty, related not so much to the abstract nature of the challenge, but rather to the fact that review may be initiated (never mind by whom) after the law has entered into force. This kind of uncertainty can, however, easily be minimised by the simple technique of establishing a legal deadline until which a new law can be challenged.70 If no such techniques are actually being used71 it may be for the reason that the uncertainty that results from abstract ex post review has never been perceived as a major problem. But even conceding, for the sake of argument, that abstract and centralised review provides for a higher level of legal certainty than the concrete and diffuse one, does it indeed follow, as Reitz claims,72 that abstract review is based on a degree of paternalism while the US model of concrete review reflects the strong anti-paternalistic stance of the American constitutional system? In Reitz’s words: “The kind of citizen required by a system limited to concrete review is a ‘tough’ citizen, one who is willing to run significant risks deliberately in order to vindicate his rights, not one who waits for the paternalistic arms of the state to take care of
48
CHAPTER 2
him”.73 We may accept that the general hostility to paternalism is higher in the American political culture than in Europe. However, it remains to be seen whether this higher American anti-paternalism can indeed explain the exclusive reliance on concrete review and, a contrario, whether the relatively higher degree of acceptance of paternalism in Europe explains the European preference for abstract review. Taking the argument one step at a time, it may be true that paternalism (that is, a conviction that the government knows what is good for its citizens better than the citizens themselves do) is inconsistent with a high degree of legal uncertainty: a paternalist government would like to signal clearly to the citizens its expectations concerning their behaviour. However, the link between paternalism and high legal certainty (which needs, as we have seen, not only abstract but also ex ante or limited-in-time review) is contingent and indirect at best. After all, any government interested in guiding the behaviour of its citizens by clear rules, paternalist or otherwise, has an interest in providing a high degree of legal certainty to those subject to the laws. This legal certainty (and thus, the efficacy of authoritative rules) clashes at times with other values, such as flexibility or individual selfdetermination; it is not, however, clear why the “paternalistic” character of rules would add extra weight to the legal certainty side of the calculus. While the antipaternalist might applaud opening the path for individuals to challenge the laws through concrete review, it is question-begging to assert that she should fear keeping open the possibility of abstract review at the same time; and thus it is doubtful whether the “[r]ejection of paternalism surely lies at the heart of . . . the US rules on justiciability”.74 The only reason why an anti-paternalist might oppose abstract review would be the fear that the government (or any other official body endowed with the authority to initiate the review) could be tempted to exercise it in a paternalistic fashion, that is, on the basis of the alleged good of the citizens who might benefit from the success of the review, even despite the citizens’ own views to the contrary. For paternalism, strictly speaking, occurs only when the authority displaces the actual preferences of the citizens while claiming that it is doing so for their own good. However, such a depiction of the official motives behind the review strikes me as convoluted, if not fanciful, in most cases: in the emblematic examples of the exercise of review in “abstract-review-only” situations, that is, in the famous decisions of the French Conseil constitutionnel, one would search in vain for any cases that would fit with such an account. And no wonder: when minority representatives in the French legislature successfully challenged the bills on media pluralism75 or the nationalisation of enterprises76 they did not appeal to any paternalistic arguments but to their own political or ideological visions, different from those of the majority. This was a routine game of democratic politics, resolved by the Conseil in these cases in favour of the minority; appeals to paternalism did not (and did not need to) figure anywhere in the discourse. As a general speculation, it is hard to see why, as a rule, the initiators of abstract review would “become detached from the concerns of the individuals whose rights are immediately at stake”, to such an extent as to risk a situation in which the citizens would actually be opposed to the goals underlying such an intervention.77
IN SEARCH OF LEGITIMACY
49
But perhaps there is another type of link between paternalism and abstract review: a fear that the exercise of constitutional review by individuals concerned would be unwise, immature, detrimental to themselves. Such fear would certainly have strong paternalist undertones, and one can understand why, in defending the projects of fundamental reform of constitutional review in France in the early 1990s, the then President of the Conseil constitutionnel Robert Badinter warned: “on ne peut traiter indéfiniment les citoyens en etérnels mineurs”.78 The main point of the proposed reform (which failed to gain the support of the Senate) was to endow each party to a legal process with the right to challenge the constitutionality of a statute (on the basis of an alleged violation of the party’s fundamental rights), provided that the Conseil constitutionnel had not pronounced on the constitutionality of this law previously. The attitude attributed (no doubt with good reasons) by President Badinter to the opponents of the reform indeed smacks of paternalism. However, as soon as there exists in a given judicial system a path of concrete review (and, even more importantly, of constitutional complaint) that is available to individuals alongside the abstract review initiated by political actors, the link between abstract review and paternalism collapses altogether. Perhaps, therefore, the French system can be accounted for by the tradition of paternalism, but the German, Spanish, Italian etc. approaches cannot. It is not that there is less paternalism in a mixed system that combines abstract and concrete review; rather, that there is no link between paternalism and the constitutional model of review at all. Similarly, a possible link between the CEE model of review and a (putative) paternalistic tradition cannot be seriously upheld. Perhaps a better explanation would be that concrete review is well suited to a narrow understanding of the role of the constitution (and, consequently, of constitutional review), that is, when the main purpose of the constitution is seen as the safeguarding of individual and minority rights against majoritarian oppression. This is a characteristically US model of constitutionalism, and one that has resulted in the well-established perspective that understands constitutional review as the last bastion of individual (and minority) rights against legislative intrusion. It is plausible that someone who endorses this view of the constitution’s purpose and this perspective on constitutional review may have a clear preference for concrete, as opposed to abstract, constitutional review. This is due to the fact that, if the whole rationale for the power of review is based in distrust towards political institutions, then it would be odd to endow those very institutions with the task of initiating the review process. When individuals feel that their rights have been violated by the legislative majority, they can stand up for themselves and press their claims in the court, leading, hopefully, to a constitutional review – or so the argument goes. However, if we broaden our view on what constitutes the proper realm of constitutionalism, and in particular if we incorporate the setting of general socio-economic goals into the scope of the constitution (and, again, of constitutional review as a consequence), then individual litigation no longer seems an adequate mechanism to commence the process of review in all cases. Hence, the argument must derive the abstract mode of review from a broader understanding of the proper scope of the proposed functions of constitutionalism, beyond the protection of minorities against majoritarian oppression.79
50
CHAPTER 2
This proposition could be plausibly defended on a number of grounds but the most obvious would be that individual citizens do not ordinarily have (each taken individually) a sufficient stake in challenges to the laws that would compel them to launch a constitutional litigation, or (and this comes basically to the same thing) that their interest in winning such litigation is, at best, only indirect and remote. What is important here is that the constitutional court is now seen not only as a protector of individual and minority rights against the legislative majority, but, more fundamentally, as a guardian of the constitution as a whole, including its separation-ofpowers rules and its guidelines for socio-economic policy, when applicable. From such a perspective, it is only logical (within the logic of collective action) that the judicial review of legislative acts should be triggered by political actors in those cases in which one cannot normally rely on individual litigation as a mechanism of satisfying individual preferences that, although widespread, are not intensive enough to yield sufficient motivations to incur the costs of private suits. This argument seems plausible though one must not exaggerate the link between the abstract nature of the review and the policy-oriented nature of the review. After all, in those constitutional systems that rely exclusively on concrete review, such as the United States, Canada or Australia, much review has a policy-oriented aspect, not directly reducible to the protection of individual rights against the majority. It is significant that in a classic and deservedly famous article on the role of the Supreme Court, Robert Dahl characterised it as a “national policy-maker”, and showed that the dominant views on the Supreme Court have never been out of line for long with the policy views dominant among the lawmaking majorities of the time.80 Under the sufficiently relaxed rules of standing, not only individuals but also groups and associations can pick up various policy-based grievances and turn them into constitutional suits; for example, in the United States, they have standing to assert those interests of their members which are germane to the association’s purpose.81 An even broader test of standing was adopted in Canada where all that a plaintiff in a suit seeking a declaration that legislation is invalid needs to show is that “he has a genuine interest as a citizen in the validity of the legislation”.82 In India, which is another important concrete-review-only country, the courts have been long used for public-interest legislation, and standing to sue has been granted to any “member of the public having sufficient interest”, where “sufficient interest” encompasses a genuine concern for the rights of others.83 This shows that the concreteness of review does not seem to be such an impediment to policy-related complaints, after all. However, even if abstract review seems better suited to those exercises of review that are not directly related to a claim of a violation of an individual right, this would only provide a partial explanation for the dominance of abstract review in CEE constitutional systems, for two reasons. Firstly, even if one adopts a broader notion of constitutionalism that encompasses a control by the constitution of large areas of policy-making, it still does not explain why one would want to involve a constitutional court into this control. Unless one equates the scope of constitution with the scope of justiciability, and one believes that any constitutional violation should be reviewed by a constitutional court, this argument for abstract judicial review is question begging. In other words, while it is, of course, undeniable that abstract
IN SEARCH OF LEGITIMACY
51
review allows a role for the court in national policy-making to a much higher degree than concrete review does, it still does not follow that one should want to involve a court in the policy-making. Secondly, the prevailing arguments in favour of establishing constitutional courts in CEE were made precisely in terms of the protection of constitutional rights against legislative violations thereof, that is, in terms that can be safely located within the traditional, anti-majoritarian logic of concrete review. This, to be sure, was not the only type of argument put forward in the constitutional discourse of these countries, but it was a dominant one; for example, the leading courts in the region liked to emphasise that they saw their main role as the protector of individual rights. As the then Chief Justice Sólyom of the Hungarian Constitutional Court declared, “We always stress that we are activists in certain areas, namely, concerning fundamental rights, where the Court does not hesitate to decide ‘hard cases’. But we are self-restrictive concerning the problems related to the political structure”.84 A commentator on the Hungarian Court could therefore accurately observe: “The Hungarian Constitutional Court has defined its own activity as that of the guardian of human rights in the midst of a quasirevolutionary transformation…”.85 And this was by no means limited to the Hungarian court only. A judge of the Russian Constitutional Court stated that the goal of protecting and guaranteeing human rights was one of the three, equally important, tasks of the Russian Court, alongside overseeing the federal-regional relationships and the relationships between the highest bodies of the Russian state.86 The primacy of rights protection as a top justification for the robust position of the Constitutional Court is very much part of the prevailing self-perception of these bodies in the emerging democracies of CEE. It is for this reason that the general thesis put forward by Martin Shapiro cannot easily be applied to the phenomenon of the constitutional courts of CEE. Shapiro’s argument is that the power of constitutional courts today derives from the fact that they are useful as arbitrators in division-of-powers disputes, in the sense that they “keep the basic institutional processes running”,87 and consequently that the acceptance of rights adjudication by political actors is a necessary cost of having this instrument in place. As the above-quoted declaration of Justice Sólyom, among others, suggests, it is not the case that the principal declared aim of setting up the constitutional courts was to supply an umpire in division-of-powers disputes rather than to articulate rights. Neither is it the case that, in CEE, rights adjudication came later in time than the division-of-power adjudication, the trend that Shapiro discerns with regard to other constitutional courts in the world. On the other hand, the fact that these courts often tend to be more deferential to legislatures on separation-ofpowers issues than on rights issues88 seems to confirm a hypothesis that, in order to gain the political capital necessary to allow the court to be activist on rights, it must “shore up its party political and popular support”89 by deferring to politicians on other issues that may affect their vested interests much more directly, namely on their powers and the procedures available to them.90 This fact may explain why the courts may afford to be “adventurous” and “courageous” in confronting legislatures on rights-related issues while they are often meek and deferential on questions of the relationship between various organs of the state. This, however, is a contingent truth: it assumes that, in order to build political
52
CHAPTER 2
capital, constitutional courts will defer to non-judicial bodies on those matters which are less salient, or less important, to the legislature and/or the executive, and that these bodies care more about their place within the system of separation of powers than about the interpretation of constitutional rights. However, the latter factor (the separation of powers as more salient to non-judicial bodies than fundamental rights) is not necessarily the case. After all, when the court intervenes in cases concerning the separation of powers, it will usually support one agent against another (the president as against the prime minister; the parliament as against the president, etc), and the body that gained the support of the court in its rivalry with other institutions (or the political party that is behind this body or its majority) will remember, and be grateful. This is part of the political capital that the court will accrue. On the other hand, when the constitutional court takes on the legislature with regards to its understanding of rights, it may put itself in an antagonistic relationship towards the parliamentary majority and the government and the president (this is what actually happened when the Polish Constitutional Tribunal found the abortion statute unconstitutional in 1997). I am making this obvious observation in order to indicate that it is far from self-evident that the courts necessarily have a strong incentive (in terms of building up political capital, especially in the early years of their existence) to be deferential on the separation of powers and activist on fundamental rights; depending on how salient the issue is to the non-judicial bodies that will feel challenged by the court’s decision, it may well be (and often is) that the reverse is the case. But (the argument might go) it is not only the salience of the issue to the nonjudicial bodies that matters; it is also significant whether or not the court is able to identify the sphere within which the other bodies may accept a solution determined by the itself; in other words, whether the court is able to find a zone of consensus (even if only a tentative and grudging consensus). As Lee Epstein, Jack Knight and Olga Shvetsova have recently shown on the basis of an analysis of the experience of the Russian Constitutional Court, such a court is most likely to hand down a decision when it can find an overlapping “tolerance interval” around the most-preferred positions of the non-judicial bodies, and of itself.91 The decisions which are broadly acceptable to all main institutional players (even if not congruent with their most preferred positions) will tend to strengthen the Court’s legitimacy. In contrast, if there is no overlap on a given issue in the attitudes (not just in the most-preferred positions but also in the “tolerance intervals”) of the main institutional players (such as the President, the Lower Chamber, the Upper Chamber, and the Court), the Court would do best by avoiding taking any decision. If it does decide nevertheless, it inevitably weakens its institutional position. Epstein et al. show how the disregard for this truth led the first Russian Constitutional Court (1991-1993) into a disaster: it tried to take a stand on issues on which there was no overlap in the tolerance intervals between President Yeltsin and the Supreme Soviet of the time, namely on the range of presidential powers and the scope of decentralisation of the system, and consequently it fell victim to Yeltsin’s anger. In turn, its case law on individual rights, on which (according to Epstein et al.) there were important overlaps in tolerance intervals, was quantitatively much less significant. In contrast, in the second stage of its existence (after 1995) the
IN SEARCH OF LEGITIMACY
53
Court prudently (by Epstein’s standards) focused on individual rights and easily targeted the positions that were within the boundaries of the overlap, while avoiding the tricky issue of separation of powers on which no overlap between the positions of the President and the Duma existed. (It also managed to identify an overlap range on issues related to federalism). Hence, the numerical explosion of individual-rights decisions in the second stage of the Court’s existence, compared to the first stage. Epstein’s research is useful for our purposes: by identifying the overlaps in “tolerance intervals” as the source of the legitimacy of courts, it shows that they are dependent for their strength upon the relationships and interactions with and between other main institutional actors. What Epstein’s conclusions do not support (and she does not suggest that they do) is the argument that, in normal circumstances, the sphere of human rights is the one in which it is easier for the court to establish its political capital while the sphere of separation of powers (both vertical and horizontal) is more risky, and the courts should tread more gently there. Which area is more conducive to the build-up of a court’s legitimacy is a fully contextdependent matter, and will vary from one country to another. After all, when the parliament and the court are deeply opposed to each other on the question of a statute or policy that is important to the government and yet that the court strikes down on as violative of constitutional rights, as was the case in Hungary with the invalidation of the economic austerity program in 1995, then we have precisely the sort of confrontation that may lead to a weakening of the court’s political capital. Whether it will happen depends (partly) on the views, strength and prospects of future victory of the political opposition at the time of the constitutional challenge. Epstein et al. consider each chamber of parliament as a separate entity, but of course, in the process of abstract review, an alliance between the Court and (in the case of an invalidation) the parliamentary minority is very likely. In consequence, what may appear as a head-on confrontation between the court and the parliament may in fact turn out to be a matter of an alliance between a majority of the Court and the parliamentary minority – but a minority that may become one day the government-building majority. In such circumstances, the political capital of the Court may be enhanced rather than weakened, even though at the time it may seem that the Court is operating in a sphere in which no overlap in tolerance intervals can be discerned. Furthermore, it should be added that the courts may weaken their political capital not only by confronting the legislature and striking down its statutes, but also by upholding certain laws, when the act of upholding is viewed as overtly politicised and especially when the law is amended by the legislature soon after an election. This may well happen in the human rights area; not necessarily only in cases involving questions of the separation of powers. Consider, as an example, a decision of the Bulgarian Constitutional Court in 1993 that upheld a controversial, and widely criticised, statute on the “lustration” of academic staff at universities and other academic institutions.92 One of the justices of the Court later commented, with the benefit of hindsight, that the law (which had been revoked by a newly elected majority not long after the decision of the Court) had been upheld by judges voting “in accordance with their political, and not legal, views”.93 Regardless of whether or not one shares this opinion, it is clear that once the Court entered into the fray (and it
54
CHAPTER 2
had no opportunity of avoiding a decision) there was no way of identifying a “safe” (i.e., located within the overlap of the tolerance intervals) way of deciding the issue. This being said, ultimately the question to be posed is not whether it is more efficient (from the point of view of a build up of its own political capital) for a constitutional court to be non-deferential on questions of separation of powers or individual rights, but rather which (if either) is the more proper field of an active intervention, considering the skills, qualifications and institutional position of the court. Off-hand, one may be excused for thinking that rights-based review is inherently more problematic because rights provisions lend themselves, more than other constitutional provisions, to different, often contrasting, interpretations. Consider some of the vexed questions and controversies regarding the interpretation of vague, general rights-provisions: Does a general equality provision (such as a general right not to be discriminated against) mandate, permit, or prohibit affirmative action? Does a general right to freedom of speech protect individuals who wish to spread hate-speech and racist propaganda? On such matters many reasonable people (who affirm a given right in abstracto) can, and do, fundamentally disagree, and that is why the authority of a particular body to make a final articulation of the meaning of a constitutional right necessarily raises the question of its legitimacy. On the other hand, structural and procedural provisions related to the separation of powers or to the integrity of legislative procedures are less ambiguous and controversial (though are not themselves unambiguous or self-evident).94 To give just two examples from CEE: when the Lithuanian Constitutional Court struck down certain provisions of the Law on the Courts that gave the minister of justice the right to nominate judges for presidential approval, because the Constitution (Art. 124) charges the judicial college with counselling the president on the appointment of judges,95 the possibility of reasonable disagreement surrounding the (un)constitutionality of the challenged provisions was rather slim. Similarly, when in late 1999 the Moldavian Constitutional Court found – at the instigation of the parliament – that the referendum called by President Lucinschi to transform Moldova into a presidential republic was unconstitutional, because the president had encroached on several parliamentary prerogatives, under Art. 88 of the Constitution, that stipulate that the parliament should have the final say on whether to call a referendum,96 again, the room for reasonable disagreement about the correctness of such finding is narrow. It is not, however, merely a matter of the scope for bona-fide disagreement with the Court’s finding, but also of how qualified the judges are to pass a judgement on a given issue, and how likely it is that this judgement will be affected by their personal, political or ideological biases. Typically, the members of constitutional courts are eminently qualified to interpret general provisions on powers and procedures because this is the stuff of legal education, and is precisely what they have been trained to do well. As Burt Neuborne said: “if the question is which organ should make a decision affecting a fundamental value, rather that what the decision should be, the functional benefits of using judges to resolve disputes would, I think, be widely conceded”. 97 These “functional benefits” stem partly from the fact that judges are well suited, by their qualifications and experience, to resolve these kinds of disputes. Thus, when a (relatively) unambiguous constitutional provision is
IN SEARCH OF LEGITIMACY
55
breached in the political or legislative process, it is a requirement of the rule of law that the constitutional court must invalidate the political decision in question.98 This argument, however, cannot be made with respect to ambiguous provisions which allow for multiple interpretations. This is not only because there is no special relevance attached to legal qualifications in making one more qualified to resolve the moral and political problems that arise from abstract provisions of rights, but also because there is a distinct risk that judges will bring their own ideological views and prejudices into the articulation of the “true” meaning of such provisions. In fact, they cannot help doing so; the question is, why it should be their ideological views and prejudices that will inform the binding articulation of vague and ambiguous constitutional rights? Therefore, the argument for stronger grounds for intervention by courts into legislative acts based on the separation of powers argument as opposed to the rights argument is made on two grounds: firstly, on the difference in the degree of ambiguity of these two types of provisions (separation of powers provisions being normally less ambiguous than rights provisions), and secondly, on the basis that the special skills and competence of the judges are better suited to dealing with procedural and separation of powers issues than with articulations of rights. These two grounds are conceptually independent of each other; the ambiguity argument is distinct from the institutional competence argument. It should be added that the ambiguity point is made about whole classes of provisions rather than about any single provision. One may, after all, contrast a very ambiguous procedural provision (for example, that no significant amendments to the proposed bill should be made by the Upper Chamber of the parliament)99 with a relatively unambiguous right (for instance, that a citizen has a right not to be extradited and a right to return to her country). But by and large, it seems obvious that rights lend themselves more readily to divergent reasonable articulations than procedural rules or rules concerning the separation of powers. This, it should be noted, is not a conclusive argument against rights-focused judicial review, but, at the very least, it demands of courts a much better justification (and, perhaps, a much greater degree of caution when they engage in the scrutiny of a law on the basis of rights, as opposed to scrutiny based on (alleged) defects in procedure or disrespect for the division of powers. The idea that the constitutional court should, whenever possible, aim to consider the challenges to a law in terms of the procedure (and consistency of its adoption with the rules concerning the separation of powers), rather than in terms of its substance, has been an important theme in the theory of judicial review in the United States – certainly not the only theory of judicial review but an influential one nonetheless. This normative theme should be distinguished from a “procedural” theory of judicial review, understood as the scrutiny of the substance of a law in terms of whether it promotes or negates the operation of democracy understood as a set of procedural devices of self-government.100 Rather, the argument is that it is more democratic, and also more in tune with the competencies of the Supreme Court, if it “examin[es] carefully the process of decision . . . even though the particulars of the decision itself [are] outside its institutional capacity”.101 The theory can be understood in a strong sense, as foreclosing the entry of the Court in the substantive-review area, and in a weak sense, as suggesting that whenever the option
56
CHAPTER 2
is available, the Court should choose to invalidate the offending law on procedural rather than on substantive-rights grounds. To be sure, the latter, weak, understanding carries with it a danger of hypocrisy: it may seem to suggest that, even if the Court wishes to strike down a provision for substantive reasons, it should try to manufacture the procedural arguments in order to render its decision more palatable and more consistent with its legitimate powers. There will be times when such a procedural argumentation will be clearly contrived and unconvincing. Either way, the theory reflects the view that the Court is much better positioned to evaluate the separation of powers and rules of procedure than it is to ascertain the meaning of substantive provisions, in particular those concerning rights and liberties. Against this theoretical background, it is striking that constitutional courts in CEE often see themselves as guardians of rights provisions more than of the separation of powers. The statement made by ex-Chief Justice Sólyom, quoted above, exemplifies this point.102 This has been noted, with approval, by a student of the Hungarian Constitutional Court, who observed that the Court “has been willing to act more deferentially in non-human rights cases in order to shore up its party political and popular support”.103 This discrepancy had been criticised by Andrew Arato, who observed that, while the Hungarian Constitutional Court has become an effective defender of liberal constitutionalism, it has not been equally committed to democratic constitutionalism. In Arato’s opinion, the court failed to take a sufficiently strong counter-majoritarian stance in those cases that called for intervention to reinforce democracy, to keep the democratic process open, e.g. in the area of governmental control of the media.104 It goes without saying that, when referring to “democracy reinforcement” and the goal of “keeping the democratic process open”, Arato applies John Hart Ely’s theory of judicial review that defends strong intervention by the United States Supreme Court in the political process, on the grounds of defending the democratic process against distortions (such as those effected by restrictions upon political communication), and remedying its procedural defects (such as the inability of some minorities to participate in the normal formation of coalitions aimed at constituting, one day, a legislative majority).105 Moreover, the Polish Constitutional Tribunal adopted a theory that different subject-matters of constitutional review may trigger different levels of deference to the legislature, and that in the case of “classical” human rights, “almost any statutory regulation calls for a careful scrutiny from the point of view of permissibility of its enactment and of its substance”.106 This is contrasted to those subject matters that “by their very nature are left to a broad political discretion of statutory regulations” such as “socio-economic issues regulated on the basis of a particular political ideal of social development”.107 The second part of the argument is unobjectionable: those regulations and policies that require, in terms of an appraisal of their desirability, a reconstruction and evaluation of the broad social vision upon which their strength and plausibility are based, seem to be largely beyond the competencies of a judge, and should be scrutinised only as an exception and with a strong presumption in favour of its constitutionality. One may draw an analogy here to Christopher Eisgruber’s distinction between “discrete” and “comprehensive” ideals:108 it is Eisgruber’s thesis that, with some exceptions, judges are less well equipped to evaluate those standards that rely upon comprehensive principles (that is, the
IN SEARCH OF LEGITIMACY
57
principles which demand an articulation of a general vision of a good system as a whole, for instance, of economic justice or electoral fairness) than those standards that constitute reasonably specific side-constraints upon governmental action. It is, of course, easier to state this distinction in abstracto than to apply it in practice, and the very characterisation of a particular rule as relying upon a “comprehensive” as opposed to a “discrete” principle will often be contested – but, as a rule of thumb for identifying the fields of judicial competence, it is not a bad starting point. This being said, however, the first part of the argument from the abovementioned decision of the Polish Constitutional Tribunal, concerning the natural competence of the court to apply strict scrutiny to regulations implicating “classical” human rights, is more controversial. Justice Lech Garlicki, who authored this decision, defended a strict scrutiny of rights-impacting statutory regulations by reference to “a constitutional assumption of leaving a maximum liberty to individuals”.109 This assumption is correct; but there is a non sequitur between it and the lowered deference to statutory choices by the constitutional court. The move from the assumption to the stricter scrutiny is understandable only if we assume that the court is, naturally and necessarily, more on the side of broader individual liberty than the parliament is. This, however, is something that has to be shown rather than presupposed: we cannot simply assume a priori that whenever the court and the parliament disagree about the liberty implications of a particular regulation then the court is, ex officio as it were, pro-liberty and the parliament necessarily a menace to individual freedom; rather, one has to acknowledge at least the possibility that it may be the other way round. The argument about “natural” competence of the Court to apply strict scrutiny to human rights-implicating regulations is, therefore, question begging. It should be added, however, that there have been instances in CEE when constitutional courts have preferred to invalidate a law on procedural rather than substantive grounds, even though these two alternative avenues for scrutiny were theoretically open to it. As an example, one may refer to the Polish Constitutional Tribunal decision in 1998 concerning the statute on the “lustration” of judges:110 an article that would have suspended the operation of statutes of limitations for disciplinary actions against politically-motivated judicial decisions in the Communist period was struck down on procedural grounds, with an almost open encouragement for the parliament to reconsider the matter in a procedurally correct way.111 The procedural defect was found in the fact that the National Council of the Judiciary (NCJ) – a corporate body representing the judges – was not properly consulted on the proposed bill even though the Constitution mandates such a consultation. Whether indeed the duty to hear the NCJ in the process of drafting the law was observed or not was in itself a controversial matter (the NCJ had presented its opinion on an earlier draft but not on the later one); but the CT chose to focus on this point in order to strike down a central provision of the law. However, it should be noted that this strategy of procedural scrutiny was subjected to strong criticism in the dissenting opinions, in which it was claimed that the procedural argument was convoluted and unpersuasive.112 One of the dissenting judges issued a warning regarding “the danger that, in proceedings on the conformity of normative instruments with the Constitution, the centre of gravity will be moved from
58
CHAPTER 2
questions concerning the substance to procedural issues, especially where . . . such solution may justify passiveness and lift [from the Court] the responsibility for the substance”.113 This is a powerful illustration of how contested and tentative the theory concerning the procedural, as opposed to substantive, strategy of scrutinising statutes by the constitutional court actually is. 4. CONSTITUTIONAL COURTS AS PROTECTORS OF MINORITIES? A favourite and often dominant line of defence of the strong role of constitutional courts – not only in CEE, but particularly strong there – is to perceive them as the defenders of constitutional rights against the policies decided on by the political branches of the state, and, in particular, of minority rights against possible majoritarian tyranny. The argument appeals to the non-majoritarian aspects of democracy: the parliament and executive are (at least, when they function well) the articulators of the majority will but these devices have to be complemented by restrictions on the majority, and this is (the argument goes) the essence of constitutionalism, and of constitutional bills of rights in particular. The argument is familiar, and may be exemplified by the account given by a judge of the Bulgarian Constitutional Court, Professor Neno Nenovsky, in an interview with this author.114 Constitutional judicial review, according to Professor Nenovsky, is not in competition with the “democratic principle in the modern sense”, because democracy requires more that simply respecting the will of the majority. Not only is the pedigree of the institutions relevant in judging their democratic character, but also whether or not the organs are continuously controlled: permanent control is also an element of democracy. The problem is that the institution assigned the task of controlling the other organs should also be subject to democratic control itself. The sovereignty of the people is expressed not only in the statutes but also in the constitution; it is the constitution that legitimates all the organs of power. This, for Nenovsky, is the main contribution of US constitutionalism which is different from the traditional European (mainly French) approach in that it places an emphasis on the constitution as the expression of popular sovereignty. The constitutional court, while not a representative body in the traditional sense, has connections (according to Nenovsky) with “the expression of the general will” within the range of functions that it performs. As the Bulgarian Constitution proclaims in Art. 1 (2), all power derives from the people, and this means that the people exercises its power through various organs, including the Constitutional Court. Further, the legitimacy of this Court is derived from human rights, which cannot be subject to majority rule. “At this level, the Constitutional Court is not dependent on the rule of the majority”; the Court should have the power to strike down a statute that is unconstitutional even if the statute expresses the interests of the majority. Professor Nenovsky concludes: “Were it not for the Constitutional Court, the tyranny of the majority would become a norm”; and one may safely claim that this represents a communis opinio of constitutional lawyers in CEE (and also of the sympathetic external observers)115 after the fall of Communism.
IN SEARCH OF LEGITIMACY
59
This conception of the role of the constitutional courts cannot, however, be taken at face value. First, it runs into some hard theoretical questions stemming from many unspoken, and not necessarily self-evident, assumptions. Second, it is confronted by a reality that does not necessarily support the thesis that constitutional courts – both in CEE and elsewhere – are on the side of minority protection when they collide with parliaments on the articulation of rights. At a theoretical level, the fundamental challenge to the conception of constitutional courts as defenders of (minority) rights is that vague and inevitably indeterminate constitutional rights provisions often lend themselves to divergent reasonable interpretations, and, when there is a collision of two institutions (or, rather, the majorities of the compositions of two institutions) in articulating a particular right, there is no reason to assume a priori a priori that one institution has necessarily a better insight into the “true” meaning of the right concerned. To use the words of Jürgen Habermas, “human rights are not pregiven moral truths to be discovered but rather are constructions”116 and, one should add, they remain “constructions” at the post-constitution-making stage in so far as their meaning remains open to differing articulations, thus calling for debate and authoritative resolution. This is a point already made earlier in this chapter, and needs to be reiterated only to emphasise that the usual rhetoric surrounding the idea that the court defends constitutional rights against majoritarian intrusions is just that: a rhetoric. It assumes that there is always a single, canonical understanding of a right, and, in addition, that the court is necessarily better placed to discern this canonical understanding. If, however, we reject as unfounded the assumption that there exists a single and objectively correct understanding of a right, we will see the court versus legislature clash on the meaning of a right for what it is: a disagreement between two institutions about how best to articulate a vague constitutional provision. When a court challenges a legislative understanding of a right, it adopts a quasi-legislative role in that it wishes to displace the parliamentary articulation of a right with its own. The institutional design of a particular system of law-making may grant the court such a role; but, even if the court has the “last word” on the binding articulation of a right, this does not imply that the parliamentary articulation was not about a right but about something else (a policy antithetical to a right), or that the parliament was proven “wrong” in its understanding of a right. To draw such a conclusion would be to infer judicial infallibility on the basis of contingent institutional design. Such institutional design (with the “last word” reserved to the court) may be recommended on the basis of a practical judgement according to which it is likely that, when the parliament and the court differ in their opinions as to the best articulation of rights, it is prudent to allow the court’s understanding to prevail. But note the contingent, tentative character of such an institutional decision. It certainly does not deny that the parliaments also articulate constitutional rights in their legislative choices; the view about judicial supremacy need not disqualify the members of parliament from a bona fide attempt to implement the constitutional mandates, as they understand them. Furthermore, the fact the members of parliament are dependent on the electorate for re-election does not negate the bona fide subjection of the parliament to the constitution because there is no reason to believe that the voters have a cavalier approach to constitutional rights, including those for
60
CHAPTER 2
members of minority groups. (I put to one side the observation, to which have already alluded in Chapter 1(3), that in systems in which judges of constitutional courts have no life tenure, those of them who will be up for re-election or for another job after having completed their term on the court will have to be popular with the majority political forces of their countries, upon which their future professional fate may depend). For one thing, some voters are members of minority groups and members of parliament have reasons to seek their votes also. More importantly, however, there is no reason to believe that, in the overall judgements made by the voters concerning their choices, ideas about justice, including those concerning the place for minorities in their societies, will not figure at all, or that it will always bow to selfish interests. As Jeremy Waldron has eloquently argued, it is simply unrealistic to believe that when the voters decide about whom to support, they always and necessarily do it upon the grounds of their interests, rather than on a combination of interests and views about justice: “People often vote on the basis of what they think is the general good of society. They are concerned about the deficit, or about abortion . . . in a way that reflects nothing more about their personal interests than that they have a stake in this country”.117 In consequence, legislative choices (insofar as they are seen as the extension of voters’ choices) must be seen as giving effect to a mix of views about majority interests and those relating to justice (the latter incorporating the views concerning the rights of members of minority groups). Of course, different voters (and consequently, different MPs) will have widely divergent conceptions of justice, but this is also what happens in the case of differences between the majority of the parliament and the majority of the constitutional court, when these two institutions clash over justice, or over constitutional rights. To infer on the basis of that clash that the parliamentary choice was really not an honest expression of about a conception of justice, but merely a policy decision based on the interests of the majority, is unwarranted. It is, therefore, no wonder that various scholars studying constitutional courts around the world have often concluded that the conception of a court as a defender of minority rights, or even of rights more generally, simply cannot stand the test of evidence. The American scholar Stephen Griffin, in his excellent work on the United States constitutionalism, refers to this problem by noting: “The emphasis on majoritarianism as the fundamental principle of American democracy in the debate over judicial review and the constitutionalist position rests on the assumption that only the Supreme Court can play a credible role in defending constitutional rights. This is clearly not the case”.118 Griffin then goes on to list various civil rights and liberties enacted by the United States Congress in the 1980s and says: “Many of these laws . . . were passed in response to numerous Court rulings that restricted the scope of laws designed to ensure the enforcement of civil rights”.119 Griffin concludes: “The [contemporary] debate [about judicial review] accepts the simplistic view that majorities are always interested in violating the rights of minorities. This makes it difficult to explain why Congress is able to produce consistent majorities in favour of civil rights and liberties legislation”.120 Nearly forty years before the publication of Griffin’s book, Robert Dahl had published an article which soon became famous, in which he conceptualised the US Supreme Court as “a national policy-maker”; Dahl rebutted the view that the Court
IN SEARCH OF LEGITIMACY
61
“stands in some special way as a protection of minorities against tyranny by majorities” – both as normatively suspect and as factually incorrect.121 It is perhaps significant that both Dahl and Griffin are, by their background, not lawyers but political scientists, which may make them more attentive to the reality of the Supreme Court’s work and more suspicious of the ideological rationalisations to which lawyers are prone. Alec Stone Sweet also approaches the study of courts from a perspective of political science and, interestingly, he too concludes (in his case, with reference to Western European constitutional courts) that, in many cases, “it is nonsense to suppose that the constitutional court functions as some kind of bulwark against the tyranny of majority rule”.122 Stone Sweet makes this observation in the context of a specific example he discusses, namely the legislative attempts in France in the 1980s to introduce affirmative action for women in local elections. The bill, brought before parliament by a Socialist government, was strongly opposed by the conservative opposition, which referred it to the Conseil constitutionnel ; the Conseil duly annulled the challenged provision on the basis of the principle of equal treatment under the law, as enshrined in the 1789 Declaration of the Rights of Man and the Citizen. Regardless of one’s views about the merits of this decision, it is certainly not possible for an observer to conclude that the Conseil behaved here as a protector of a minority123 against majoritarian tyranny. Has the performance of the constitutional courts in CEE supported the claim that they were, consistently, engaged in defending minorities against the majority? It is hardly possible to answer in the affirmative. A more detailed account will be presented in Chapter 8 but at this point one may only briefly observe that the evidence for such a proposition would be difficult to come by. Apart from the shining exception of the Bulgarian MRF case (in which the Constitutional Court defended a Turkish-based party against delegalisation),124 there have been virtually no significant decisions by constitutional courts in CEE along these lines. In fact, there have been surprisingly few decisions dealing with ethnic/national problems at all, even in the places when one would expect them. A natural place to look for such cases would be the Baltic states (in particular, Estonia and Latvia, with their large Russian-speaking minorities); but there, constitutional courts played a very minimal role in imposing a regime designed to accommodate the Russians. The case of Estonia is quite instructive in this regard. As the Estonian scholar Vello Pettai has shown, the Constitutional Review Chamber (CRC) was very timid in tackling the Language Act, and fundamentally avoided any principled appeal to minority rights in dealing with constitutionality of the provisions which were arguably discriminatory against the Russian minority.125 The CRC struck down the provisions of the law on technicalities; the parliament easily re-enacted the law free of technical defects, and it was only international pressure that subsequently compelled the parliament to amend the law. In other post-Communist countries, there have been very few ethnic-related decisions by constitutional courts; and those that there have been can hardly support the thesis that the central role of these courts is to shape a generous system of minority protection. For instance, in Romania in 1995 the Hungarian minority party UDMR, along with some other opposition parties, attempted to introduce into the draft law on education a provision on the right of the Hungarian minority to have a
62
CHAPTER 2
state Hungarian-language university. They did not succeed in the legislative process, and challenged the bill in the constitutional court, but the challenge failed. (Of course one may suggest that it is due to the general weakness of the Romanian Constitutional Court, and that a stronger court would take on the legislature more aggressively; this, however, is an unverifiable speculation). As another example, one could point at the decision of the Ukrainian Constitutional Court of December 1999, in which the Court strengthened the constitutional status of the Ukrainian language in Ukraine, and established an affirmative duty on all public bodies to use only Ukrainian throughout the country (even though in Eastern and Southern regions the Russian language is widely used both in private and public contexts).126 Of course, the ethnic/national dimension is not the only aspect of possible domination of a minority by the majority; the religious is another. Again, however, constitutional courts in CEE have not been, by and large, active fighters for religious tolerance based on the separation of state and religion. In the country in which the Churches (or one should say, the Church) are the strongest, Poland, the Constitutional Tribunal has been anything but a champion for religious tolerance, and time after time it has caved in to the pressure from the Catholic Church. Invalidating crucial provisions of the (relatively liberal) abortion law; upholding the introduction, by ministerial decree, of religious teaching in public schools; upholding the ban, in the Broadcast Law, on expressions offensive to Christian values127 – all these decisions were seen, rightly, as establishing a privileged position for the Roman Catholic faith, and amounting to discrimination against other religions or non-believers. This is not to say that the record of constitutional courts in CEE is negligible, as far as the protection of constitutional rights is concerned. I will attempt a more comprehensive account of this record in Chapters 6-10, and indeed it will be shown that, overall and subject to many reservations, the record in this field has been positive. But this is not the point discussed here; the object of this chapter has been to explore some of the main arguments aimed at justifying granting a strong power to constitutional courts as legitimate umpires of the constitutionality of legislation, when they exercise abstract review under the general constitutional provisions of rights. It is one thing to say that the operations of the constitutional courts had, overall, beneficial consequences, and another to argue about their legitimacy on the basis of fundamental precepts of democratic theory. The argument that, in a democratic system, there must be a protector of minority rights against majoritarian abuse, and that constitutional courts are well suited to perform such a role, might be a good legitimating argument to support the existence of strong constitutional courts – but, for the reasons spelled out earlier, it fails to perform that role satisfactorily in the discourse on the legitimacy of judicial constitutional review. 5. CONCLUSIONS Constitutional courts in CEE, as elsewhere, have faced the legitimacy dilemma, which has been particularly acute when they performed the role that represents their main raison d’être: the articulation of the true meaning of constitutional rights, and
IN SEARCH OF LEGITIMACY
63
the invalidation of legislation on the basis of its inconsistency with those meanings. Based on the tacitly accepted – and never fully defended – fiction of the objectivity of the rights articulations, these courts have had to forsake the strategy that would have offered perhaps the most candid and convenient defence of their legitimate role in overturning democratic legislation in this way; namely, the strategy of depicting themselves as quasi-legislative bodies, adding an extra chamber to the existing parliamentary process. They “had” to forsake this strategy because it would undermine their pretences to quasi-judicial character, and thus the privileged standing of lawyers to sit on these “courts”. The contrast between their legislative function on one hand, and the quasi-judicial staffing, procedures and rituals of these bodies on the other, calls into question the reasons for adopting the “Kelsenian” rather than the decentralised system of judicial review in the first place. None of the main rationales usually provided for this particular choice of institutional design is fully convincing, and neither does the sum of these less-than-persuasive arguments suffice. The question is, if there is going to be a system of judicial review of legislation, in particular under constitutional bills of rights, would a model other than the system of abstract, ex post and final review by specialised constitutional courts be more conducive to the solution of the legitimacy dilemmas? This question is the focus of the discussion in the next chapter.
CHAPTER 3 THE MODEL OF JUDICIAL REVIEW AND ITS IMPLICATIONS
Despite all the differences and national variations, the constitutional courts in CEE share some basic, fundamental features; indeed, the exceptions to the basic model are relatively unimportant. They will be dealt with below, in a chapter that will discuss critically the dominant model of constitutional review in postcommunist countries of CEE, in terms of three significant dimensions: the power of constitutional courts to conduct abstract review, in addition to concrete review and to consider constitutional complaints by citizens; their power to review the constitutionality of laws ex post, that is, after they have already entered into force; and the finality of their findings of unconstitutionality. As will become evident, this chapter builds upon the account given in Chapter 1 but my purpose here is to illustrate in more detail certain important implications of the model of judicial review chosen for reasons that (as suggested in Chapter 2), although supplied by the dominant constitutional doctrine, may not be entirely persuasive. 1. ABSTRACT REVIEW All constitutional courts in CEE conduct abstract review, though most of them combine the powers of abstract and concrete review.1 However, the European model of concrete review is centralised: most typical is the pattern whereby an ordinary court, while considering a case and encountering a problem in terms of the constitutionality of a relevant rule, stays the proceedings and asks a constitutional court (either on its own initiative, or at the behest of one of the parties) to pass judgement on the validity of the law. Hence, the constitutional court considers such a “concrete” case in an “abstract” fashion; that is, it normally does not take into account the factual circumstances of the case but merely compares the contentious law with the constitution in abstracto. (It is clear that “abstractness” of the review is a different matter from its “centralised” character; while these issues are indirectly related – it is difficult to have a model that is diffuse and abstract at the same time – of particular interest here is the question of abstractness, not that of centralisation). Different courts deal with such cases to varying degrees of abstraction; at one end of the spectrum, for example, the Hungarian Constitutional Court considers references from ordinary courts in purely abstract terms, and treat this as a virtue rather than a
66
CHAPTER 3
disadvantage.2 In consequence, the reasoning of the constitutional court in such cases does not differ significantly from its reasoning when exercising a strictly abstract review at the initiative of the president or of a group of members of parliament. This being said, the focus of this part of the chapter is not on the modalities of a concrete review, or on the degree to which the concrete review in centralized context is “contaminated” by abstractness, but on abstract review par excellence. The question is, are the constitutional courts, when exercising abstract review, behaving in a judicial or quasi-legislative fashion? And also, does review of this type magnify the legitimacy dilemmas as outlined in the previous chapter? To be sure, the contrast between abstract and concrete review must not be exaggerated; for example, the Supreme Court of the United States – our paradigmatic case here of concrete review – also exercises at times what may be called “abstract” review. This happens, for example, when the parties make challenges to statutes facially on grounds of excessive breadth or vagueness, even though the same statute would be constitutionally applicable to the party when properly, narrowly construed.3 An aspect of abstractness is also present when a party challenges a statute on the basis of a claim that it has a “chilling effect” upon its private behaviour; for instance, where a general prohibition on particular sexual conduct has not led law-enforcement officials to an actual legal action against a private person, but rather a private person, in anticipation of the possibility of such an action in future, claims that the very prohibition unconstitutionally affects his or her freedom of behaviour in private matters. The ensuing review is, strictly speaking, “abstract” because no specific application of the statute in question has preceded the review; a (putatively) unconstitutional intervention in private life is purely hypothetical and anticipated, rather than real. On the other hand, there is an element of “concreteness” to the case just described because a private party is present, and the court can at least visualise a specific application of a statute to the situation of this party in general: the rules pertaining to standing to sue assure that the reasoning in which the court will engage is not totally “abstract”, because the grievance is specific (although based on hypothetical events) and related to a particular purpose. Of course such a visualisation of concrete consequences is equally available to the European courts engaged in abstract review par excellence: when constitutional judges consider the prohibition on abortion or on the public display of religious symbols in schools, they can place themselves in the shoes of those who will feel aggrieved by the consequences of their decision; there are no set limits to the moral imagination. This indicates that “abstractness” is a matter of degree. The crucial difference, however, between “concrete review” systems, such as that of the United States, and “abstract review”, such as exists in the majority of European systems, relates to the fact that concrete review is not made at the behest of public officials, and in particular, of a group of parliamentarians (assumedly, those unhappy with the recently adopted statute). The legislators in the United States are allowed to sue only to vindicate some concrete individual interests connected with their official duties (such as the right not to be excluded from the Congress) but never as representatives of their constituents and never to test the constitutionality of laws adopted despite their objections. The ground for this is that, when “the individual legislator is really
THE MODEL OF JUDICIAL REVIEW
67
only complaining of a failure to persuade fellow colleagues”,4 the method is to pursue a legislative, rather than judicial, solution. This, therefore, renders more remote the prospect of a constitutional court performing the role of the “third chamber”. As the Supreme Court of the United States once proclaimed, “[i]t was never thought that . . . a party beaten in the legislature could transfer to the courts an inquiry as to the constitutionality of the legislative act”,5 and this is exactly what happens in cases of abstract control exercised by a constitutional court where the process has been initiated by parliamentarians or other political officials. As the above paragraphs indicate, the abstract versus concrete distinction is separate from a classification of systems of judicial review into centralised and decentralised (or diffuse): the former exists when there is a single body endowed with the power of constitutional review (as with continental European constitutional tribunals); the latter, when every court has the power to decide on the constitutionality of an act applicable to the case before it (as, e.g., in the United States, Canada, Australia, Sweden, Japan and India). Decentralised review is always concrete, but centralised systems can be abstract or concrete, or mixed. An interesting case of such a hybrid system is provided by Estonia where, apart from the normal functions of the National Court (or, more precisely, its Constitutional Review Chamber) in terms of abstract constitutional review, any lower court can petition the National Court regarding the constitutionality of a law applicable in a case before it. In itself there is nothing surprising about this, as most European (including most CEE) constitutional tribunals can be activated in this way. What is, however, peculiar is that, if a lower court, in the trial of a case, concludes that an applicable law violates the Constitution, the court not only petitions the National Court to determine its constitutionality, but also “shall declare it to be in contradiction with the Constitution”.6 It follows that every Estonian court has the power to pronounce on the constitutionality of any law (a decentralised and concrete review), and only subsequently will such a declaration trigger proceedings before the National Court. However, for the sake of argument, let us consider here a distinction between “abstract” and “concrete” review in their pure forms. What are the implications of this distinction for the position of the constitutional court? Offhand, it seems obvious that in the process of considering a law in abstracto the court behaves much more as a quasi-legislator than as a judicial body, and that the implications for the general allocation of powers are thus much more grave than when the review is limited to the adjudication of concrete cases. One of the reasons for this is that abstract review, when (as is often the case) addressed to a recently adopted law, brings the court right into the center of a political controversy. It is for this reasons that Bruce Ackerman has argued that CEE constitutional courts – and the Hungarian Court, in particular – should abandon their role of abstract review altogether and be limited to concrete review only. As Ackerman wrote: “if a court must wait for a specific complaint by an ordinary individual, it may take a while before a bitterly politicized dispute makes its way into the judges’ chambers”; this pause, and the resultant “cooling down” of the political passions, “provides the court with a valuable political resource to sustain itself as part of the new constitutional order”. 7
68
CHAPTER 3
According to Jürgen Habermas, the main reason that concrete review is less problematic than the abstract review when exercised by constitutional courts is that “rulings on constitutional complaints and the concrete constitutional review initiated by individual cases are both limited to the application of (constitutional) norms presupposed as valid”,8 and, for these forms of review, Habermas goes on to note that “the distinction between discourses that apply norms and those that justify them offers at least an argumentation-theoretic criterion for demarcating the respective tasks that the judiciary and the legislature can legitimately accomplish”.9 I am not sure whether this particular distinction, between the application of norms and the justification or criticism of norms, indeed marks the fundamental distinction between concrete review (and constitutional complaint) on the one hand, and abstract review on the other. While I agree with the conclusion that concrete review is more compatible with the endeavour to keep legislative and judicial functions separate from each other, I am not persuaded that this is because the argument concerning the application of a norm is more compatible with concrete review rather than with the abstract one; consequently, I am not convinced that the presumption of constitutionality of a statute is more at home within concrete review rather than abstract. In all likelihood this is the case, but for rather more contingent reasons than Habermas suggests in the passage quoted immediately above. After all, there is no general theoretical reason why a court, when scrutinising a statute as a by-product of hearing a specific court case, should adopt a stronger presumption of validity of that statute than a court subjecting the statute to abstract scrutiny. As soon as the court engages in a consideration of the constitutional validity of a statute, the level of deference that it pays to the legislative choice reflected therein is no longer governed by the path by which the matter arrived before it. The distinction between application and justification of norms, and the resulting strength of the presumption of constitutionality, does not stand (it seems to me) in any particularly close connection with a concrete/abstract review distinction. The reasons why abstract review raises more acute legitimacy problems than concrete review have to do with the different rationale available to a court engaged in concrete review for striking down (if it so decides) a statute, and the related availability of precautions against excessive judicial activism. Consider the original rationale for the system of judicial review as established in the United States in 1803, under Marbury v. Madison.10 Contrary to the popular opinion, Chief Justice John Marshall did not base the Supreme Court’s power to invalidate the acts of Congress on an understanding of the Court’s role as a watchdog of the constitutionality of legislative acts. The conventional argument that the existence of constitutional constraints necessitates the granting of power of the Court to declare when Congress has overstepped its competence does not figure in Marshall’s reasoning; not explicitly, anyway. If Marshall’s opinion in Marbury v. Madison is read carefully, it becomes clear that the whole construction of this implicit power of the Court is based on one, rather simple, argument: the court (any court, not just the Supreme Court) has to apply the law in order to decide a specific case; if there is a conflict between two laws that control the case at hand, the court has to decide which law should be given precedence; if the Constitution and a lower act clash on the same issue, the Constitution must prevail. As Griffin summarises it, the
THE MODEL OF JUDICIAL REVIEW
69
argument is based on the proposition that “the judiciary must side with the Constitution in the case of a direct conflict between it and an ordinary legislative act”;11 this obligation, however, is a direct result of the court’s duty to decide concrete cases rather than of any notion of judicial supremacy in matters of judgments on the constitutionality of acts. Consequently, the declaration that an act of Congress is invalid is presented as an outcome of the practical necessity of having to decide a particular case: “Those who apply the rule to particular cases, must of necessity expound and interpret that rule. If two laws conflict with each other, the courts must decide on the operation of each,” explains Justice Marshall:12 and, as the Constitution is addressed to the legislatures as well as to the courts, the latter have no choice but to declare invalid any act found to be inconsistent with it. On the face of it, this is an argument which is qualitatively different from one based upon the idea that the Court has to be a watchdog of the Constitution, and to oversee congressional behaviour under that Constitution. It is because of this nature of Marshall’s discourse in Marbury that Neil Devins concludes: “Marbury, while establishing judicial review, does not come close to suggesting that elected government also cannot interpret the Constitution or that elected government must treat Supreme Court decisions as definitive”.13 There is hardly any place for the idea of judicial supremacy in Marshall’s view: the Court has no choice but to discard the act in order to apply the Constitution, but Congress or the executive may insist on its own interpretation thereof, different from that of the Court. Martin Shapiro describes in the following manner a hypothetical explanation given by a constitutional judge in a concrete-review system, in accordance with the Marbury formula: “‘We judges do constitutional review only because we cannot help doing so in the routine disposition of ordinary cases…’”.14 As Shapiro immediately observes, abstract review “most clearly strips reviewers of their judicial cover” contained in that explanation. It is important not to overstate this point. The very act of deciding that one needs to apply the Constitution as against a statute involves a judgment that the statute is contrary to the Constitution – and, as such a judgment will almost always be a controversial matter, especially if made under vague rights provisions, such a judgment thus inevitably appears as an act of a watchdog of the Constitution. After all, the power to declare statutes unconstitutional does not necessarily follow from the judicial duty to apply laws in specific cases.15 But there is a difference in the discourse accompanying such a defence of judicial review: it is a discourse of reluctance and unavoidability/inevitability. Such a discourse need not lead to a heightened deference to a legislator (this is the point made earlier in reference to Habermas), but nevertheless it places the whole enterprise at a different level than in an abstract review. For, in an abstract review, the judge does not have to apply any law to the case at hand; she is free to exercise her judgment as to whether she approves of the way in which the legislator tried to articulate the constitutional norms in question. In contrast, in concrete review the approach is more of a “tragic choice”, triggered by the exercise of a routine, uncontested judicial task: a judge is torn, as it were, between the duty to apply the statute and the duty to apply the Constitution (as she understands it), but her direct motive is not to evaluate the job done by the elected legislators.
70
CHAPTER 3
What does this matter, beyond “mere” rhetoric? For one thing, the rhetoric of judicial review is no small thing: it places a judge in a certain political perspective vis-à-vis the legislators, and affects the whole relationship between the constitutional judiciary and the elected representatives. But the matter goes well beyond the rhetoric. In the case of a concrete review where the court simply declines to apply the statute because of its perceived unconstitutionality, the judge not so much “strikes down” the law but rather renders it a dead letter. The talk of “striking down” a statute is therefore somewhat careless,16 and it even happens that the outvoted judges announce that they would apply this very statute in later cases, despite its being now found unconstitutional by the majority.17 More importantly, perhaps, this rationale for judicial review informs the institutional precautions against judicial activism. If support for judicial review is based on Marshall’s approach, and thus makes sense only with respect to concrete judicial review, then it also makes good sense to erect precautions against judicial transgressions beyond the role justified by the rationale for said concrete review. In the United States Supreme Court’s jurisprudence, it has meant that the federal courts have jurisdiction only if a number of conditions of justiciability are met: they cannot decide lawsuits that are “moot” (because the harm, if it ever was, ceased to exist), or “unripe” (because the injury to a complaining party is as yet uncertain), or where parties cannot establish their “standing” or when the subject-matter is essentially “political,” etc. These conditions of justiciability all follow from the constitutional description, in Article III of the United States Constitution, of the role of federal courts in deciding “cases” and “controversies”, and the power of judicial review of legislative and executive acts is subject to constraints stemming directly from this role. As Alexander Bickel noted in his classic book, the concept of standing follows from the Marbury v. Madison rationale for the existence of the power of the judiciary to enforce the Constitution against the other departments of government.18 Here is how Bickel develops the rationale for judicial review, which he associates with Chief Justice Marshall’s insight: One of the chief faculties of the judiciary, which is lacking in the legislature and which fits the courts for the function of evolving and applying constitutional principles, is that the judgment of courts may come later, after the hopes and prophesies expressed in legislation have been tested in the actual workings of our society; the judgment of courts may be had in concrete cases that exemplify the actual consequences of legislative or executive actions. . . . The concepts of “standing” and “case and controversy” tend to insure this….19
In contrast, “abstract” judicial review need not be subjected to any such constraints; the rationale for abstract judicial review relies more directly upon the watchdog role of the constitutional court; thus, the constraints mentioned in the preceding paragraph do not apply here. In fact, none of the standard criteria of justiciability as exemplified by the U.S. Supreme Court jurisprudence apply, or need to apply, to abstract review. The two maturity-related requirements, that the issue be neither “unripe” nor “moot”, have no place whatsoever in abstract review; if the court is supposed to be a watchdog of the parliamentary behaviour then the very fact that a law deemed unconstitutional by a recognised challenger is on the books makes the matter eo ipso “ripe” for consideration by the court, and as long as it has not
THE MODEL OF JUDICIAL REVIEW
71
been repealed, it cannot be considered to be “moot”. Furthermore, in abstract review there is no need to adopt a “political question” doctrine as a factor of restraint that would limit the role of judges; after all, the whole point of having abstract judicial review is that the judges are invited to pronounce about the conformity with the constitution of political choices par excellence made by the parliament. To adopt a “political question” doctrine would therefore contradict the very rationale of having an abstract review in the first place, and it is no wonder that such a doctrine has not been developed, at least explicitly, in those systems preferring such a model of constitutional review. Of course, in practice, a constitutional tribunal that exercises an abstract review may manifest a great deal of restraint (including on the grounds that to decide a particular case would involve making a political judgment),20 while – on the other hand – a supreme court that strikes down an act in the process of concrete adjudication may be very activist and non-deferential in its treatment of the legislative branch. However, all things being equal, the concrete review lends itself better to a restrained review, and, therefore, has a lesser impact upon a shift in the allocation of powers from the legislature to the judiciary in terms of the authoritative articulation of constitutional norms. This last point was actually recognised by the US Supreme Court when it noted the link between its doctrine of justiciability and the principles of separation of powers: to accept jurisdiction on matters where a party does not have a narrowly defined standing but merely “suffers [an injury] in some indefinite way in common with some people”, would be “not to decide a judicial controversy but to assume a position of authority over the governmental acts of another and co-equal department, an authority which plainly we do not possess”.21 And, as the rules of standing do not follow in any obvious fashion from the principle of abstract review – though they admittedly differ from country to country – the reluctance to assume such a “position of authority”, as described in the above quotation, is not necessarily coherent with a system of abstract review. The argument thus far has proceeded on the basis that it is a good thing to maintain a clear separation of the law-making and the adjudicatory power in a state, and that concrete judicial review comes closer to this separation than the abstract one. However, of course, one may object to the very presupposition (concerning the need to constrain the law-making function of the courts as much as possible) as question-begging; and, if such a position is adopted, then obviously there is no reason to applaud the justiciability-related restraints upon judicial review. Nevertheless, if constitutional courts are to maintain their self-characterisation as judicial (or quasi-judicial) bodies, then the rules of justiciability related to concrete review help contain their intervention into the legislative process, within the limits resulting from their perceived role as bodies that apply the law (including the Constitution). If, in contrast, we relax this rationale and adopt instead a view of constitutional courts as legitimate participants in the law-making process, then all other pretences of a judicial character pertaining to their composition and procedure should be abandoned also; for example, there is then no justification for recruiting its members from within the legal profession, or for designing the procedure before the constitutional courts in a form mimicking that of the regular courts, etc. It does not follow from the foregoing that those courts that conduct exclusively concrete review have thereby escaped all of the legitimacy problems that can arise in
72
CHAPTER 3
connection with constitutional review; the persistence of the debate about “countermajoritarian difficulty” in the United States testifies to the contrary. However, the narrow point that I am making here is that, when a constitutional court is constrained by various justiciability conditions that apply logically to the concrete but not to the abstract review, they are more likely to recognise their proper role as that of applying the Constitution and other laws to the concrete circumstances of a particular case rather than as a watchdog of the Constitution against the legislative misbehaviour of the legislatures. Consider one particular argument about the meaning of the “political question” doctrine in the United States. A European legal scholar, Fritz Scharpf, commenting several years ago upon various puzzles and inconsistencies within this doctrine as established by the Supreme Court, offered a “functional” approach that explicitly acknowledged the constitutional character of the review and that would rely upon the general following criterion for the rules of justiciability: whether the fact-situation of the case, the stage of its development and the qualifications of the parties in view of the availability of other potential litigants and of the intensity of their interest in the issue itself, were optimally or at least adequately suited for the full presentation and clarification, within the limitations of the adversary process, of this particular constitutional question.22
From this perspective, the rules of justiciability serve as a way of sharpening the focus within which the court may conduct its critical scrutiny: they ensure that the constitutional question is raised in circumstances that do not merely mimic those that characterise the legislative deliberations. Rather, the focus is on the specific consequences of applying a particular law to a certain situation: the court ascertains the violation of the constitution from the perspective of an aggrieved citizen rather than that of a disgruntled parliamentary minority. The constitutional question is then raised in a way that makes the judicial procedure reasonably well-suited to identify the violation of a right: it is contextual, personalised and specific. This is how Scharpf’s description of the functional rationale of justiciability conditions should be understood; of course, however, he does not (nor does he aspire to) provide an answer to the question of why a court should consider a constitutional question in the first place. It is not my concern at this point, either; the point is rather to say that, once we have accepted the very principle of constitutional judicial review, the adoption of the model of abstract review raises the likelihood of running into certain problems that, under a system of concrete review, are more readily minimised. That is why it is important to emphasise what this section of the chapter has not sought to establish. My purpose was not to argue about the advantages of concrete over abstract review per se. Rather, it was to urge that in the judicial or quasijudicial setting, and without the legitimacy assets of a legislative body, a system of abstract review is more problematic than a concrete one, at least in terms of the relative legitimacy of the procedures. However, at the end of these remarks, I wish to bracket my agnosticism as to the moral-political primacy of one model of judicial review against the other, and suggest that, even if we abstract from the legitimacy problems involved, there are some good reasons for giving priority to concrete review over its abstract counterpart as a superior method of testing the moral and political consequences of laws already adopted. The argument runs along the lines of
THE MODEL OF JUDICIAL REVIEW
73
distinguishing between law-making and law-evaluating functions: one may say that it is a good idea to have a mechanism of reviewing the wisdom of a law some time after its entry into force, even if only to check whether the unforeseen negative consequences do not defeat the expected gains that had justified adopting the law in the first place. This rationale for judicial review comes close to what Pasquale Pasquino23 describes as “lenient legislation”, a term he applies to a review conducted by the Italian Constitutional Court when it considers the questions of unconstitutionality raised by trial courts in the course of judicial proceedings.24 As Pasquino describes it, the Constitutional Court behaves then not so much as a counter-balance to the elected bodies but rather is “reconsidering the statute law in its concrete enforcement, commenting, interpreting, explaining and up to a certain point modifying it in order to make it compatible with the principles of justice and the set of individual rights written down in the Constitution”.25 The operative words in this description are “in its concrete enforcement”, because it is only the concrete application of the law that allows any possible defects, unforeseen at the point of enactment, to emerge. In Pasquino’s words, the Court “reconsiders the legislative provision in the light of its concrete application, not a priori and abstractly like the legislator does”;26 from this perspective, the division of labour between the constitutional court and the legislator is much clearer in concrete than in abstract review. This is a similar insight to that expressed in an already-cited passage from Bickel, that “the judgment of courts can come later, after the hopes and prophesies expressed in legislation have been tested in the actual workings of our society”.27 Such a mechanism is helpful precisely when the courts can acknowledge the evidence from specific cases and complaints, and they are reasonably well-suited to such a function; they do normally look at specific cases, and assess their legal significance. This is the habit – and an institutional skill – that the courts actually possess; and the rationale behind a system of concrete judicial review thus relies on the hope “that the hard, confining, and yet enlarging context of a real controversy leads to sounder and more enduring judgments”.28 A judicial (or quasi-judicial) body may be particularly well suited to an evaluation of whether some moral principle (that can be also characterised in constitutional terms: the vagueness of constitutional language usually permits such a characterisation) is implicated by the operation of a given law, as evidenced by specific applications of that law. This is the sort of institutional competence that Michael Moore had in mind when he noted that judges are better positioned for this kind of moral insight [related to the rights persons possess] than are legislatures because judges have moral thought experiments presented to them everyday with the kind of detail and concrete personal involvement needed for moral insight. . . . One might well think that moral insight is best generated at the level of particular cases, giving judicial beliefs greater epistemic authority than that possessed by legislative beliefs on the same subject.29
This is a good statement of the moral and political theory behind the power of concrete judicial review. No doubt, the theory as outlined by Moore is very controversial. Do the concrete-review courts, such as the US Supreme Court, indeed consider the questions of constitutionality in such a tight connection with the circumstances of each specific case as Moore would require them to? 30 Are the
74
CHAPTER 3
legislatures unable to debate and decide on rights-related issues with specific cases forming an important part of the background of this process? Is consideration of such issues that is characterised by a high degree of abstraction necessarily an inferior level of reflection upon the nature of rights? Notwithstanding its controversial nature, the theory summarised by Moore contains important resources in terms of defending of a system of concrete review; these resources, however, are obviously not available to those who seek to support such a review process in abstracto. 2. EX-POST REVIEW The great majority of contemporary constitutional courts in Western democracies have only an a posteriori power of review of acts, that is, they cannot consider laws until they have entered into force. There are a few exceptions: notably the Conseil Constitutionnel in France, which can only consider (and, if it so decides, invalidate) the lois before their promulgation. The Portuguese Constitutional Court has, alongside ex-post review, a power of ex-ante review when petitioned by the President. For some time, between 1980 and 1985, the Spanish Constitutional Court also had (in addition to an ex-post review) a power of a priori review of so-called organic acts (that is, the acts that develop constitutional rights, and that are located, in terms of their status, roughly half way between the constitution and the ordinary statutes).31 However, it was abolished due to the perceived abuse of this possibility by opposition political parties: the best way to prevent an act from entering into force was to challenge it in the Constitutional Court which, in the face of a large caseload, simply could not deal promptly with the referrals.32 There are also some examples of non-judicial bodies that can review in a non-binding fashion the constitutionality of proposed legislation, such as the Law Council in Sweden, where the convention developed that the government would shelve criminal or civil legislation ruled unconstitutional by this body. In the post-communist world, the Romanian Constitutional Court has the power to adjudicate on the constitutionality of laws before their promulgation but after the adoption of the law by Parliament (alongside the ex post power, but the latter can be triggered only by the courts).33 This is the only case of a CEE country where the abstract control of statutes is only ex ante. According to some commentators, it means that the Constitutional Court plays the role of a sort of “mediator” between the parliamentary majority and minority, and becomes embroiled in political controversy.34 As the Court’s decisions taken in the course of abstract preliminary review are not final, any such decision of unconstitutionality in fact means that a law requires a higher threshold of adoption (a requirement of 2/3 majority in each chamber to override the judgment of Constitutional Court). Even though no such judgments have yet been overridden in this fashion, there have been cases in which a new bill was eventually proposed to incorporate the Court’s objections.35 In some other CEE countries an ex ante control exists in a weaker form, alongside an ex post abstract review. In Estonia, if the Parliament adopts a law that the President refuses to sign and sends back to the House (i.e., if the Parliament has
THE MODEL OF JUDICIAL REVIEW
75
overridden the presidential veto), the President can then refer the law to the Constitutional Review Chamber.36 This form of control exists alongside abstract ex post review that can be triggered by the Ombudsman, if the lawmakers refuse to amend a statute that according to the Ombudsman violates the Constitution.37 In Poland, pre-enactment review can be conducted at the request of the President (but only if the President has not vetoed the law previously). However, an ex ante review is seen in Poland by the Constitutional Tribunal as an exceptional case of abstract review, and this is reflected, among other ways, in the adoption of a stronger presumption of constitutionality of the statute than is the case in ex post abstract review.38 Finally, in Hungary an ex ante review had existed until 1998 in different forms, including not only the possibility of review after the vote and before the promulgation (at the initiative of the President, Government, fifty MPs or a parliamentary committee), but also review of bills that had not been voted on (a matter to which I will return below). Now it exists only in a minimal form, similar to that in Poland: only the President can initiate constitutional review of laws already passed but not yet promulgated. What is the significance (if any) of the distinction between an ex ante and ex post review for the purposes of the discussion of the democratic legitimacy of constitutional judicial review? It is useful to consider the leading example of the ex ante model – namely the French Conseil constitutionnel. The first thing to note is that the Council was not intended to be a protector of constitutional rights; indeed, the 1958 French Constitution does not even contain a bill of rights, and the Council was meant to operate merely as a guarantor of the separation of powers. The extension of its role, including the protection of constitutional rights (justified by reference to the preambles to the 1946 and 1958 Constitutions, the 1789 Declaration of Rights, as well as “fundamental principles recognised by the laws of the Republic”) was due to its own judicial activism, beginning with the landmark decision of 16 July 1971 when, for the first time, a bill was struck down as in breach of fundamental rights.39 Furthermore, and following from the limitation of its power to that of an a priori review only, the Constitutional Council stands out, in comparison with other European courts, in that it does not possess any power of “concrete” review; that is, it cannot review the question of constitutionality arising out of an application of a challenged law to a specific legal case. In these two respects, the Conseil is much closer to being a system of lawmaking than of judicial power. Indeed, a classic study by Alec Stone makes a convincing case for considering the Conseil as a “third chamber” of the legislature, rather than as a judicial or even a quasi-judicial body. The Conseil can invalidate a decision of the parliament in abstract terms rather than in the process of litigation; it can suggest positive solutions that would remove the defect from the law, and therefore is not merely “negative”;40 it decides on matters in which the disagreement boils down to different policy choices and where the constitutional norms in question are highly indeterminate;41 it defines policy objectives and goals for the Parliament to pursue,42 its deliberations are activated by a political action, usually by a group of opposition senators or deputies unhappy about the law,43 etc. All of these are characteristics of a legislative body, and Stone is very convincing in showing that a number of
76
CHAPTER 3
arguments that may seem to deny the “third chamber” characterisation, in fact are equally applicable to a number of entities whose “legislative” nature is undeniable. But how different, from this point of view, are the cases of those constitutional courts that possess an a posteriori power of review only? Contrary to expectations, the difference seems to be formal rather than substantive. The German Constitutional Court can do all the things that the French Conseil can do – yet it also possesses all of the powers related to the concrete review. The only difference between an a priori and an a posteriori abstract review is the absence of promulgation by the President in the case of the former. But if this promulgation is compulsory, as it is in the case of laws certified as constitutional by the Conseil, then the difference is technical: in both cases a body may displace a decision that enjoys majority support in an elected branch of the government. One obvious difference is that the a priori model brings about more stability to the system; once the law has been promulgated, no future challenge is possible. The a posteriori system introduces more uncertainty and instability to the law. Although this, in itself, does not affect the separation of powers, it is certainly an important matter. On the other hand, an a priori system seems to create an incentive for frivolous or obstructive uses of constitutional review by the opposition. It is instructive to note that this was one of the major factors in Spain’s decision to abandon a priori review; bills referred to the Constitutional Court before they became effective caused delays in the introduction of reforms devised by the Spanish government in 1983-85. As Stone says, “these referrals delayed the reforms for ludicrous periods of time”;44 this, however, is a matter that can be remedied by, for example, strict time limits imposed upon the constitutional court (as in France, where the Conseil is required to rule within one month). Furthermore, in any event, the gain in terms of stability and certainty is achieved only if the a priori review is the only available type of review in a given system; as soon as it is supplemented by ex post forms of review, this certainty is inevitably and irretrievably contaminated. A more interesting question is to ask what would be the consequences if constitutional courts were permitted, or even required, to express their views regarding the constitutionality of proposed bills before the vote in the legislative chamber(s) had been taken; in other words, if they had the power to issue advisory opinions at earlier stages of the law-making process. Some constitutional rules in CEE countries expressly prohibit such an involvement (e.g., in Slovakia45 and in Estonia46). The case of Hungary is illuminating: the Constitutional Court there had possessed such an authority, but it renounced it itself at an early stage in its life. One major problem that the Court had faced was that a bill would be sent at an early stage of its drafting, and it would then undergo many changes, not necessarily due to the Court’s judgment; the judgment, however, would apply only to the early version in which it had been submitted to the Court.47 The Court attempted to avoid such a possibility, without renouncing the very principle of preventive review, by establishing in 1991 that “a claim can . . . be submitted only before the final vote in Parliament”.48 At that time, the Court decided not to consider the motion of the Union of Free Democrats regarding the constitutionality of the reprivatisation law, on the basis that the project was still at a too early stage; in doing so, it appealed to the concept of the separation of powers49 and proclaimed that the review can only
THE MODEL OF JUDICIAL REVIEW
77
occur when a project is final (indeed, after the motion to review, more than 400 proposals of changes were made regarding this draft law).50 The advisory function may be also performed by the court in an indirect fashion, without making this function explicit. In Bulgaria, this may be achieved through an abstract interpretation of the Constitution by the Constitutional Court. This task, performed by the Court in abstracto and not in connection with any particular statute already adopted, may be seen as “an instrument of improving the legislation”; in the process of preparing a law the bill’s sponsor may ask the Court to provide an interpretation of the Constitution, and therefore indirectly pronounce on the constitutionality of then proposed statute.51 The reasons given against allowing any advisory role for the constitutional courts at the law-making stage usually cite the principle of separation of powers, and, more specifically, “the danger of direct political involvement” of the court.52 As a Hungarian scholar puts it, preliminary norm control “has the potential to threaten the principle of separation of powers; in essence, the C[onstitutional] C[ourt] would replace the legislature by seriously interfering in its work during the legislative procedure”.53 A Czech Constitutional Court judge has reacted in the following way to the very idea of his Court’s possible involvement in issuing advisory opinions: “It would make us too powerful. We are quite powerful the way we are now, and I am happy with the [amount of] power that we have”.54 Practical reasons are also often cited, that raise the spectre of overload of the agenda of constitutional courts. As a constitutional justice in Bulgaria noted in an interview, if the power of advisory opinions were given to the court, it is likely that every legislative project would be sent there for review, thus resulting in a situation in which the court would become a “permanent advisory council” of the parliament. “It will replace [sic] our function completely”, he added.55 However, one wonders whether it would make such a great difference, once the constitutional courts have been given (or have usurped) a strong power to displace legislative judgment after the act has been enacted. Perhaps an early warning system might help avoid a subsequent invalidation? Interestingly enough, one of the arguments against the issuing of advisory decisions by courts is that it constitutes a waste of judicial resources,56 but at the same time, by removing the uncertainty over the constitutional status of a proposal, a system of advisory opinion may save other resources (legislative and of private parties); therefore, in consequentialist terms, there is a trade-off there, and the costs and benefits may well balance each other out.57 Further, perhaps an advisory role of the court would de-dramatise the process of distorting the policy designed by the representative branch? It is likely (as Stone Sweet suggests in his comparative analysis of Western European constitutional tribunals) that even an ex post review strongly affects the choices of lawmakers, by encouraging them to anticipate, and avoid, grounds for future invalidation of their acts.58 If this is the case, the added correction resulting from the power of advisory opinions might not be such a qualitative change, and an open and explicit opinion expressed by the court at an early stage would have the effect of bringing more transparency to the process. It would also help save resources consumed in legislative decision-making (considering that the costs of decision-making by a constitutional court are lower than the costs of parliamentary decision-making) and
78
CHAPTER 3
provide vital information to lawmakers. Furthermore, contrary to some concerns, it is hard to see why the power to give advisory opinions ex ante would compromise the independence of the court and turn it into “an organ loyal to the Parliament.”59 The power to give expert advice is not contrary to, but rather, presupposes, a degree of independence. As a matter of fact, some (non-constitutional) Supreme Courts (e.g., in Poland) have the express power to issue advisory opinions concerning proposed laws – and this has not been understood as compromising separation of powers, or judicial independence in any way. The most frequent argument against an ex ante review is that it necessarily entangles the court in the political bargaining that is carried out at the stage of legislative drafting. As a Russian constitutional judge says: “We would be involved in political games”.60 However, the way in which such a “politicisation” is anticipated in this particular judge’s explanation is very telling. President Putin, explains Judge Vladimir Yaroslavtsev, can currently count on the majority of deputies who will comply with his legislative wishes. Suppose that the opposition will then be able to send the bill to the Constitutional Court and the latter agreed with the opposition’s objections. “What to do in such a situation? The President has decided on the steps to be taken; the Constitutional Court decided [differently]: what to do?”61 This, according to the judge in question, leads to political instability; but of course such a picture presupposes that the President is in a dominant position, in full control of a loyal parliamentary majority, and therefore can impose his own understanding of the meaning of constitutionality upon the law-making process without the risk of any irritating vetos. The fact (referred to by the judge in the same interview) that, in Russia, even after the laws have been enacted so few challenges thereto originate from the legislative minority, in spite of its full constitutional standing to do so, confirms further that the hostility towards an ex ante review is in fact underwritten by the acceptance of the dominant position of the executive in the law-making process. It is worth adding that, in a system of a decentralised and “concrete” judicial review, not all courts resist the idea of providing advisory opinions to legislators. Such resistance might seem unsurprising as the very nature of a concrete review system presupposes that a court may decide concrete cases only; legislative proposals are even further removed from concrete cases and controversies than abstract, valid laws are, in that they may or may not develop into real cases in the future. However, while the Supreme Court of the United States has, from the very beginning, rejected such a possibility on the grounds of separation of powers,62 some state constitutions in the United States nevertheless permit their highest courts to issue advisory opinions.63 Be that as it may, it seems that the principle of separation of powers provides stronger arguments against the availability of advisory opinions in a system of concrete judicial review than in an abstract system. This is because one might argue that “the ideal of an independent, apolitical judiciary would be undercut if judges expressed an opinion about a law that might later come before them in a lawsuit”.64 No such concern, however, applies to a system in which judges are called upon to review the law in abstracto. Incidentally, the idea of an “advisory” role for the court vis-à-vis the legislature has recently found its proponents even among the federal US judges – despite the
THE MODEL OF JUDICIAL REVIEW
79
fact that it is the federal judicial system in the US that has been most vehemently opposed to the adoption of this type of function. Such is the nature of the conception of a “constitutional remand”, advocated consistently and continuously by Judge (and Professor) Guido Calabresi of the US Court of Appeals for the 2nd Circuit – even though it would apply only to one particular case of putative unconstitutionality, namely, when changed circumstances may have eroded the force of the original rationale of the legislation in question.65 (Also, it would apply only to those cases in which, according to the Court, the law is neither plainly constitutional nor plainly unconstitutional66 – although, of course, such a characterisation can easily be concocted at a judge’s will). It is, therefore, an example of a proposed “advisory” role of the Court, but not, strictly speaking, “ex ante” – in fact, it applies to legislation that has been in force for such a long time that a suspicion of desuetude is justified. (It can, however, be understood as “ex ante” in the literal sense – because it is “ante” the proposed legislative action, namely, the action of rescinding the old legislation). The idea is that if a law has been on the books for quite some time, and there is a strong suspicion that its continuing validity results from legislative inertia rather than deliberate legislative support, then it is proper for the Court, if it has its doubts as to the law’s constitutionality, to ask the legislature whether it wishes to renew its commitment to the legal rule in question. Judge Calabresi, in making this proposal, refers both to American constitutional writings by Alexander Bickel and Harry H. Wellington,67 and also to the practice of the Italian and German constitutional courts that make use of such a technique to “put their parliaments on notice that a serious and thoughtful legislative review and reconsideration was in order”.68 This, according to Judge Calabresi, is a reasonable middle ground between “jumping in and striking the laws down” on the one hand, and “leaving them undisturbed and thereby allowing legislative inertia to dominate”, on the other.69 In such circumstances, if the legislature decides not to re-enact the remanded law, it is the legislature and not the court that is responsible for the invalidation, thus avoiding all of the problems concerning democratic legitimacy; if, however, the legislature reenacts the law despite the remand, this fact should weigh on the court’s mind as a reason to uphold the rule.70 The Court has provided its advice; the elected representatives of the people have considered it, and then acted as they thought appropriate: they will then “face the consequences of their decision before the people”.71 3. FINAL REVIEW In the post-communist world, only the Romanian Constitutional Court has a lessthan-final power of review, as its decisions concerning unconstitutionality can be overridden by a parliamentary supermajority consisting of two-thirds of both chambers. It applies only to the decisions of the Court taken in the process of abstract review (which in Romania precedes the promulgation); any judgments rendered in the course of concrete review are final. It should be added that, thus far, not a single finding of unconstitutionality by the Court in Romania has been overridden by the legislature (although this was not for the want of political will but because of insufficient numbers in the Parliament). Also in Poland, the decisions of
80
CHAPTER 3
the Constitutional Tribunal were non-final and subject to being overridden by Parliament under the regime of the so-called “Little Constitution”, now superseded by the new, fully-fledged Constitution of 2 April 1997.72 The parliament in Poland was successful in overruling decisions by the Tribunal eleven times until the possibility of so doing was constitutionally discontinued.73 Non-finality is universally considered a sign of institutional weakness in a constitutional court,74 and as inconsistent with the very principle of the rule of law.75 The general view among the judges of constitutional courts in the region, and clearly the majority view within the doctrine, is unqualifiedly hostile towards the possibility of a legislative override of constitutional courts' decisions. In my various discussions with justices of constitutional courts in CEE, I have not yet met one who would favour the introduction (or, in the case of Romania, the maintenance) of the legislative override. As a constitutional judge in Russia told me, to allow such a possibility would turn the court into “something like a committee for constitutional affairs” and its decisions would then be “just like recommendations”.76 It should be also added that the persuasive force of the Romanian system as an example to other CEE countries seems extremely weak; in fact, it is practically non-existent. In a neighbouring Bulgaria, for example, the idea of a Romanian-style legislative override was never publicly put forward in the debate on the proposed Constitutional Court law, and as a prominent Bulgarian legal scholar has noted, it is likely that “it was not known” that Romania even had such a system.77 However, for those who deplore the anti-democratic consequences of judicial power, the non-finality of judicial pronouncements can offer a way of reconciling democratic decision-making with constitutional review. The restriction of the power of constitutional tribunals to review parliamentary acts in this way means that legislators and the general public are asked to have a second look at proposed legislation, and take into account the constitutional aspects that perhaps had not been considered in sufficient detail the first time around. It is a power that slows, but does not derail, the operation of majority rule. However, even if we put to one side the possibility of a formal parliamentary override, the power of constitutional review should not be seen as a matter of a dichotomy – either the constitutional tribunal’s decisions are final, or they are only tentative – but rather as a point on a continuum. At one end of the spectrum, the tribunal’s decision adds only an insignificant cost to the legislative process and the will of the legislators is subverted only to a minimal degree; at the other end, the cost of overriding the non-majoritarian body is very high. Of course, the court’s decision is never “final” in the literal sense; in lawmaking, there is no such thing as having “the last word”. For one thing, any judgment can always be overridden by constitutional amendment. The degree of “finality” of decisions of constitutional courts may then be measured by the degree of entrenchment of the constitution: the easier it is to amend the constitution, the less final are the court’s decisions. On the one side of spectrum, there will be cases of an “absolute” entrenchment: when the constitution itself precludes a possibility of amending certain provisions. Such an absolute entrenchment automatically solidifies the “finality” of the constitutional court’s decisions, and, in consequence, the court’s power. This is the case, for example, with the German Basic Law’s provisions on
THE MODEL OF JUDICIAL REVIEW
81
rights (and also on the federal system): the Basic Law explicitly states that amendments to these provisions are “inadmissible”.78 As other examples, coming close to “absoluteness”, one might cite Spanish constitutional rights provisions that may be amended only by a two-thirds majority in each chamber, followed by a dissolution of the parliament, followed by a successful referendum over these amendments; the procedure was described by one Spanish scholar as “diabolically complex” and making it “practically impossible” to amend the relevant parts of the Constitution.79 In France, “the republican form of government” is protected against any amendments.80 The same effect may be achieved, more indirectly, by labelling certain constitutional rights “irrevocable”; the term that, for example, the Bulgarian Constitution reserves for “fundamental civil rights”.81 The “absoluteness” (or irrevocability) of a constitutional provision is, legally speaking, a suspect construct: while a constitutional provision (A) may declare an amendment of some other provision (B) inadmissible, it cannot protect itself against a future amendment, and amending the anti-amendment provision A will open, legally speaking, a way to amending the “absolutely entrenched” provision B. Hence, the absoluteness of an entrenchment should be viewed as an emphatic moral and political declaration but not as a legal prohibition on any future amendments, strictly speaking.82 On the other side of the spectrum, there will be constitutions that are weakly entrenched, that is, those that envisage a reasonably easy amendment process: for example, in France (with respect to any constitutional provision other than that providing for the republican form of government), a three-fifths majority of both chambers sitting together is sufficient.83 However, even weakly entrenched constitutions (from a formal point of view) enjoy a degree of protection against amendments, which often is of a political rather than legal character. In more stable democracies there may be strong political conventions to “deter politicians from constantly tinkering with the constitution to promote short-term or partisan ends”.84 Even absent such conventions, often the very fact of changing the constitution may be seen to be politically risky even if legally quite feasible. There is, in particular in times of general political or social instability, a certain value attached in public perception to the continuity and stability of the constitution, and even a minor amendment may be seen as symbolically detrimental because making an inroad to an implicit convention about the need to protect the stability of the Constitution as such. As one Russian constitutional judge told me: It is very dangerous, in the Russian society, to try change the Constitution. . . . We have so many elements of political instability here, that it is important to keep the Constitution as the cornerstone, . . . as an element of stability. . . . It is not the right time to try to change the Constitution.85
The fact that a constitutional court’s decision may be overridden by a constitutional amendment is sometimes used as an argument to defend robust constitutional review against the charges of democratic illegitimacy: after all, the democratic bodies have a means of prevailing over them, if only through amending the constitution.86 However, choosing a constitutional path for overriding a court’s decision is a risky and costly affair, and the subject-matter of a specific court decision (even if the decision is resented by the parliamentary majority) may not
82
CHAPTER 3
lend itself to a constitutional treatment; this is why the examples of amendments introduced specifically to override constitutional court decisions are very few and far between. As far as Western Europe is concerned, Alec Stone Sweet has identified only one case of legislative overturning of a constitutional court’s decision in the four countries of his study (Germany, France, Spain and Italy), namely in France in 1993, when the right-wing majority in the Assembly and the Senate revised the Constitution in order to permit the promulgation of a controversial immigration and asylum law (that had been earlier annulled by the Conseil constitutionnel).87 In Central and Eastern Europe, I know of only one example of a constitutional court being overridden by constitutional amendment: in Hungary in 1990, the Court found unconstitutional the absence of a right to vote in general elections by Hungarian citizens abroad, and referred the law back to the Parliament demanding a correction. As a consequence, the National Assembly adopted a constitutional amendment that explicitly deprives those persons of a right to vote.88 Constitutional amendments may be costly and burdensome, but not necessarily much more costly than the supermajority needed to override (through a nonconstitutional procedure) a court’s decision in, for example, Poland (until 1997) and Romania. As a matter of fact, in Poland until 1997 the requirements for a constitutional amendment were exactly the same as those for a decision overriding a decision of the Constitutional Tribunal, and this fact served as the basis for one commentator’s remark that “the override [was] tantamount to [a constitutional] amendment.”89 In Romania, the difference is that, while both the override and the constitutional amendment require the same parliamentary majority, the amendment also requires a referendum.90 It should be noted that in Romania – currently the only CEE country in which decisions of the Constitutional Court concerning the unconstitutionality of statutes can be overridden by the parliament – the views on this institutional arrangement are divided, although it would probably be accurate to say that the majority of lawyers and commentators consider it to be an aberration. The prevailing opinion among lawyers and other observers of the Constitutional Court seems to be critical of the parliamentary override, on the basis that it weakens the authority of the Constitution. As one observer has noted, the availability of an override is “a mistake”, because the “Constitutional Court should control all the constitutional affairs”.91 The same observer believes that the only reason why the parliament never has actually overridden any Court’s decision of invalidation is that the Court never has actually struck down “any really important laws”.92 An eminent constitutional lawyer and an ex-judge of the Constitutional Court himself, Professor Mihai Constantinescu, claims that “this is not a good solution” and as a result, the Constitutional Court’s decisions are not really “decisions” but only “suspensive vetoes”.93 At the same time, he admits that the override is a “largely theoretical possibility” because of an extremely harsh requirement of two-thirds in both chambers.94 That is why, he suggests, the possibility of an override never really concerned him during his period of tenure as a judge. The President of the Constitutional Court has stated that the availability of override is “very negative”, and that while this provision is on the books, “you cannot have a powerful Court”.95 In an interview, he expressed the hope that “in future this will be abolished”, and added that it was the unanimous view of
THE MODEL OF JUDICIAL REVIEW
83
all of the judges on the Constitutional Court.96 However, the opposite view can also be found, even among lawyers who are otherwise supporters of a strong and independent role for the Constitutional Court. For example, Mr. Horatiu Dumitru, now (at the time of the interview) a private lawyer, and formerly a lawyer for the government closely involved in proceedings before the Constitutional Court, believes that the override expresses the fact that the parliament is “the single representative body”, and that an override debate gives the Parliament an opportunity to reconsider the law.97 He also argued that the possibility of an override does not affect the Court’s decisions. As an example of a mechanism that aims at reducing the finality of constitutional court decisions in order to inject a degree of democratic deliberation into the essentially non-democratic process of judicial review, one may consider the “notwithstanding” provision of the Canadian Charter of Rights and Freedoms. This provision, in s. 33 of the Charter, states that “Parliament or the legislature of a province may expressly declare in an Act of Parliament or of the legislature . . . that the Act or the provision thereof shall operate notwithstanding a provision included in” the Charter’s catalogue of freedoms and rights.98 These declarations can be in effect for up to five years, which is the longest period of time for which a government can stay in power without going to the polls, but they can be renewed indefinitely. In effect, the Supreme Court’s decision that a particular statute violates one of the Charter rights is subject to overruling by a parliamentary process. Section 33, admittedly inserted into the 1982 Charter as a matter of political compromise and used sparingly so far, may be seen, as one American enthusiast of the provision has described it, as “an effort to have the best of two worlds: an opportunity for a deliberative judicial consideration of a difficult and perhaps divisive constitutional issue and an opportunity for electorally accountable officials to respond, in the course of ordinary politics, in an effective way.”99 The benefits of this approach seem significant: it allows the court to register its constitutional protest, puts the burden upon the legislature to face the constitutional issue explicitly, and symbolically identifies the problem in a manner highly visible to the electorate, yet it does not distort the legislative will as the requirement of a supermajority in order to override the court’s decision necessarily does. It seems like a good compromise between ordinary politics and constitutional concerns, which enhances popular deliberation over constitutional norms without distorting the democratic will. If we believe that the articulation of constitutional norms is a matter of concern not only for the constitutional courts but also for legislatures, executive branches, and the general public, then the s. 33 compromise may be seen as an attempt “to make ordinary politics and constitutional law penetrate each other”100 in a way that benefits society overall. It should be acknowledged, however, that the actual experience with the “notwithstanding” clause (as contrasted to an ideal theory underlying the clause) is, according to a number of Canadian experts, rather disappointing. According to Jeremy Webber: it is fair to say that section 33 has achieved very little legitimacy outside francophone Quebec, with the partial exception of Alberta. . . . . Section 33 has … made very little difference to the way in which rights are conceived under the Charter, at least in
84
CHAPTER 3 English-speaking Canada. The mythology of judicial review on rights grounds remains very strong – the sense that courts protect rights, unproblematically defined, against legislative or executive oppression.101
The United States Constitution provides for a mechanism of majoritarian constraint on judicial review, in the form of the Article III power of Congress to regulate the jurisdiction of the federal courts. Theoretically at least, Congress might use this power to foreclose judicial consideration of constitutional challenges to legislation, but, in fact, although some constitutional lawyers have no doubts as to the constitutionality of such a power of foreclosure,102 this has never served as a significant limit upon judicial review. There are various reasons, both political and legal, why the Article III power has never played an analogous role to the Canadian Charter’s section 33 in insulating controversial legislation from judicial review.103 However, the very fact that such a power exists suggests that, even in a system that is seen as the model for strong judicial review, the “finality” of the judicial decision to invalidate a legislative act is qualified. Some writers believe that the Supreme Court’s decisions are never really the “last word” on the matter; rather, they serve to initiate a complex dialogue between the courts and the elected branches of government, in which the latter may attempt to counter the effects of the decision. In a recent study of such interaction between the Supreme Court, the legislative and the executive, Neal Devins has shown that the legislative and executive branches have successfully restricted the impact of the Supreme Court’s landmark decision on abortion,104 and, in consequence, have forced the Court to re-examine and qualify its own, earlier decision. As Devins concludes, “once a Supreme Court has decided a case, a constitutional dialogue takes place between the Court and elected government, often resulting in a later decision more to the liking of political actors.”105 Devins is correct in saying that, on issues where constitutional interpretation is at stake, “the last word is never spoken,”106 and that the articulation of a “true” meaning of constitutional norms is as much a task for the legislature, the executive, and the general public as it is of the Supreme Court. It is also the case that the legislative and the executive branch have numerous methods of prevailing over the Court in its interpretation of constitutional rights,107 although sometimes it may take a long period of time, as, for example, the protracted resolution of the child labour issue in the United States indicates.108 Finally, one should add that the finality of tribunals’ decisions may on occasion be seen by legislators as an advantage rather than as a countervailing, antagonistic power. The fact that legislatures work “in the shadow of judicial review” may provide them with a good excuse for not taking the decisions that the electorate demands – by anticipating the tribunal’s objections or by shifting the responsibility for an unpopular decision to the tribunal. It may provide a convenient excuse: “We wanted to adopt this law, or this policy, but the tribunal would not let us do it.” Or, conversely, the tribunal’s strong authority may free the parliament to behave irresponsibly. Individual members of a parliament can display their “correct” attitude (valued by the majority of their constituency) by voting for proposals that they know will not actually become law because the tribunal will strike them down as unconstitutional.109 Ironically, the constitutional court’s capacity to prevail over
THE MODEL OF JUDICIAL REVIEW
85
the legislature may often serve the latter’s interests quite well, although perhaps not for the right reasons. 4. CONCLUSIONS In this chapter, I have assessed the main features of judicial review in CEE in terms of three contentious oppositions that are applicable to the issue of judicial review globally. These are: abstract versus concrete review; ex post versus ex ante review; and lastly the decisions of constitutional courts as final versus the possibility of override. I argued that, in terms of the first opposition, the choice of abstract review went to the very heart of the legitimacy dilemma; a concrete review, confining itself to the consideration of a particular case, lends itself to the creation of various constraints, both of a rhetorical and institutional character, that minimise the legitimacy problem of constitutional courts. In turn, even if abstract review is adopted, there is no compelling reason to object strenuously to an ex ante review, even of a purely advisory character at the early stages of drafting a bill: it may be a prudent way of saving legal and political resources by establishing an early warning system that would alert the legislators as to possible concerns over the constitutionality of certain proposals. Finally, the insistence on the finality of the decisions of constitutional courts reveals a misunderstanding of the nature of inter-institutional constitutional dialogue, which should be aimed at reasoned deliberation over the most plausible articulation of vague constitutional mandates. In sum, the current system of judicial review as established in the post-Communist countries of CEE does much to magnify the legitimacy problems, and little to minimise the potential for clashes with the legislatures. The nature and contours of those clashes will be discussed in more detail in the next chapter.
CHAPTER 4 CONSTITUTIONAL COURTS AND LEGISLATION
In any attempt to examine the role of constitutional courts in CEE, it is crucial to locate the analysis within the context of their relationships with the relevant legislatures. For example, whether these courts can be properly characterised as “activist” cannot be judged in isolation from the other institutions but only by reflecting upon whether the courts enter into the political space which is occupied (or at least, claimed) by another institution: “activism” is a necessarily relational notion. The question of the legitimacy or otherwise of judicial review of constitutional rights cannot be considered in isolation from our baseline views about what are the legitimate prerogatives of other institutions, and in particular those endowed with (what we normally consider to be) the legitimate mandate to make universally binding law. The relationship of constitutional courts to legislatures is therefore central to the main concern of this book. It is important to emphasise that the dynamic relationship between these two institutions is mutual: not only do the parliamentary systems determine the way in which the constitutional courts operate (as will be discussed in parts 2 and 3 of this chapter), but the courts also affect the parameters of legitimate law-making by parliaments (part 1). 1. THE IMPACT OF CONSTITUTIONAL COURTS ON LAW-MAKING The experience of CEE constitutional courts vindicates the thesis established by Alec Stone Sweet with respect to their West European counterparts, namely that “constitutional courts ought to be conceptualised as specialised legislative organs, and constitutional review ought to be understood as one stage in the elaboration of statutes”.1 This seems quite obvious – although not to many legal scholars who often prefer to perceive constitutional courts as judicial organs; following the legal fiction propounded by the courts themselves, they tend to situate them within the judicial branch within the general tri-partite scheme of separation of powers (as discussed, in more detail, in Chapter 2(2)). Those, however, who study the role of European constitutional courts from the perspective of political science, and are interested in the real impact of those courts upon the broad legal and political decision-making processes in society (like Stone Sweet himself), realise that the impact of those courts upon legislation goes far beyond the traditional paradigm of judicial bodies. Significantly, it was a Bulgarian political scientist, not a legal scholar, who, when discussing the Bulgarian Constitutional Court, observed that it “has acted as a second chamber of parliament, resolving conflicts within the legislature”.2
88
CHAPTER 4
Rumyana Kolarova drew this conclusion mainly from the fact that a great majority of appeals to the Constitutional Court were made by the opposition (twenty one out of twenty seven cases decided by the Court in its first year and a half of existence): “All parliamentary decisions on controversial issues were immediately appealed by the opposition of the moment”.3 However, the invalidation of the choice of the parliamentary majority is not the only way in which these courts can be said to participate in the law-making process. Another method, occasionally used, has been the actual rewriting of the laws in question or, less radically, the setting of specific deadlines for parliaments and general guidelines about how the law adjudged defective should be amended. Before considering any examples of such behaviour, we should note that it is anathema to those who endorse Kelsen’s idea of constitutional courts as merely “negative legislators”, and also to those who situate themselves within the American-style model of concrete and ex-post judicial review which rejects the possibility of any positive impact of a decision by the court upon the substance of legislation. Such a position was recently reasserted without any equivocation by Judge Altimari of the US Court of Appeals for the 2nd Circuit: Just as we ordinarily do not issue advisory opinions, we should not suggest to Congress that it ought to adopt proposed legislation. Our role is limited to interpreting and applying the laws that Congress passes, and striking down those that we conclude are unconstitutional.4
This was in response to his fellow judge Guido Calabresi who (in a concurring opinion in the same decision) reasserted his idea that, when an appellate court is faced with legislation that might have been sound at the time of its adoption, but which, due to changed circumstances, may lost its original justification, the courts might simply ask the legislature whether it still wishes to uphold a law that has become, with the passage of time, constitutionally dubious. In supporting this proposed technique, which was already mentioned earlier in Chapter 3(2) here, Judge Calabresi gave the examples of the German and Italian Constitutional Courts that, from time to time, “announce[] that laws . . . [are] heading towards unconstitutionality. . . . In this way, the continental Courts have put their parliaments on notice that a serious and thoughtful legislative review and reconsideration was in order and that failure to undertake such a review might in time result in judicial action and perhaps even nullification of the laws”.5 The technique of “putting the parliaments on notice” has been frequently used by some CEE constitutional courts, and not only in the narrow context of “changed circumstances”, as Judge Calabresi would postulate, but also in cases in which brand new legislation has been found defective by the court. As examples, discussed at greater length elsewhere in this book, consider the judgement of the Hungarian Constitutional Court in 1994, compelling the parliament to rewrite the “lustration” law,6 or in 1998, to repair the abortion law by specifying more clearly the criteria of “emergency circumstances” that might justify a pregnant woman’s demand for abortion.7 Furthermore, these demands for rewriting the law often come with relatively detailed guidance about how it should be done. After the 1993 law on local government in Slovenia was challenged, the Constitutional Court declared that a particular article which stipulated that referenda (constitutionally required for the
CONSTITUTIONAL COURTS AND LEGISLATION
89
establishment of new municipalities) should be conducted in some municipalities before local elections and in others only after such elections, was unconstitutional because, in the latter case, elections would take place before the people had determined by referendum the boundaries of the new municipalities.8 The decision caused a degree of consternation among some deputies as it declared explicitly which legislative solutions the Court would consider constitutional and which it would not. There have been even more overt assumptions of the role of “positive legislator” where certain courts have, at times, not simply told the parliaments how to improve the statutes concerned, but have actually rewritten them themselves. In 1994, the Constitutional Court of the Czech Republic annulled certain provisions of the 1991 Act on Extra-Judicial Rehabilitation, the act that allowed persons to make claims for the return of property that had passed into the hands of the state as a result of certain specified laws enacted during the communist era. The Court held that a requirement of permanent residence in the Czech Republic as a prerequisite for making such claims was contrary to the equality provisions of the Constitution, and actually gave the exiled citizens six months as from 1 November 1994 to claim their former property. The then Minister of Justice Jiri Novak criticised the ruling saying that the Constitutional Court should have only interpreted, not created, the law, and that setting of deadlines was the prerogative of Parliament. However the position of the Court prevailed, partly perhaps due to the fact that President Havel announced that he respected the decision. At times, a “dialogue” of sorts develops between the constitutional court and the parliament, in which the legislation moves back and forth between both institutions until certain equilibrium is attained. The Estonian lawmakers tried, in 1993, to use the law on housing co-operatives to help centralise property, in order to facilitate its subsequent privatisation. The legislation tried to force the former state-owned cooperatives to surrender apartments that they had built with their funds so that their tenants could privatise them under the country’s voucher system. The Constitutional Review Chamber (CRC) found a key provision unconstitutional (under the right to property), declaring that co-operative and collective organisations were entitled to control their property if they had contributed to its creation. This decision fundamentally changed the process of privatisation for these organisations, giving them new rights to retain or even to reclaim their property. The deputies drafted a new law in an attempt to ameliorate the situation: the new measure of December 1995 implicitly admitted that the apartments were going to be expropriated, but attempted to sweeten the deal for the co-operatives by allowing them to use the tenants’ privatisation vouchers for the purchase of the land that the tenants might otherwise possess. This compromise did little to satisfy the co-operatives, or the Legal Chancellor who lodged another appeal; the CRC, however, this time sided with the Parliament, arguing that the Constitution’s main requirement for the expropriation of private property had now been satisfied, in that (1) the goal of the law, to allow all tenants in Estonia equal access to housing privatisation, was clearly in the public interest, and (2) the Parliament’s new offer to the co-operatives and organisations amounted to ‘equitable and appropriate compensation’ for their losses. The appeal was thus dismissed, and the court helped to strike a compromise between the parties involved in the controversy.
90
CHAPTER 4
Many scholars studying West European constitutional courts have noted the indirect but nevertheless clear impact these courts make on legislation by making the legislators so aware of possible invalidations that they avoid enacting laws that may later be struck down. In anticipating the possible reaction by the court, legislators will attempt to draft the law accordingly. As Martin Shapiro has observed with regard to French judicial review: There is a substantial amount of evidence that … the French assembly is so mindful of the Constitutional Council’s pronouncements, and the Council so explicit in stating its future intentions, that French statutes are rigorously tested constitutionally before the Assembly approves them by what amounts to a constitutional dialogue between the Assembly and the Council.9
Although it is very hard to find evidence that such anticipatory behaviour has occurred in CEE legislature/court interactions as well, it is certainly the perception of some of the courts themselves that their jurisprudence affects the shape of new legislation in this manner. The then President of the Constitutional Tribunal of Poland, Professor Andrzej Zoll said in 1997: [T]he [Constitutional] Tribunal is treated [by the parliamentarians] with great respect and often I notice that our rulings are analysed by the deputies before they pass a law. They want to see if a pending statute will run afoul of any previous decisions of the Tribunal and if the Sejm will again run a risk of having its legal acts questioned.10
A similar opinion was expressed by the President of the Romanian Constitutional Court, Lucian Mihai; in his view, Romanian MPs increasingly take into account the fact that the Court may, in a preliminary review, strike down the bill. As an example, Judge Mihai draws attention to the draft law on local administration proposed by the government at the beginning of 2001. According to the draft, any mayor against whom any criminal charge is brought would have to be suspended, whereas according to the currently valid law such suspension only takes place if the official act of indictment by public prosecutor is filed. The opposition acted promptly and, as Judge Mihai reports, threatened that they would challenge the bill in the Constitutional Court if it were to be adopted (on the basis of inconsistency with presumption of innocence). As a result of this threat, in Judge Mihai’s opinion, the government modified its draft accordingly.11 But again, it is difficult to find hard evidence that this is a universal phenomenon in CEE, and no doubt the constitutional justices have an interest in playing up their role in influencing the pre-enactment legislative process. 2. DETERMINANTS OF THE “STRENGTH” OF JUDICIAL REVIEW To ascertain the relationship between the legislature and the constitutional court, we may talk about the “strength” of judicial review – by this, I mean the quantity and the importance of confrontations between the court and the lawmakers reflected in the number of successful referrals to the court (particularly parliamentary minority) and also in the relative significance of the laws invalidated for unconstitutionality. The strength has a quantitative dimension – a dimension of scope that refers to the range of issues within which the court’s preferences collide with those of the
CONSTITUTIONAL COURTS AND LEGISLATION
91
parliament – and a qualitative dimension, that is, the salience of the issues that produce disagreement between the two institutions.12 There are a great number of factors that determine both the range and the significance of the clash between the court and the parliament; obviously, the institutional rules of access to the constitutional court and the range of actors entitled to initiate the review are among the more important ones. Here, however, I am more interested in those factors determining the strength of judicial review that are anchored in the design and operation of the legislature itself. In general, one can claim that the strength of judicial review is, to a high degree, a function of the structure of the legislative system (namely, the more “veto points” there are in the lawmaking process, the weaker the clashes between the constitutional court and the legislature), and also of the nature of the party system as it operates in the parliament (namely, the greater the push for creating coalition governments rather than strong one-party majorities, the weaker the clashes between the constitutional court and the legislature.) As to the first factor, Alec Stone Sweet has convincingly shown that the strength of constitutional review is largely a function of how centralised the control by the government over the legislative process is.13 The more “veto points” a system contains (such as a second chamber, the availability of presidential veto, etc) the less dramatic is any disagreement between the constitutional court and the legislature likely to be. This is because “[w]here the court is only the second, or even the third, filter of legislative ambition, the opposition might well be able to achieve its objectives without making use of the abstract review referral….”.14 In a somewhat different context – that of a concrete review – Stephen Griffin makes the point that the fragmentation of the US political and legislative system, with in-built checks and balances, renders it more likely that the Supreme Court will find an ally in one of the branches of government and thus dilute the strength of its disagreement with the lawmakers in general: “even if there are important policy implications [of a Supreme Court decision], the Court can still escape damage if it is allied with political forces or one of the branches that objected to legislation. These allies can protect the Court by using the checks and balances the Constitution provides. . . . The effects of a fragmented political system filled with checks and balances thus work to the Court’s advantage . . . .”.15 The general argument is that the centralisation of the system of lawmaking – where one political actor can effectively control the process from the early legislative initiative up to the final enactment of the law – makes the constitutional court a more likely arena for resolving disputes concerning legislative preferences, and less “protected” against the effects of such a head-on clash. The evidence from CEE countries to support these hypotheses is very fragmentary, partly due to the fact that the force of variables related to the regime type is overshadowed by the force of variables related to the degree of consolidation of democracy in different CEE states, with the latter variable affecting the independent position of the constitutional court and thus the strength of judicial review. For example, the strength of the presidency only partly affects the existence of a “veto point” which may deflect the clash of legislature and constitutional court. Almost all consolidated democracies of CEE have opted for a parliamentary system with a relatively weak presidency. An exception is Poland, with a semi-presidential system, where the presidential veto power has indeed at times effectively released the
92
CHAPTER 4
Constitutional Tribunal from the need to consider (and, putatively, invalidate) controversial laws. However, not all of the powers of this relatively strong presidency are capable of weakening the legislature-constitutional court clash; indeed, the presidential power to initiate an ex-ante review (incidentally, also present in Hungary) may exacerbate the disagreements because the burden of invalidation then falls on the Constitutional Tribunal rather than on the President. The other Central European consolidated democracies assign a weaker role to the president within the parliamentary system than does Poland. In some cases even though the president may be elected directly by citizens, implying a relatively high degree of political legitimacy, for example in Slovenia, he nonetheless does not possess any power of veto over legislation approved by parliament. Nor in Hungary or Lithuania do the presidents enjoy such a power of veto (although Lithuania is often classified as an instance of semi-presidentialism, due to the fact that the president has reasonably strong powers in the field of international relations, dissolution of assembly and legislative initiative). The presidential powers of veto in Estonia, Romania (another state often characterised as semi-presidential) and Slovakia are more limited than in Poland because in these countries his suspensive veto can be overridden by a simple parliamentary majority rather than by a qualified majority of three-fifths as is the case in Poland (in Bulgaria and the Czech Republic an absolute majority is required); hence the capacity of presidents to act as “veto points” are more limited in these countries. Other than Poland, semi-presidential systems in the region, such as Russia, Ukraine, not to mention Belarus, also do not experience strong clashes between the constitutional courts and their respective executives/legislatures; here, however, this is not due to the “deflecting” effect of strong presidential powers but rather to the less consolidated nature of the democratic institutions in these countries. All of this illustrates that, with the exception of Poland and, to a lesser degree, the Czech Republic and Bulgaria, the institution of President does not really constitute a significant “veto point” affecting the strength of judicial review. Alec Stone Sweet has also shown that, in those West European bicameral systems where the upper chambers possess effective powers to stop the legislation favoured by the lower chamber (Germany and Italy), the most controversial pieces of legislation may be screened out by the upper house, thus reducing the potential for conflict between the constitutional court and the parliament.16 In two of the Central European bi-cameral systems (Poland and the Czech Republic) the powers of the upper chambers are more limited, as they possess only suspensive vetoes visà-vis legislation (although, unlike the Senates in France or Spain whose veto can be overridden by simple majority of the lower chambers, the Senate’s veto in Poland and the Czech Republic can be overridden only by an absolute majority thereof). In the third bicameral Central European system, in Romania, both chambers posses the same legislative powers; this, of course, renders the Senate much more significant and effective as a “veto point” than the Polish or Czech Senates are, especially due to the fact that, in practice, the two chambers often “compete with one another for the final word on particular pieces of legislation”.17 Nonetheless, even the merely suspensive veto of the Polish or Czech Senates may, at times, deflect the force of the clash between the lower chamber and the Constitutional Court. An example is
CONSTITUTIONAL COURTS AND LEGISLATION
93
provided by the Czech Republic in which the Senate (incidentally, existing now for reasons that, in the opinion of many observers, have largely lost their force)18 occasionally sides with the Constitutional Court when the latter collides with the lower chamber. As the President of the Senate explains, this is largely because “the Senators are less bound by the strict party discipline than the members of the lower chamber”,19 and also because the relatively long term of office of Senators (6 years) brings their perspective closer to that of the judges of the Constitutional Court.20 There have been therefore cases in which the Senate formed an “alliance” with the Court against the lower chamber. As an example, consider the law of 1998, which reduced the salaries of public servants by annulling the traditional “salary supplement” (the so-called “13th and 14th monthly pays”) for all public employees, as part of the government austerity measures. The Constitutional Court protested against the inclusion of judges’ salaries in these restrictions, on the basis of judicial independence, and also on the basis of the special role played by judges in the protection of democratic values and the rule of law.21 As a consequence, the Court rewrote the law by exempting judges from the universal pay reduction. We may put to one side the reasonable suspicion that the Court's decision was partly a selfserving exercise (after all, the judges on the Constitutional Court would themselves benefit from this particular decision!).22 The Court’s decision was, as could be expected, unpopular with the general public and with politicians, including President Havel who went as far as to publicly criticise the judgement.23 The following year, the Assembly of Deputies (the lower chamber) enacted the austerity law, which basically replicated the previous year’s measures, including the removal of pay supplements for judges. At this point, the Senate returned the law to the Assembly specifically on the basis that the parliament should respect the decision of the Constitutional Court. This example becomes all the more striking upon consideration of the fact that the 1998 law (subsequently questioned by the Constitutional Court) had originated in the Senate itself. The Assembly chose, however, to ignore both the position of the Senate and the earlier decision of the Court (which had applied to an earlier, although relevantly similar, law) and enacted the 1999 law in its original form. This new statute came again before the Constitutional Court, which, by this time, had changed its mind and found the law consistent with the Constitution. In this case of a parliament-constitutional court clash, it was the Court who eventually caved in; but not before discovering an ally in the upper chamber of the parliament, and realising that its objections had been endorsed by the upper chamber and presumably seriously considered by the lower. 3. CONSTITUTIONAL COURT AND THE PARLIAMENTARY MINORITY While, in general, the position of a constitutional court exercising a robust judicial review puts it in a potentially (and often, actually) antagonistic relationship vis-à-vis the legislature, this is of course only partly true: the real potential institutional hostility lies between the court and the parliamentary majority after a recently enacted law has been judicially invalidated. At the same time, such an invalidation creates grounds for an alliance between the court and the parliamentary minority,
94
CHAPTER 4
especially in cases where minority MPs have initiated the challenge to the law in question. Alliances of this type are a universal phenomenon whenever abstract review can be initiated by members of parliament, for the simple reason that “[o]ppositions judicialize legislative processes in order to win what they would otherwise lose in ‘normal’, unjudicialized processes (they would be outvoted)”.24 This is a cheap and (subject to some reservations, more on which below) risk-free avenue for achieving the political aims temporarily unattainable by the opposition through a legislative path. Stone Sweet hypothesises (with regard to West European constitutional courts) that the likelihood of clash between the court and the majority of legislature is higher soon after the alternation of power in the parliament (from the left to the right, or vice versa), because in such periods the court is likely to reflect the political views of the past but not the current majority.25 This hypothesis, however, is obviously only of limited applicability in those systems in which the selection of judges is done only partly by the parliament,26 or where the majority has to build a consensus with at least a part of the opposition to agree upon a judicial nomination.27 Nevertheless, events in CEE often seem to confirm Stone’s hypothesis. In Bulgaria, after the return of the postcommunists [the Bulgarian Socialist Party, or BSP] to power in December 1994, there was a distinctive rise in the ratio (and importance) of referrals to the Constitutional Court by the president (at the time, Zhelyu Zhelev) and the opposition. In this situation, where the presidential veto could easily be overridden by the parliament (in Bulgaria an absolute majority of parliament is required to do so, and the BSP held such a majority at the time) and, in fact, was so routinely (sometimes, even the day after the lodging of the veto by the president!),28 the Constitutional Court became the only countervailing power to the parliament dominated by the BSP. As Professor (and Constitutional Court judge) Todor Todorov later commented about this period: [T]he thrust of every one of these policies initiated by the BSP – the attempts to undermine judicial independence, to roll back the process of reintroducing private property, and to resubjugate the newly ‘independent’ media – inevitably set the parliamentary majority against the Court…. Neither the opposition’s parliamentary manoeuvres nor the president’s vetoes could pose a significant challenge to the new [BSP] majority….29
It has to be noted, however, that the Court displayed a high degree of restraint in invalidating the laws, probably in order not to further sharpen the political conflict.30 As another (former) Justice of the Bulgarian Constitutional Court observed, the opposition in Bulgaria developed a habit of threatening the majority that if a certain law were adopted, it would immediately be sent to the Court.31 The effectiveness of this “threat” is, naturally, a function of the composition of the Court; the same former justice observed that as “recently the Court has been connected with the parties in power”, the government and the majority “hope that the justices of the Constitutional Court will be loyal”.32 The relative composition of the court and the legislature is a variable that may affect the attractiveness of judicial review to the parliamentary opposition; however, as a general rule, it seems obvious that the opposition must treat the court as its natural ally, particularly when the legislative system includes relatively few “veto
CONSTITUTIONAL COURTS AND LEGISLATION
95
points” and internal “checks and balances”, and the political majority is reasonably disciplined and stable. This is the case, for instance, in Romania; and it is no wonder that the President of the Romanian Constitutional Court could proclaim: “[T]he [Constitutional] Court is an effective defender of parliamentary oppositions. All of the draft laws that we considered were brought before us by minorities, by minority factions”.33 Similarly, in Russia, a commentator could observe that the method of members of the Duma of challenging the constitutionality of a piece of legislation in the Court “is often used as a type of extra-parliamentary procedure, with losers on a certain issue bringing the case before the Court”.34 He noted, however, that “the Court has shown considerable restraint by refusing to enter the political fray, and has held fast to a strict interpretation of the law or statute in light of the Constitution”.35 This being said, whether or not the constitutional court can be indeed used as an ally of a parliamentary minority depends upon a number of factors; not only on the perceived likelihood of success in the court but also on the anticipated public response to a constitutional challenge. In Russia, under President Putin, the existence of a close cooperation between strongly pro-presidential Duma majority and a very popular President renders the very fact of a challenge by a weak and demoralised minority politically costly. This is the most likely explanation for the low number of minority-initiated constitutional challenges to adopted statutes under the Putin presidency. As one knowledgeable commentator has explained, “the President is very popular and so they [i.e. the parliamentary opposition] have to act carefully. If they object to a statute, it will be [seen as being] against a decision of the President, and they will not win any political points as a result [of such challenge]”.36 Another expert has noted that if the opposition appeals to the Constitutional Court and the Court then dismisses their challenge, then this will negatively “affect the legitimacy of the opposition”.37 If, as now, the President enjoys a high degree of popularity, he may even dissolve the Duma as a reprisal for such a challenge and “this poses a grave danger to the opposition”.38 He added that if a parliamentary opposition genuinely wishes to “discredit” a given law, it is more likely to go to the media rather to the Constitutional Court.39 This last example refutes the hypothesis that a parliamentary minority has an obvious, immanent incentive to challenge laws in the constitutional court whenever it is outvoted in the parliament. While the financial costs of such a challenge are usually negligible, the political costs may be high, and the maxim “there is no harm in trying” does not necessarily apply. This is also the case of some constitutional courts in Western Europe. In Portugal, for example, constitutional challenges to newly enacted laws have only very rarely been initiated by parliamentary minorities even though, from a formal point of view, such a challenge is exceptionally easy (one-tenth of the MPs can petition the Court to decide on constitutionality of a law in the process of an ex-post review). This is, as one observer has explained, due to the fact that the opposition parties did not want to be seen to be resorting to a counter-majoritarian (and therefore, presumptively non-democratic) mechanism: such action would weaken, in the eyes of the electorate, their claim to become a party of government at the next elections.40 (Interestingly, the only opposition party that made frequent use of the possibility of referral to Constitutional Court was the Communist Party).
96
CHAPTER 4
The Portuguese example clearly shows that the public perception of any challenge by an opposition party of a new law before the constitutional court depends on various factors related to political culture. In Portugal, such a challenge is seen as a sign of the weak attachment of the relevant party to the democratic rules of the game; whereas in Russia, a challenge is seen as a sign of “trouble-making”, as a symptom of political isolation, and becomes part of negative political baggage carried by a party. Elsewhere, the public interpretation of such a challenge may be different still: in Poland, for example, the party supporters usually expect that their party will challenge the controversial law in the Constitutional Tribunal whenever it has such an opportunity; such a challenge is not generally seen (even by opponents of the challenging party) as a violation of democratic rules of the game or of political weakness. Rather, it is seen as a perfectly legitimate political tool; indeed the failure to bring the law in question before the Constitutional Tribunal may be seen as a sign of the opposition party’s negligence, lack of skills or of political will to win. The tentative conclusion of this comparison is that the incentives for opposition parties to enlist the constitutional court as its political ally against the majority and the government depend upon the public perception of these parties and on the public interpretation of the meaning of the constitutional referral, as much as they do on the comparison of the (political) composition of the parliament with that of the constitutional court. 4. THE QUESTION OF JUDICIAL ACTIVISM AND RESTRAINT How “activist” are the most dynamic constitutional courts in CEE? Is the judicial activity of these courts significantly altering the preferences of the parliamentary majorities, and – more importantly – departing from the views of the constitution makers? Elsewhere (especially in the United States) a reliance on the “original intent” in the area of constitutional interpretation is largely – and deservedly – in disrepute, but when the constitution is brand new and the constitution makers are still very much around, the hostility to the very idea of the original intent is less understandable. After all, it is the constitution that provides the basis of the legitimacy of judgements by the constitutional court. It is difficult to establish what the criteria of “judicial activism” should be, and the concept itself is suspect to many legal scholars and practitioners,41 but we do not need to get embroiled in the controversy over the term. What is important is that the phenomenon that it is supposed to denote here is substantial and raises understandable concerns, namely, the substitution of the view of the parliamentary majority with that of the court’s majority concerning the proper articulation of the meaning of a constitutional right when these two views collide and when it is possible for the court, within the set of argumentative resources available to it, to uphold or to strike down the law. So this understanding of “activism” is not negated by the fact that, for example, courts do not normally control their own agenda, and that they cannot refuse to deal with the matters brought to their attention by politicians:42 “activism” is not a matter of an unwarranted intrusion into a field that the court has not been invited into, so to speak. As a working test, I suggest that an
CONSTITUTIONAL COURTS AND LEGISLATION
97
inquiry into the “judicial activism” of constitutional courts involves two criteria: the importance of the laws invalidated under the rights provisions and the nature of the reasoning leading to such invalidation. As far as the first criterion is concerned, the relative importance of a norm is admittedly in the eyes of beholder; thus, whether a rule that has been struck down is relatively significant or relatively trivial is a matter that cannot be ascertained in a non-controversial fashion. Take, for example, the Estonian Constitutional Review Chamber that struck down the Tallinn city regulations concerning the removal of illegally parked vehicles;43 some would probably say that this is a relatively trivial matter, in the broader scheme of things. Others might look at the decision more closely, and, having ascertained that the conclusion has been reached under an interpretation of the right to private property, will conclude that it posits a fundamental and potentially far-reaching principle of demarcating individual autonomy and the state’s police power. At the end of the day, what matters for the characterisation of a court as “activist” is not so much a proportion of the relatively “trivial” matters decided by it but rather the very fact that, even if very rarely, it has reversed some truly fundamental political choices on central public issues. And there is little reasonable disagreement that some of the Central European constitutional courts have, at least occasionally, displaced the parliament’s will on fundamental matters: important aspects of laws on abortion,44 the death penalty,45 “lustration” (the screening of officials suspected of improprieties under the auspices of the ancien régime),46 criminal prosecution of former communist officials responsible for crimes against the people during the communist period,47 economic austerity measures,48 fiscal policy,49 citizenship requirements,50 personal identification numbers for citizens,51 indexation of pensions,52 have all been struck down. This is a random list; a more compelling, but by no means comprehensive, account will be provided in chapters 6-10 which show the significance of the laws scrutinised – and not infrequently invalidated – by constitutional courts in the region. The importance of the laws overturned is only one part of the test to gauge the “activism” of the constitutional courts. Another, equally essential factor, is the nature of the reasoning that led to invalidation decisions. After all, if the courts are constitutionally mandated to verify the constitutionality or otherwise of statutes, they may have no choice but to overturn the laws that, on their face, clash with constitutional provisions. Of course, the mutual compatibility of two legal provisions (one of which is in the constitution) is always a matter of interpretation, and people may disagree in good faith over the correct articulation of their meaning: either the provision that is subject to the constitutional challenge, or the one that forms the basis for a possible invalidation decision. Rather than getting embroiled in a theoretical discussion about what renders judicial reasoning “activist” (and whether such a characterisation makes sense at all), I will give some examples of certain patterns of reasoning characteristic of constitutional courts in the region discussed here, and appeal to certain intuitive, commonsensical and relatively uncontroversial views about how these patterns are symptomatic of judicial restraint or judicial activism of the courts in question. I will, further, discuss in some detail two important decisions on the death penalty and on
98
CHAPTER 4
abortion, from Hungary and Poland respectively, in order further illustrate the patterns of judicial reasoning of the constitutional courts in these countries. In all fairness, one should note that the rhetoric of judicial restraint is usually present in the case law of constitutional courts – and often the rhetoric is adhered to in the actual structure of argument. The courts, when engaging in judicial review, often emphasise the presumption of constitutionality of statutes. For example, the Polish Constitutional Tribunal, in a 1997 decision on collective agreements in the workplace, noted: The burden of argument is on those who challenge the constitutionality of a law and unless they produce a specific and convincing legal argument to prove their points, the Constitutional Tribunal will recognise the laws under challenge as constitutional.53
The presumption articulated in this sentence forms a fixed point in the selfperception of many constitutional judges in CEE countries. A justice of the Bulgarian Constitutional Court has stated that, in his court, the presumption of constitutionality is a “tacit assumption” that forms a common position of all the justices: the law will be declared unconstitutional only when justices “are sure about it”; if they have any doubts, they adopt an “in dubio pro reo” principle and uphold the law.54 Consistently with this presumption, the courts occasionally act upon the principle that if a statute allows for different interpretations, the one that is most consistent with the constitution has to be chosen.55 For example, the Czech Constitutional Court established it as a legal principle that when the Court is faced with a choice between interpreting a statute in conformity with the Constitution (even if there is also an interpretation available that is contrary to the Constitution) or striking it down, it should choose the first path, and regular courts should then follow this “constitutionally conforming” interpretation.56 The interpretation by the Constitutional Court should then be binding: otherwise, as Pavel Holländer has argued, the Court will have an incentive to abandon judicial restraint and overturn the statute, even in cases in which it could have upheld it subject to a binding interpretation.57 An example from Czech Republic illustrates this technique:58 the cumulative application of two criminal statutes under a “plain meaning” interpretation would have allowed for a suspect to be held in detention for a total of 48 hours even although the maximum possible period of such detention provided for by the Constitution was 24 hours. An ordinary court, faced with this situation, interpreted the statutes in the “normal” way, allowing for 48 hours. A constitutional complaint was then brought before the Constitutional Court which, in turn, decided that the ordinary courts should have interpreted the statutes in a manner that conformed to the Constitution. There was also an interesting twist to this case: after this decision of the Constitutional Court, the Supreme Court lodged a challenge (in the process of concrete review) to one of these statutory provisions. The Constitutional Court accepted this referral and did indeed annul the statutory provision. This idea of “interprétation conforme” – a doctrine adopted in a number of CEE systems, as well as in their Western European counterparts – is generally seen as an example of deference by the constitutional court, which may then portray itself as doing its best to avoid outright annulment of a statute.59 In a similar fashion, the
CONSTITUTIONAL COURTS AND LEGISLATION
99
courts often acknowledge that legislators enjoy a wide scope for legitimate discretion. The Hungarian Constitutional Court, in its 1991 decision on abortion, stated that “[w]here the law should draw the line between the unconstitutional extremes of total prohibition and unrestricted availability of abortions is for the legislature to decide”.60 The constitutional courts also like to declare that, within the domain of legislative discretion, it is a political rather than constitutional responsibility that controls the legislator.61 They also characterise their own role as, at best, a “negative legislator” rather than a positive one (repeating a well known Kelsenian formulation of the role of a constitutional court)62 and emphasise that they should attempt to “minimise” their “intrusion into the decision-making powers of other state bodies”.63 Most of all, they never tire of reminding their audience that the grounds of their decisions are not “political” but “strictly constitutional”, implying that the political or moral preferences of the judges do not enter into the process of review.64 These declarations should not, however, always be taken at face value, as they often remain just that: declarations. In terms of the substantive basis of decisions, the distinction between the “political” and the “constitutional” is rarely tenable. When, for example, the Polish Constitutional Tribunal considered in 1997 certain collective-bargaining rules, sure enough it reasserted its role as limited to the scrutiny of constitutionality rather than of the substance of the rules: The Constitutional Tribunal wishes to assert once again that it is not its task to decide on the substantive wisdom of legislative solutions. It is up to the legislator to enact laws that will respond to the political and economic tasks adopted, and to find such legal solutions that will, according to the legislator, best serve the attainment of these goals. . . . It is only when the legislator exceeds these limits of discretion and breaches a specific constitutional rule, principle or value that an intervention by the C[onstitutional] T[ribunal] is permissible.65
The reference to a “constitutional value”, alongside a constitutional rule and principle, immediately shows just how illusory is the principle of restraint proclaimed in the first part of the passage just quoted. Although in this particular case the challenge was eventually unsuccessful, it has indeed become an established doctrine of the Constitutional Tribunal that statutes are controlled by reference to “constitutional rules, principles and values”,66 thus providing the Tribunal with a vague and broad standard in terms of which they may analyse specific statutes. More generally, the declarations of restraint should be treated with a dose of scepticism, not necessarily because they are deliberately deceptive – this certainly is not claimed here – but rather because the intentions expressed in them are rarely capable of being tested against the reality of abstract judicial review, especially when conducted in terms of constitutional rights. How reliable is the principle of deference expressed in the presumption of constitutionality? The presumption itself does not, and cannot, tell us what it takes to override it; it itself is untestable. The presumption of constitutionality tells us, at best, that the judge will strike down the law only if she hears good arguments against it: unless such good arguments are provided, the law will stand. The burden is thus on those who argue against the law, not on those who defend it. This formula, however, tells us nothing (nor can it) about what arguments will be accepted as good ones by the judge. There is no
100
CHAPTER 4
necessary connection between the adoption of a presumption and the ease with which this presumption may be rebutted, and the easier the rebuttal, the weaker the restraint. Consider the analogy to Thayer’s famous formula for judicial self-restraint in the United States: that a court can strike down a legislative enactment not “merely because it is concluded that upon a just and true construction the law is unconstitutional” but only when a legislature has made a “very clear” mistake, “so clear that it is not open to rational question”.67 However, what is to count as a “very clear” mistake is a matter that the formula must leave open, and whether a judge has sincerely established that the statutory measure meets this condition is ultimately untestable. This is particularly evident in cases of scrutiny of legislation in terms of constitutional rights when the challenge to a law is often based on a fundamental moral disagreement: anyone who disagrees with the moral choice made by the legislator in articulating the constitutional right will naturally tend to deem this choice a “very clear mistake”. Within the presumption of constitutionality, a judge having a fundamental moral disagreement with the legislative choice may find it excessively easy to rule the presumption of constitutionality overridden. In any event, the CEE courts have not always maintained even the rhetoric of judicial restraint, especially when their articulations of constitutional rights were at stake. In answering a question concerning charges of “activism” of the Hungarian Constitutional Court, the then Chief Justice of the Court, László Sólyom, responded: “Activism can mean different things. We always stress that we are activists in certain areas, namely, concerning fundamental rights, where the Court does not hesitate to decide ‘hard cases’”.68 There have been indeed some important decisions of the constitutional courts that are unmistakably “activist” – in the sense that the court had the option of upholding the statute, within the recognised conventions of judicial reasoning, and yet decided to overturn it. If a set of recognised conventions of judicial reasoning makes it possible for the court to uphold the statute in question, but also makes it possible to overturn it, a tendency to choose the latter path may be seen as an indicium of “activism”. This happens when, for example, a court grounds its decision in very abstract, general and vague constitutional notions, the specific articulation of which reasonable people may disagree over, even though it had the option of founding its decision on narrower and less ambiguous grounds. The idea of “human dignity”, used as a sufficient basis for overturning specific statutes, is a good example. The Hungarian Court has appealed to this notion in a number of different contexts. For example, in a 1990 decision, the Court proscribed trade unions from representing employees without their consent; it relied on “human dignity” and established the following general principle: “When none of the … named fundamental rights are applicable for a given state of affairs” then the “general personal right [to dignity] … may be relied upon at any time by the Constitutional Court”.69 This decision is all the more remarkable given that, as one (friendly) commentator of the Court has noted, rather than relying upon the right to dignity, the Court could have easily held that the constitutional clauses that specifically relate to the rights of unions “to safeguard and represent the interest of employees” only authorise representation with consent.70 Arguably, the most telling example of the use of vague and ambiguous notions in order to overturn a clear legislative and constitutional intention is provided by the
CONSTITUTIONAL COURTS AND LEGISLATION
101
same Court’s decision invalidating the death penalty.71 To reach this – laudable, from the point of view of this author – result, the Hungarian Court had to face the problem that the Constitution contained a stipulation that “no one may be arbitrarily deprived of life and human dignity” (art. 54 (1)). Although silent on the specific issue of the death penalty, this provision clearly implied that a “non-arbitrary” deprivation of life was constitutionally permissible. A judge faced with this textual implication, who is nonetheless intent on striking down the death penalty as unconstitutional, can theoretically reach this result in one of two ways, neither of which is quite satisfactory. She may either (a) claim that death penalty is necessarily arbitrary and thus prohibited under art. 54 (1), or (b) find another constitutional provision that would constitute a basis for invalidating the death penalty, and give precedence to that other provision over art. 54 (1). In the decision of the majority in the Hungarian example, the latter path was taken. The Court relied on art. 8 (2), which proclaims that “rules on fundamental rights and obligations shall be determined by laws which, however, shall not impose any limitations on the essential contents of fundamental rights”, in connection with the right to life and dignity. The Court ruled that the death penalty necessarily infringes the “essential content” of the right to life, and thus art 54 (1), to the extent that it may be read as permitting a nonarbitrary deprivation of life, is superseded by art. 8 (2). Of course, however, whether a non-arbitrary enforcement of the death penalty violates an “essential” aspect of the human right to dignity is a moral proposition about which reasonable people may, and do, disagree. The Court opted therefore for a controversial moral judgement over a more precise and narrow constitutional construction. Interestingly, in his concurring opinion, Chief Justice László Sólyom used both strategies; it is, perhaps, no wonder that applying both of them creates a clear impression of overkill (a mot juste in this context). As to strategy (a), he argued that capital punishment necessarily intrudes upon an “essential” area of life and dignity because they constitute “an absolute value” and form an “indivisible and unrestrainable right”. However, if this were true then any law that adversely affected life and dignity, however marginally, would have to be invalidated! As to strategy (b), he argued that capital punishment is necessarily “arbitrary”, not in any empirical, sociological sense (for example, because it fails achieve its purported aims) but rather “conceptually”. According to him, “capital punishment is arbitrary not because it limits the essential content of the right to life but because the right to life and dignity – due to the characteristics [of the two components] – is from the outset unlimitable”.72 This is a strange statement, and can hardly be read in any way other than as an expression of strong moral disapproval of the death penalty. Such moral disapproval is plausible and resonates with many people’s feelings – but as it happened, not with those of the majority of Hungarians, nor with the majority of the Hungarian MPs acting both in its law-making and constitution-making capacity. The reason why their moral judgements, as expressed in legal practices, should be replaced by those of the Court is the real issue that should have been addressed in this decision, but was not. As another example, consider a momentous 1997 decision of the Polish Constitutional Tribunal on abortion law.73 The Tribunal struck down as unconstitutional some liberal aspects of the then Polish abortion law, basically finding any
102
CHAPTER 4
abortions other than those justified on strictly defined medical grounds (because of threat to mother’s health, or the genetic defects of the foetus) or resulting from rape as contrary to the Constitution. The decision was all the more remarkable as it ran contrary not only to the majority opinion of the legislators (the Parliament was dominated by the centre-left coalition at the time) but also to the clear implications of the text of the Constitution. At the time that the decision was handed down, a socalled Little Constitution (an interim constitutional document, virtually free of constitutional rights) was in force, and it contained no reference to the “right to life”, much less a right to life from the moment of conception. More importantly, a new, fully-fledged Constitution had already been adopted, and accepted in a national referendum, and was to enter into force as from October 1997.74 Although, formally speaking, the new Constitution was not binding on the judges of Constitutional Tribunal, it provided a good insight into the views of the constitution-makers. The new Constitution did make a reference to a right to life but, importantly, demands to include the proviso “from the moment of conception” had been considered and expressly rejected by the drafters. At the very least, constitutional judges knew that the constitution-makers had chosen not to prohibit abortion. Faced with these textual constraints, the Tribunal nevertheless proceeded to argue that the availability of abortion on grounds other than those of (1) the danger to the life or health of a pregnant woman, (2) genetic defects of the foetus, or (3) the pregnancy being a consequence of rape, are contrary to the Little Constitution. In the absence of any reference to the right to life in that interim document, the Tribunal decided to base its conclusion, somewhat improbably, on the interpretation of the concept of the democratic “state based on law”, or Rechtsstaat, proclaimed in the first article of the Constitution. The centrepiece of the reasoning of the majority (and one should add that the decision was accompanied by three strongly-worded dissenting opinions) was that the Rechtsstaat presupposes a community of people, and that the essential attribute of individuals is their life, which has to be constitutionally protected “at each stage of its development”. While the “value” of life is not subject to degree as a function of different stages of its development, the intensity of the protection can be varied, depending on the conflict of this value with other constitutional values and interests, according to the Tribunal. The connection between this principle and the line drawn between some kinds of abortion (such as abortion necessitated by the health of a pregnant woman, which are permissible) and other kinds (abortions because of the “hard life conditions or difficult personal situation” of a pregnant woman) rests on a value judgement that could not have been inferred from the Constitution itself. The Court decided to ignore the ‘authentic interpretation’ of the meaning of the Constitution, provided by the very recent process of the drafting of a new Constitution, and established as law its own judgement on a very controversial matter. A critique by legal philosopher Jan WoleĔski is apposite here: as he points out, general clauses such as “democratic law-based state” have an important role to play in the interpretation of specific legal texts but their role in “generating new rules” is “a very delicate matter”. 75 WoleĔski further observes that there is no clear connection between the general Rechtsstaat principle and “any specific conclusions concerning the protection of life before birth”;76 furthermore, the principle of the “state based on law” may be seen as to
CONSTITUTIONAL COURTS AND LEGISLATION
103
support restraint in lawmaking and legal interpretation by courts regarding matters as controversial as abortion.77 He concludes that “the Tribunal has elevated its own, rather unclearly articulated axiology to the level of a legal principle. I do not think that this is within its legally defined competences”.78 Another symptom of the activism of some of the Central European constitutional courts is their predilection for “balancing jurisprudence”, especially of the least deferential kind; namely, for assessing whether the legislative measures in question are necessary to attain the approved legislative purpose. As is known, the use of balancing “is transforming constitutional discourse into a general discussion of the reasonableness of governmental conduct”, and is therefore the kind of reasoning that situates the court in a characteristically legislative role.79 Whenever the conflicting constitutional values (rights, principles, etc) are balanced against each other, the judgments as to the reasonableness or necessity of restricting one in order to implement the other have the effect of transforming the argument over the legality of a given measure into an argument about its wisdom, “with the consequent risk that the Court might encroach on choices that are, in principle, reserved to the legislator”.80 I will explore the main patterns of balancing jurisprudence in Chapter 10, when discussing the use of proportionality to elucidate the statutory limits of constitutional rights. At this point, however, an example of an activist decision based on non-deferential balancing may be helpful. A 1997 decision of the Slovenian Constitutional Court concerned a proposed referendum on amendments to the law on the re-privatisation of real estate, mainly agricultural lands and forests.81 The amendments, proposed by three parties, and which obtained the support of over 50 thousand voters in the referendum, were meant to water down the re-privatisation law of 1991 by (among other things) introducing a ceiling on the size of the lands returned to their former owners, and also by banning the return of land “of feudal origins”. The Court, exercising its power of review over referendum questions, largely disarmed the proponents of the referendum. It struck down the central question, aimed at introducing the limit of 100 hectares of land or forests by refuting, on allegedly empirical grounds, the rationale provided by the proponents of the referendum. This was that the return of very large areas was not within the capacities of the state, and would hinder the return of smaller pieces of land. This is clearly the type of judgment based on a cost/benefit calculus that is characteristically the domain of the political branches of the state, not the court. As to the prohibition of the return of lands “of feudal origins”, the Court inserted its own proviso that this prohibition must not apply to land previously owned by churches and other religious institutions. Contrary to the intentions of the authors of the referendum, the Constitutional Court argued that “it would not be constitutionally permissible to equate the nationalised property of the church and religious communities in view of their role as institutions of general benefit and their position in the Slovenian legal system, with estates of feudal origin”.82 In other words, a preferential exemption for the churches has been carved out by the Constitutional Court on the grounds of a positive assessment of the social role of the church. This assessment by the Court pre-empted a judgment by the general public on whether the prohibition of return of feudal property should apply also to religious institutions. This judgment made use
104
CHAPTER 4
of the “strict scrutiny” style of reasoning (although this concept was not used explicitly by the Court), namely, on the basis that the “proposed measures must be un-avoidable in a democratic society, dictated by urgent public need [and that] the . . . measures . . . must, in compliance with the principle of proportionality, be appropriate and unavoidable in order to reach the legislator’s objectives….”83 If the standard of scrutiny of a statutory regulation is whether or not the measures adopted by legislators (or, as the last case illustrates, the measures contemplated by referendum questions) are “unavoidable” and “necessary” if the approved goals are to be achieved, the pattern of reasoning of the judges becomes virtually the same as that of legislators. Such a standard also exhibits a high degree of distrust in the judgement of the legislature and reduces the likelihood of the regulation in question being affirmed as constitutionally valid. It is almost always possible to establish the availability of a different measure to the one adopted by the legislature, and, if such a demonstration is sufficient to defeat the legislation, no trace of deference to the preferences of the parliamentary majority can be found in the court’s approach. 5. CONCLUSIONS The design of constitutional courts in CEE, their aspirations, the high prestige that they quickly acquired and the relative weakness of the largely discredited political branches of government – mainly, the parliaments – all created an environment in which the constitutional courts gained a significant role in law-making. Nor was this role confined to striking down provisions that conflicted with the value choices shared by the majority of judges, as opposed to the majority of MPs; it also included “putting the parliaments on notice” that they should change specific laws, indicating the directions of these changes, and even, at times, “rewriting” the laws themselves. This, naturally, placed these courts on a collision course with their respective legislatures, or rather, with the political majorities, as parliamentary minorities often found the courts to be useful allies in their struggle to overturn the laws on which they were outvoted. Both the strength of the clash and the nature of the alliance between the courts and political oppositions largely depended on specific local factors: for example, the strength of court-legislature conflict apparently decreases with the fragmentation of the law-making process within the parliamentary system (e.g. in bi-cameral parliaments) and with a presidential power of veto over legislation. The tendency to establish alliances between the majority of the court and the parliamentary opposition, on the other hand, depends on the relative costs incurred by the opposition in challenging (or, as the case may be, in abstaining from challenging) the law in question before the court: an act that, in theory, should be almost cost-free may in fact lower rather than elevate the political position of the minority parliamentary parties. These institutional constellations moved the most dynamic and powerful of the courts in the region – especially those in Hungary, Poland, the Czech Republic and Slovenia – in the direction of a marked judicial activism, understood as a willingness to strike down important laws even if, under the available conventions of judicial reasoning, a decision upholding the provisions in question was an option genuinely
CONSTITUTIONAL COURTS AND LEGISLATION
105
available to the courts. This became clear not simply because of the significance of some laws being struck down by the constitutional courts in the region but also due to the type of arguments frequently used: appeals to very vague, indeterminate constitutional ideals and values (such as human dignity or Rechtsstaat), and a proclivity for constitutional balancing in which controversial judgements of policy figure prominently.
PART II
CHAPTER 5 JUDICIAL REVIEW AND PROTECTION OF CONSTITUTIONAL RIGHTS
As already stated in Chapter 2, the dominant justification for the robust position of the constitutional courts of post-communist states in CEE is based on the role of those courts in the protection of individual rights – in particular, those explicitly entrenched in the respective constitutions. It has been accepted, virtually without critical scrutiny, that constitutional courts must have strong powers to monitor the constitutionality of legislation if constitutional rights are to be meaningful. It is now time to consider the plausibility of this general claim in more detail. The claim, of course, is not specific to CEE; in fact, it is by far the most popular argument used to support the existence of, or demand for, strong constitutional courts. The argument – presented here in a deliberately simplified form – goes usually like this: democracy is not based on blind respect for unrestrained majority will, and individual and minority rights are among the most important constraints upon the majority. The political majority is capable of looking after its own interests, but can we be sure that it will give sufficient protection to the minority and individual dissidents, whether the dissidence is understood in political, moral, religious or personal lifestyle terms? The majority should not be allowed to always prevail over those who disagree with its preferences and choices, and the values reflected in constitutional rights – the argument goes – reflect this “precommitment” regarding the outer borders of the majority’s reach. As the majority cannot be trusted with observing predetermined limits upon its powers, an independent, non-majoritarian institution is needed to police, monitor and enforce those limits. The specific institutional design of such a body may vary from country to country but the general principle remains the same: if perfecting “constitutional democracy” is our aim, then the parliament corresponds to the noun, and the judicial (or quasi-judicial) body to the adjective in that term. This is a familiar argument, and the counter-argument is also familiar. It is said that countries without judicial constitutional review (e.g., the Netherlands, Sweden, the United Kingdom) protect rights as well as, and sometimes perhaps better than those with strongly entrenched bills of rights that allow judges to strike down unconstitutional statutes. Naturally, any such comparison runs the obvious risk of drawing illegitimate conclusions from a seemingly evident observation. The conclusions would be legitimate if all other things were equal, and if the only variable was the presence or absence of judicial review. However, the condition of ceteris
108
CHAPTER 5
paribus is never satisfied in comparisons between different countries: “everything else” is never equal, and those other unequal factors can significantly affect the level of rights protection. It might be said, for instance, that there are some factors – such as the nature of the political system, tradition, political culture, societal attitudes – in the Netherlands, Sweden, Australia or the United Kingdom that compensate for the absence of rights-based judicial review. Furthermore, if they had robust judicial review, perhaps the protection of rights in these countries would be even better. Or so those who favour robust judicial review could claim. The purpose of this chapter is not to provide a conclusive and universally valid answer about the actual role of judicial review in the protection of individual rights. Rather, it will aim to reflect upon what parameters must be taken into account to render the argument, or the counter-argument, regarding the role of judicial review in the protection of constitutional rights valid. It is also important, at the outset, to clarify the relationship between the defence of judicial review on the basis of rights protection, on one hand, and the defence of the democratic legitimacy of constitutional courts on the other. Theoretical and constitutional discussions about judicial review often merge these two dimensions. They usually focus on the question of legitimacy: is it legitimate for non-elected judges (who are therefore not accountable through normal representative-democratic channels) to frustrate the will of the democratically accountable representatives of the people and overturn their legislative choices? If a country’s constitutional text does not explicitly provide for such a judicial role (as in the United States), a court that has “usurped” such power can be (and often has been) directly confronted with the concerns of its critics. If the constitution does provide for judicial review of the constitutionality of statutes (as in most European countries) these concerns are addressed to the constitution’s authors, who were responsible for making this institutional choice. Either way, the question is distinct from that of whether judicial review does or does not help to protect individual rights. We might consider the operation of judicial review to be illegitimate and yet rights-protective. Or, vice versa, we might believe it to be legitimate but detrimental from the point of view of rights. These are conceptually and politically two distinct questions, although they can be merged in a trivial way. A trivial connection between legitimacy and rights protection would be established by saying that an illegitimate system of judicial review necessarily tramples upon individual rights; that is, it conflicts with our right to have coercive public decisions made only by bodies with the democratically legitimate authority to do so. But the word “necessarily” indicates that this is a trivial proposition; no meaningful conclusion about the overall quality of rights protection in a hypothetical, illegitimate system of judicial review follows from it. In this chapter, I will therefore resist any temptation to merge the question of the protection of rights with the question of the legitimacy of judicial review, and focus on the former only. Of course, a strong conclusion that constitutional courts are necessary, or even only significant, guarantees of protection of individual rights goes a long way towards supporting the need for constitutional courts, all things considered. In my view, however, nothing is gained by insisting that such a “need” is eo ipso a proof of these courts’ legitimacy. I will therefore explicitly narrow my discussion down to the issue of the role of these courts in promoting constitutional
PROTECTION OF CONSITUTIONAL RIGHTS
109
rights. More specifically, I will attempt to identify the sort of facts and data that we need to know in order to make an informed judgement about that role; in this sense, this chapter may be seen as providing a framework within which the successive chapters – on the role of CEE constitutional courts in articulating various types of constitutional rights – can be located. 1. TWO THEORIES ABOUT JUDICIAL REVIEW Theoretical conceptions of the role of judicial review in the protection of individual rights range from an enthusiastically positive response to an unqualifiedly negative one. These two extreme positions are best exemplified by the theories of Ronald Dworkin and Jeremy Waldron, respectively. For all the substantive – and fundamental – differences between the two theories, one common aspect renders them equally unsatisfactory: both are relatively insensitive to the facts of the protection of individual rights in existing systems of judicial review. It is not that these two legal philosophers disregard the reality of the legal systems – mainly the United States and the United Kingdom – that they use as their starting points (or as their sources of illustration). On the contrary, both scholars are deeply immersed in the life of these legal systems, and both certainly care about the substantive decisions that the relevant supreme courts and legislatures produce. However, their conclusions about the effect of a system of judicial review on the protection of individual rights in general are not greatly influenced by the reality of these patterns of decisions. The strength of these two theories is therefore independent of empirical facts about the effects of judicial review. Ronald Dworkin enthusiastically endorses the active and powerful role of the US Supreme Court, and advocates the establishment of a similar system in the United Kingdom.1 An empirical problem would thus seem to be raised by those decisions of the Court that are deplorable in terms of rights protection, and also by the instances where the Court was inactive (when it could have acted) towards statutes that had an adverse effect upon individual rights. These decisions should compel an advocate of a “moral reading of the Constitution” to reassess the value of judicial review; after all, Dworkin himself stresses that the “moral reading” thesis is about what the Constitution itself means, and not about whose views on the meaning of the constitution should be binding.2 He has been an energetic critic of many recent (and historical) United States Supreme Court decisions.3 Yet it is hard not to notice that Dworkin, as a critic of this or that decision, has little in common with Dworkin qua a theorist of judicial review. Precisely because Dworkin’s normative vision is so clearly divorced from his concern with specific effects, he can proudly, but implausibly, proclaim: “For two centuries American judges have ruled both national and state legislation invalid because it invaded the rights of freedom of speech or religion or of the due process of law or of the equal protection of law that the United States Constitution recognizes”.4 His normative ideal does not seem to have been affected by the fact that the Supreme Court only began to take the First Amendment seriously well into its second century of operation, and that its record in this area – as in other areas listed
110
CHAPTER 5
by Dworkin, including the equal protection sphere – has been mixed, to put it mildly.5 Nor does Dworkin seem disturbed by cases implicating important issues of rights in which the legislature was more rights-protective than the Supreme Court, such as when rights-enhancing legislative measures were invalidated by the Court, or when the legislature enhanced rights that had been affected by restrictive decisions of the Court.6 A general theory of judicial review should be tested against the actual outcomes that a review produces. The achievements during the liberal Warren and Burger eras must be compared to the losses (from a liberal perspective) produced by the Rehnquist Court, or even the Taney Court.7 It is likely that the overall balance will be positive, but this needs to be shown. Consider the case of affirmative action. A liberal who believes (as Dworkin does) in the rightness of raceconscious remedial affirmative action could be excused for thinking that this right would have been better protected if judicial review had not been exercised by federal courts in the United States, and if recently invalidated laws or policies had been allowed to stand.8 A critic of specific decisions who nevertheless defends the Court’s strong role towards legislation therefore has prima facie reasons for embarrassment. This can be avoided only by showing that there is no conflict between the criticism and the defence, which can be done in two ways. One way is simply to state that, on balance, individuals’ rights are better protected under a system of judicial review because the sheer number and significance of rights-protective judicial decisions greatly outweighs the number and significance of decisions that weaken legislatively conferred rights.9 This is notoriously difficult to prove in abstracto, and some would simply disagree with the outcome of such an equation, at least with respect to the United States.10 The other option is to dismiss the “wrong” decisions as aberrations, as cases of system failure that are unavoidable in any human institution. This alternative precludes the need to count and weigh a particular decision’s impact on rights, but it makes Dworkin’s thesis unfalsifiable and hence unverifiable. It acquires an internal self-validation quality and thus becomes immune to confrontation with the reality of the practice of judicial review in a given legal system. This is what I mean by its (relative) insensitivity to facts. Dworkin’s insensitivity to facts is all the more puzzling because he openly endorses a result-driven test for judging institutional design: “The best institutional structure is the one best calculated to produce the best answers to the essentially moral questions of what the democratic conditions actually are, and to secure stable compliance with those conditions”.11 This seems to be a sensible beginning for a fact-sensitive balancing of good and bad (that is, rights-protective and rightslimiting) decisions. But Dworkin draws a conclusion from the result-driven test that sounds like a non sequitur. He asserts that established constitutional practice actually reveals that US judges have a strong authority to conduct constitutional review: “If the most straightforward interpretation of American constitutional practice shows that our judges have final interpretive authority ... we have no reason to resist that reading and to strain for one that seems more congenial to a majoritarian philosophy”.12 This sounds like conservatism pure and simple: the fact of a given practice validates its value and demands that it be preserved. The question, however, is not
PROTECTION OF CONSITUTIONAL RIGHTS
111
whether these judges do have the final authoritative power over constitutional meanings, but whether – in light of a result-driven test – they should keep exercising such a power. The fact of authority and compliance is not a substitute for a critical review of an established practice. And while the practice is certainly established, it is far from being uncontroversial. This is something that Dworkin admits elsewhere when he refers to “the contemporary debate among American constitutional lawyers about the legitimacy of the United States Supreme Court’s power to overrule the decisions of elected legislators”; it is, he says, a debate “dominated” by “unspoken assumptions” about the centrality of majority rule in a democracy.13 Jeremy Waldron’s right-based criticism of judicially enforceable bills of rights exemplifies the opposite pole in the controversy surrounding judicial review.14 Waldron relies upon the point that the judicial reversal of democratically adopted laws denies an important individual right, namely, a right to democratic selfdetermination. As he claims: “this arrogation of judicial authority, this disabling of representative institutions ... should be frowned upon by any rights-based theory that stresses the importance of democratic participation on matters of principle by ordinary men and women”.15 Waldron’s argument is unimpeachable as far as democratic legitimacy is concerned: there is a chronic legitimacy deficit in any system that allows democratically unaccountable judges to displace choices made by a democratically elected legislature. Even if the system of democratic representation and accountability is defective, the defects can hardly be remedied by establishing an even less democratically accountable body.16 However, as indicated earlier, the question of legitimacy is conceptually and politically distinct from that of the effect of judicial review upon the protection of rights. And while Waldron’s argument serves as a powerful objection to the legitimacy of constitutional judicial review, it is not conclusive as an argument that judicial review adversely affects the protection of constitutional rights. At best, it adds into the equation a particular type of right that seems to be infringed when judges reverse statutes; namely, the right to democratic self-government through electoral representation. This is an important right, but not the only one, and perhaps not even the most important one. Thus, Waldron’s argument is incomplete in the context of the protection of rights: the loss of self-government rights in a specific legal system might be more than compensated for by the superior judicial protection of other rights that had been disregarded by the legislature. As Stephen Griffin has observed: Deciding to place the protection of basic rights in the hands of the judiciary is also a decision to remove such issues from the agenda of the elected branches. This restricts the basic right of citizens to participate in important political decisions respecting the content of such rights. While this consideration is by no means decisive, it provides a salutary reminder that the decision to adopt judicial review involves restricting some basic rights in order to promote others. This immediately raises the question of whether the rights to be promoted are of greater importance than the political rights that are restricted. 17
Obviously such a calculus cannot be made in abstraction but only by reference to a particular legal system. Waldron’s argument serves as a reminder that the judicially produced loss of self-government rights has to be included in our
112
CHAPTER 5
calculation, but it does not determine the result in the final weighing and balancing of different rights. Furthermore, it is not necessarily true that the right to democratic selfdetermination must be defeated by establishing a system of robust judicial review. Such an effect is only inevitable if a violation of this right is, by definition, understood to occur whenever the legislative majority’s view is not final. This is what Waldron seems to have in mind when he says that “if the process is nondemocratic, it inherently and necessarily does an injustice, in its operation, to the participatory aspirations of the ordinary citizens ... whether it comes up with the correct result or not”.18 In this case, however, the connection between judicial review and a violation of the right to democratic self-determination is trivial. And yet, it is always open to us to question the operation of the system of representation in a given country. It may be possible to show that the preferences of the majority undergo such radical distortions in the political process that a right to be accurately represented by one’s parliamentarians is not actually respected in a given society. This might be due to particularities of the electoral system, the influence of wealth on the process of representation, a biased media, self-interest and myopia of the representatives, and a great number of other political and social factors. In such a society, judicial review might offer individuals a better way of producing results that correspond to the majority’s actual preferences in the legal system. And if that is what the requirement of democratic self-determination is fundamentally about, under some factual circumstances it may better served by judicial review than by the democratically accountable institution. Suppose that a great majority of people favours strict gun control, or wishes to allow doctor-assisted suicide under some conditions, but that the elected politicians systematically oppose these legal measures due to pressure from powerful (though minority) interest groups.19 Recourse to a non-democratic institution, such as a court empowered to check the constitutionality of laws, might help to overcome the blocked political process. It might help to produce an outcome that, in these respects, better represents majority preferences than the legislative process does. If a particular authority is anointed with democratic validity only by virtue of its electoral pedigree, such an outcome might be thought to lack legitimacy. But it nonetheless passes the test of democratic self-determination. It will no doubt be noted that this last suggestion resembles a celebrated theory by John Hart Ely who attempted to support the power of judicial review by appealing to the integrity of the democratic process. His claim is that the United States Supreme Court’s power to overturn legislative decisions can only be justified when it helps to remedy the malfunctions of democracy, such as defects in the functioning of communication channels or systemic disregard for the interests of under-represented minorities.20 Although this chapter is not the right place to review Ely’s theory and consider the claims by his critics (including Jeremy Waldron himself),21 a brief comment may be useful for further argument. As a general theory of judicial review, Ely’s thesis strikes me as erroneous. The basic problem concerns the existence of reasonable disagreement about the devices and processes of democracy. The question about why an unrepresentative body should have the last word in the debate about the best procedural devices for
PROTECTION OF CONSITUTIONAL RIGHTS
113
democracy merely replicates a dilemma – which Ely recognises as fatal to many theories of judicial review – about an unrepresentative body having the last word on the substance of laws. Further, the values of process are often indistinguishable from the values of substance.22 For example, freedom of speech is a procedural device that is necessary for the effective functioning of a democracy, but it is also a substantive interest of individuals that is protected by the constitution. So when, for example, the legislature compels broadcasting stations to respect Christian values (as it did in Poland),23 is it imposing constraints upon the channels of political communication or, rather, upon individuals’ rights to publicly express themselves as they wish? A natural answer would be: “Both”, but the process-oriented theory of judicial review would have us disregard the latter effect and focus on the former. The problem with the former interpretation is that virtually any speech might be seen to be directly or indirectly related to the political mechanisms of democracy. If this is so, the process-based argument collapses into a substance-based argument, and one is indistinguishable from the other. However, Ely’s theory may be instructive for our purposes in that it points to the fact that there can be democratically endorsed distortions of the representation process (even if not everyone agrees that they really are distortions). The existence of disagreement might be fatal to the problem of legitimacy (why should a court have the final word when the two institutions disagree about the proper devices of political representation?), but not to the problem of rights-protection. From the perspective of someone concerned with how to design institutions that protect individual rights, there is nothing irrational about using a court to remedy legislatively endorsed distortions of the right to self-determination. Returning to Waldron: it should be remembered his argument is not merely directed against judicial review but also, more fundamentally, against the very idea of a constitutional bill of rights. At that level, however, his critique simply reproduces the fact-insensitivity of his critique of judicial review. According to Waldron, demands for the constitutional entrenchment of a particular right reflect a particular combination of self-assurance (a conviction that the right is fundamental) and mistrust. The latter is implicit in [the proponent’s] view that any alternative conception that might be concocted by elected legislators … is so likely to be wrong-headed or ill-motivated that his own formulation is to be elevated immediately beyond the reach of ordinary legislative revision.24
Waldron finds this mistrust incompatible with crediting all citizens with autonomy and responsibility. Of course, whether any proponent of a specific institutional arrangement has reason to mistrust the legislators is contingent upon whether that person believes that the legislators are so “wrong-headed or ill-motivated” when they draft laws that they detract from, rather than contribute to, the citizens’ ability to exercise autonomy and responsibility. It is one thing to mistrust one’s fellow citizens; it is another to mistrust one’s elected legislators.25 Whether the latter sort of mistrust is symptomatic of the former is entirely context-dependent. It is not a matter of principle.
114
CHAPTER 5 2. THE FACT-SENSITIVITY OF A THEORY OF JUDICIAL REVIEW
Suppose that you have a relatively clear view about how constitutional rights should operate in your country in specific circumstances. In other words, you have a view about how to translate broad constitutional pronouncements of rights – equality before the law, human dignity, freedom of speech, freedom of association and so on – into specific outcomes of which you approve. What is important is that you approve of these articulations (that is, specifications of the preferred method of application of a general constitutional right to concrete issues) qua constitutional rights and not as free-floating values; in other words, you believe that constitutional rights will be imperfectly implemented if the political system translates the broad constitutional pronouncements in a different manner to that which you believe to be the correct one. Suppose, for example, that the constitution declares a right to freedom of speech, but does not make it clear whether defamatory statements about public officials are or are not constitutionally protected, and you believe the proper articulation of that constitutional right demands that such statements should be constitutionally protected (within specified limits). Or, suppose that the constitution provides for a right to equal treatment but does not make it clear whether preferences in favour of a traditionally disadvantaged group count as violation of that right or not, and you believe that they should not. In both these cases, it is important that you do not merely think that the preferred outcome (non-prosecution of speech that is defamatory of politicians and protection of affirmative action respectively) is politically or morally right, but that you think it is the correct articulation of these constitutional rights in the specific fact-situations. In other words, you genuinely hold these outcomes to be the correct interpretation of your constitution, rather than simply your ideological preferences. Your support for any particular constitutional right – and through an accumulation of the support for various particular rights, your support for a whole set of rights – is coloured by your view about the correct articulation of those rights. You have no reason to value a particular right unless it is (or you believe that it plausibly can be) articulated in a way of which you approve. For instance, there is no sense in valuing “freedom of speech” unless it can be articulated in a valuable way. If you believe that it is important to protect individuals who criticise officials, and that for this reason those individuals must enjoy a degree of immunity against defamation suits pressed by politicians, then you would only value a “right to freedom of speech” if it is (or can be) articulated in a way that entails such an immunity. Of course, every abstract “right” does many different things, and even if the right to freedom of speech in your country is not, and cannot realistically be, articulated in your preferred way with respect to the defamation of officials, it can do many other useful things; for example, it might protect private speech or protect journalists against having to reveal their sources. But these other things are also positive by virtue of a valuable articulation. Unless some such articulation can be made, you would have no reason to value a right to freedom of speech in your constitution. Constitutional rights are precious only by virtue of substantively valuable articulations.
PROTECTION OF CONSITUTIONAL RIGHTS
115
Provided you care strongly about constitutional rights, your view about whether an institutional system that is designed to enforce the constitution promotes or hinders these preferred articulations, on balance and in the long run, will obviously inform your evaluation of the institutional system. (Of course, the same institutional system does many other things besides protecting individual rights, such as providing for the smooth and efficient operation of a governmental system by allocating powers among different branches and institutions. Your overall evaluation must take into account these other functions of the system. This, however, stretches beyond the focus of this chapter, concerned as it is solely with the role of judicial review in protecting constitutional rights; thus we will proceed as if you would only be concerned with rights when evaluating your country’s institutional system). On balance and in the long run: these are important provisos. “On balance”, because to make this assessment you will have to compare the net result of the existing institutional system with the expected net results of alternative institutional systems. Put simply: if you live in a constitutional system with a robust system of judicial review (a system in which the judicial institution may displace legislative articulations of rights), you need to compare the net outcomes of your system with a scenario in which everything else is equal except for the unconditional supremacy of the parliament. “In the long run” is a difficult proviso to specify because its very nature is unclear. It is obvious that you must take a dynamic and historical approach, rather than a snapshot of one particular moment in time that might be an aberration or exception; on the other hand, however, an unduly long time-frame distorts the picture because people harmed by today’s legal system cannot be consoled by the promise that the system will produce more good than harm in the long run. How can we go about deciding whether the gains of judicial review exceed the losses from a rights-protection perspective? The “score card” of a constitutional court would arguably take the form of Table I. It would not simply call for a comparison between the number and significance of “correct” decisions (invalidations that are conducive to the implementation of a constitutional value that we endorse: Box 1) and “incorrect” ones (invalidations that are detrimental to an implementation of a constitutional value that we endorse: Box 2); the calculus would have to be subtler and more complex. “Incorrect” decisions (Box 3) would have to include the invalidations that are not conducive to a value that we share, as well as decisions that uphold a provision when the court should (from the point of view of a constitutional value that we share) and could (from the point of view of the legal resources available to it, such as the accepted conventions of legal reasoning) have invalidated it.
116
CHAPTER 5
Table 1 The calculus of gains and losses resulting from constitutional court decisions:
Invalidating
Upholding
Rights-enhancing decisions (“gains”) (1) Invalidations “wrong” statutes
of
(3) Upholding of “right” statutes
Rights-weakening decisions (“losses”) (2) Invalidations of “right” statutes (4) Upholding of “wrong” statutes
The second category might initially seem to be an inappropriate factor to place on the negative side of the score card. After all, one might claim that, if a decision erroneously upholds a rights-problematic statute, it does not detract from a country’s system of rights protection because the legislature would have enacted the provision anyway if there had been no constitutional court to prevent it from doing so. But this is not so. The existence of a constitutional court somewhat relaxes the responsibility – a special duty of care, so to speak – of the political branches to avoid creating legislation that might infringe constitutional rights. The very fact that a court can review the statute might encourage the other branches to be more cavalier with lawmaking; if bad laws are likely to be struck down anyway, the stakes are not as high. Legislators might try to test a particular provision while knowing that judicial scrutiny is likely, something that they may not have risked this in the absence of such scrutiny. Legislating in the shadow of constitutional review affects the motivations and the risk calculus of legislators. The erroneous endorsement of a rights-implicating provision is therefore a negative – rather than a neutral – factor in the calculus of costs and benefits of constitutional review. This is an important argument developed by Mark Tushnet in his recent book.26 Note that Boxes (3) and (4) of Table I deal with specific decisions by courts in the process of conducting judicial review; that is, decisions in which the courts decide to uphold the validity of specific statutes. This is the argument that, compared with a no-judicial-review situation, a loss occurs when a court decides to uphold a “wrong” statute (Box 4); on the other hand, by upholding a “right” statute (Box 3), the court makes it more difficult to launch a legislative initiative to amend the statute in a way that is negative (from our point of view). For example, a court partly “entrenches” a liberal abortion law that it upholds, and thus makes it more difficult to render the law less liberal in the future. A similar effect occurs in Box 4: when a rights-positive cause loses in the constitutional court, the rights-detrimental law acquires additional support and it becomes more difficult to annul it in future. As Tushnet suggests, “the rejected claims of rights simply drop out of political consideration instead of becoming ordinary political claims like any other”.27 This is a loss from the point of view of rights protection. It is not, however, enough to confine ourselves to a definite set of judicial-review decisions in the calculation of gains and losses from the point of view of one’s preferred interpretation of rights. Judicial review can affect the implementation of rights not merely through the impact of actual decisions (that is, when specific cases
PROTECTION OF CONSITUTIONAL RIGHTS
117
have already reached the court), but also by its very existence (regardless of whether a challenge to a rights-implicating statute has been actually launched). The fact of judicial review just being there – “judicial overhang”, as Tushnet calls it – can have rights-positive or rights-negative consequences.28 On the positive side, the existence of judicial review can constructively influence the motivations and incentives of legislators if it makes them more cautious of rights than they might have been in its absence, and they end up adopting statutes that more closely accord with our understanding of rights. Further, the existence of judicial review – and publicly available information about the case law of the court(s) exercising this power – can have an educational effect and promote the “right” understanding of constitutional rights among the legislators and the general public. However, this does presuppose a degree of awareness about judicial decision-making that is often not reflected in the general public or legislators.29 On the negative side, one has to consider cases in which legislative irresponsibility is generated by judicial review; that is, when the awareness of possible review makes legislators less attentive to constitutional rights, with the possible result that a sub-optimal law will never be invalidated. Another negative consequence could be legislative apathy in the implementation of constitutional rights (along the lines: “if something is wrong, the court will remind us of it”).30 The very existence of judicial review can also have a negative educational effect; it might help to generate the perception that rights discourse is an obscure activity reserved for lawyers, and that deliberation about the political values that give rise to specific articulations of rights is something over which neither the population nor its elected representatives have any control. As Ronald Dworkin puts it, ”[t]here is little chance of a useful national debate over constitutional principle when constitutional decisions are considered technical exercises in an arcane and conceptual craft”.31 However, while this last effect is regrettable from the viewpoint of a participatory conception of democracy, it is not necessarily detrimental to any specific articulation of preferred rights. It might be a neutral matter from an individual’s perspective on which rights should be articulated, and it will only register on the “losses” side if one has reason to believe that public apathy about rights in general will detrimentally affect one’s own set of preferred articulations of rights. This may or may not be the case. Table 2. The calculus of gains and losses resulting from the very existence of the system of judicial review:
Affecting legislative incentives and behaviour Affecting society at large (educational role)
Rights-enhancing effects (“gains”) (1) Promoting consideration of rights in legislation (3) Promoting pro-rights attitudes
Rights-weakening effects (“losses”) (2) Promoting legislative negligence towards rights (4) Promoting apathy visà-vis rights
118
CHAPTER 5
It must again be emphasised that the effects in Boxes (1) and (3) count as “gains” only if the very existence of judicial review promotes an articulation of “rights” that accords with our preferred understanding of particular “rights”. In this sense, the calculation in Table II is just as substance-dependent as that in Table I. For example, suppose someone believes that, properly understood, the right to equality mandates affirmative action in some circumstances but the court in her country consistently passes decisions that invalidate affirmative action under the constitutional right to equality (and thus count as losses under Table I, Box 2). The fact that this attitude of the court affects a “consideration of rights” by the legislature (in the sense that it discourages parliament from even trying to establish affirmative action by statutes) cannot count as (1) in Table II; it must be (2). It would not make any moral sense to say: “I believe in the rightness of affirmative action, but I still believe that it is better to have legislators consider rights when deciding about affirmative action, even if a result of this consideration – operating as it does in the shadow of judicial review – is that politicians will refrain from enacting affirmative action programs”. If the acceptance of affirmative action is the correct interpretation of the right to equality, then “judicial overhang” (which, by virtue of its existence, paralyses legislators from instituting affirmative action because they know that it is likely to be invalidated) cannot count as a gain from the point of view of rights protection. 3. RIGHTS PROTECTION AND DISAGREEMENT ABOUT RIGHTS Tables I and II in the previous sub-chapter illustrate the consequential calculations that can be conducted by someone who has a preferred articulation of rights and who uses it as a yardstick for assessing the net benefit of a system of judicial review. But this is not the end of the story. We have thus far been disregarding the consequences of significant disagreements about the preferred articulation of constitutional rights. Suppose that I have a set of preferred interpretations of constitutional rights; let us call this S-1. Imagine that the constitutional court of my country consistently gives effect to articulations that contradict these preferred articulations; we can call the Court’s actual set of authoritative rights articulations S-2. Would I still have any good moral and political reason to support robust judicial review in my country? The answer is, “it depends”. I must assess the chances of having my preferred interpretations enforced by the legislature. If they are higher than the chances of convincing the court to adopt my own understandings, then I acquire a good prima facie reason against supporting the system of judicial review. It is, however, only a prima facie reason; in real-life situations there is likely to be a degree of overlap between my preferred interpretation of the rights (S-1) and that of the Court (S-2). After all, if the Court gives effect to S-2, there must be a constituency that espouses S-2 and expects the Court to enforce it. It is hard to imagine that there is no overlap in the articulations of rights offered by different constituencies in one society: such a radical division would be a sign of a very fundamental breakdown in social cohesion, and it seems unlikely that any system of judicial review could operate in the context of such fundamental dissensus. In reality, there must be a degree of overlap between different constituencies. Therefore, the fact that the Court articu-
PROTECTION OF CONSITUTIONAL RIGHTS
119
lates at least some constitutional rights in the same way as I do gives me at least one good reason to support judicial review. This reason can still be overridden by my dissatisfaction with the Court’s articulation of other constitutional rights but the picture is no longer uniformly negative towards judicial review. I will now need to consider how important those other rights are to me in comparison with the right with respect to which the Court’s articulation coincides with mine; I will also need to consider how likely it is that the legislative interpretation of all those rights would be closer to my preferred articulation than to that of the Court. Note that although this reasoning identifies an area of consensus, the overlap recognised in the preceding paragraph has nothing to do with Rawlsian “overlapping consensus”;32 as a mater of fact, it is just the opposite. Overlapping consensus in Rawls’s scheme occurs as a result of the congruence of conclusions reached by different people on the basis of different grounds. Rawls postulates that we should ignore the disagreement over the justifications in favour of the consensus regarding the conclusions. That type of agreement has little relevance to a rights advocate who reflects upon whether or not she should support a system of judicial review. Her problem is not that people have different reasons for supporting the same conclusions, but rather that other people have reached different conclusions on the correct articulation of constitutional rights, and that the Court has chosen to give effect to those people’s articulations rather than to her own. In other words, the Rawlsian “overlapping consensus” is part of the political theory and, as such, is addressed to the authoritative bodies in the society (such as the courts); in contrast, this chapter is about which institutional system a citizen concerned with rights should support when faced with an institutional conclusion that differs from her own interpretation. Now suppose that I have identified a degree of overlap between S-1 and S-2, and I know that a sizeable social constituency supports S-2. Of course, I believe that members of this group are mistaken and that constitutional rights are suffering due to their error. Neither of these facts – the overlap or the existence of a constituency favouring S-2 – are conclusive factors in my support for judicial review. It would be perfectly rational for me to demand that the S-2 constituency acts through democratic political mechanisms only, and that it attempts to win the support of a legislative majority for the S-2 conception. At this point, much depends on my theory of rights. If I believe that the standard of justification required to support rights-implicating action is different to that required for many other types of action (as most rights theorists indeed believe), and that rights-implicating actions require the giving of more careful reasons than do other authoritative decisions, I may have reason to believe that the judicial decision-making process is more likely to properly consider rights than is the political process. This is a familiar type of argument: courts are said to be more immune from political pressures, less subject to shortterm political incentives, more at home with reasoned deliberation, more transparent in their giving of reasons, etc. As Ronald Dworkin puts it: “A judge who is insulated from the demands of the political majority whose interests the right would trump is ... in a better position to evaluate the argument [of principle]”.33 In contrast, legislatures are thought not to be “ideologically committed or institutionally suited to search for the meaning of constitutional values, but instead see their primary
120
CHAPTER 5
function in terms of registering the actual current preferences of the people. . . .”. 34. Michael S. Moore summarises the argument thus: “The institutional features of judicial office – notably job security – make judges better able to focus their deliberations on the moral aspect of any problem, putting aside all the questions of political expediency with which legislators must grapple”.35 I do not want to labour these well-known arguments. Instead, I would like to emphasise their contingent nature: they are context-dependent and hinge upon the institutional design of a specific court vis-à-vis the legislature in the same country. The giving of reasons is almost non-existent in some judicial review systems (consider, for example, the brief, terse, purely legalistic grounds for decisions given by the French Conseil constitutionnel; it has no place for dissenting opinions, or discussion of the moral issues involved, and the French voter can surely learn more about the motives of his or her elected legislators than those of the Conseil constitutionnel judges). But even that paragon of deliberation and public reasongiving, the United States Supreme Court, has led Ronald Dworkin to admit to the “tentative” nature of the suggestion that “judicial review may provide a superior kind of republican deliberation”.36 However, the contingency of the argument does not indicate that it is weak; it only means that it will apply to some countries and not to others. We began this chapter by claiming that any theory of judicial review must be fact sensitive: the facts about the comparative institutional properties of courts and of legislatures are among the most important ones that need to be considered. From the point of view of protecting constitutional rights, it is crucial to establish a link between the protection of rights and the institutional differences between courts and legislatures. Such a link is usually asserted in the following way. Rights, the argument goes, are based on sound moral grounds rather than on mere expressions of preferences that might be purely interest-based. The institutional situation of courts, which have to support their decisions with reasons and are thus subject to public scrutiny, may suggest that they are in a privileged position when it comes to reasoning in terms of rights. As one American constitutionalist has stated, “the duty to write opinions gives judges an incentive to examine the reasons for their decisions, since they know they will have to justify the result in a document subject to public criticism”.37 One could therefore conclude that there is a link between a particular institutional feature of courts – a duty to justify their decisions in a publicly transparent fashion – and the tendency to consider a given matter from the angle of the implicated rights, as opposed to testing it merely against the preferences of the constituency of a given institution. “Reason-giving”, however, is too crude a concept to be understood as a factor of improving the rights-orientation capacity of a given institution. It also matters what sort of reasons a given institution (a court, a legislature, etc) is qualified, and expected, to give. Reason-giving connects with a rights-orientation if an institution is expected to provide general moral reasons for its authoritative decisions: rights are then seen to reside in those moral rationales for a particular authoritative reading of the constitution, of a “moral reading” of the constitution, to use Ronald Dworkin’s words.38 We may indeed justify an institutional design in which a particular
PROTECTION OF CONSITUTIONAL RIGHTS
121
institution will review the enacted statutes under the criteria of their consistency with the proper reading of constitutional rights, where that reviewing institution will be required to provide good moral reasons for its decisions, as a guarantee that it will be guided by rights-considerations rather than pandering to the preferences of its constituency. But there is nothing particularly judicial about such a review; indeed, such a vision is almost antithetical to the institutional capacities, qualifications and skills of courts.39 This is because courts (including constitutional courts) are typically not expected, nor do they normally attempt, to provide the sort of general moral scrutiny of statutes that is called for in an institutional vision just outlined. To the contrary, they normally represent their decisions as purely “legal”, based on specialised technical reasoning that attempts to stay away from direct appeal to moral and political reasons. Nothing in this argument, of course, hinges upon a formal characterisation of constitutional courts as part of the judicial branch; indeed, it was suggested in Chapter 2(2) that constitutional courts in CEE, similarly to their West European counterparts, in many respects resemble legislative chambers sitting in judgement on the laws enacted by lower chambers. The point is that in some other respects these courts do behave in a quasi-judicial rather than parliamentary way, and from the point of view of the issue currently under discussion here (the deliberative character of constitutional courts and its relationship to rights-orientation) these courts are quite defective as instruments of the sort of broad moral reason-giving called for when rights are at stake. The sort of reasons that constitutional courts usually provide for their decisions, even if quite fundamental moral dilemmas are at stake in a fairly transparent fashion, are the reasons that normally do not figure in actual arguments for a particular reading of a constitutional right. Rather, they are technical legal arguments, mimicking the reasoning of “ordinary” judges. True, they do give reasons for their decisions, but these reasons are so stylised, formalistic and technical that they do not qualify as the reasons that we – the concerned citizens – appeal to when arguing about the best articulation of an abstract constitutional right. To be sure, these stylised, formalistic and technical reasons normally offered by courts often are just a disguise for open moral or political arguments; they hide the sort of moral reasons that are normally expected to figure in the reasoning about rights articulations. This being said, it does not render courts any more qualified to orient public deliberation in the direction of rights: the general lack of judicial candour is not a redeeming feature but rather an institutional defect of courts in this particular regard. This is because the connection between reasoning and rightsorientation resides not merely in having reasons, but also, and crucially, in giving the reasons for one’s decision: the public transparency and explicitness of the reasons underlying a particular decision is seen as a factor maximising rights-orientation. If, however, the reasons publicly offered by a given institution hide rather than reveal the actual reasoning leading to an authoritative decision, then the potential for connecting this institution with progressive rights-orientation is lost, or at least seriously compromised. The connection between the deliberative nature of a given institution and the deliberation on rights in society at large is eroded when the former is constrained by canons and conventions that have no equivalent in the public discourse on rights, and therefore cannot resonate with the deliberative
122
CHAPTER 5
process in general. In other words: for the reason-giving feature to be part of the institutional structure aimed at maximising the rights-orientation of an institution it must be seen as an aspect of this institution’s public accountability; the canons of appropriate judicial reasons publicly offered, which tend to disguise the actual reasons behind a decision, undermine the possibility of such accountability. The giving of reasons is just one of a number of institutional circumstances that may affect the comparative incentives and capacities of courts (vis-à-vis the legislature) to reason in terms of rights rather than mere preferences. The circumstances of selection and tenure of judges is another: those judges who are selected in a process that links them with particular institutions (such as the president), or groups (such as parties in the parliament) that have a clear interest-orientation, may feel constrained to pursue interests rather than values in their decision-making; while judges who have limited tenure and depend upon interest-based political forces for their re-election or a comfortable post-tenure appointment may have disincentives against pursuing the best moral values in their rights-articulating decisions (as indicated earlier, in Chapter 1(3)). In contrast, constitutional judges whose appointment results from a process which mixes and merges the input of various institutions, and those who have a secure life tenure, may find it “easier … to act on their moral convictions”.40 The established structure of argument and deliberation might be yet another factor: even if judges have an incentive to reason in terms of rights, they might lack the capacity to do so if proceedings are structured in a way that limits their ability to engage in a serious consideration of all of the aspects that might influence their preferred articulation of the constitution. In a highly adversarial model of appellate judicial proceedings – such as in the United States – those same factors that are often cited as improving the impartiality of a trial can simultaneously handicap judicial inquiries into a wide range of moral issues that might be relevant to the rights in question. Deliberation can actually be impoverished rather than improved if judges, in contrast to the legislature, “get their information solely from the briefs and records prepared for the case sub judice”, “are prohibited from seeking outside advice (except by way of amici curiae briefs)”, and if “[o]nly the parties to the case may be heard in each matter, and public participation in the process, whether by letter-writing or by demonstration, is very much discouraged”.41 Surely the link between certain institutional features and a rightsorientation in judicial deliberation must be contingent upon the degree to which the judiciary is subject to such limitations on access to sources of information and alternative arguments. Suppose, arguendo, that we have successfully established a link between the institutional design of the court and the tendency to take rights seriously. All this tells us is that, in a system with judicial review, our political system is more likely to give effect to some understandings of rights than it would in the absence of judicial review. But returning to the hypothesis that we introduced earlier, if the rights articulation of our particular court is more likely to be S-2 than S-1, do we have a reason to support such a system? In other words, is it better for us to have S-2 than weak (or non-existent) protection of rights across the board? In order to answer this, we need to pose a fundamental question: what good is produced by protecting rights in abstracto that is distinct from the good produced by protecting specific
PROTECTION OF CONSITUTIONAL RIGHTS
123
articulations of rights? The usual answer is: the very idea of rights presupposes limits upon the exercise of political power, and so protecting rights – any rights – limits political power, regardless of whether we agree on the specific limits of those rights. But is this answer compelling? Consider Waldron’s response to this point, given in the context of his criticism of the contention that we need counter-majoritarian measures to give effect to the idea of rights: [W]e should not underestimate the extent to which the idea of rights may pervade legislative or electoral politics. The idea of rights is the idea that there are limits on what we may do to each other, or demand from each other, for the sake of the common good. A political culture in which citizens and legislators share this idea but disagree about what the limits are is quite different from a political culture uncontaminated by the idea of limits, and I think we sell ourselves terribly short in our constitutional thinking if we say that the fact of disagreement means we might as well not have the idea of rights or limits at all.42
To translate Waldron’s point into the language of the question I have just formulated: it is better to have a political system in which rights matter, that is, they are taken seriously and are protected, regardless of whether we agree with the dominant articulation of those rights, than to have a system “uncontaminated” by rights. Rights express an idea of “limits” upon the sacrifices we might impose and demand in the name of the common good, and this idea is valuable per se. However, I am unsure as to whether this idea of “rights as limits” gives us as much mileage in supporting the very idea of rights, as Waldron seems to suggest. Our disagreement over the proper articulation of rights is not merely disagreement about the proper limits of rights; rather, we disagree about whether a particular articulation is a limit in the first place. Take, for example, a disagreement over the constitutional status of abortion as a consequence of specific articulations of two intersecting rights: the right to life and the right to privacy. A “pro-choice” advocate will believe that the proper articulation of these two rights will result in the constitutional protection of women’s reproductive decisions: imposing a duty on women to give birth against their wishes would transgress the state’s limits on what we may do to each other. A “pro-life” advocate will claim that to tolerate abortion is to impose an intolerable penalty on the foetus, and would violate the limits on what the government may permit one person to do to another. Proponents of both views can argue about the limits on the sacrifices that can be extracted for the purpose of the common good, and each might consider that the opposing position violates these limits. The pro-choice person’s “limit” is a restraint on governmental action, and the pro-life person’s “limit” is a restraint on individual action, but this is not relevant to the very idea of rights as “the limits on what we may do to each other”. It would be relevant if we believed that the only intelligible rights are those that give us claims against official action, but this would be an excessively restrictive conception of rights. Here is another example. Imagine that we disagree about whether the right to free speech should protect me if I defame politicians (unless a politician can prove actual malice on my part). Suppose I believe that this articulation gives effect to a limit on what the government can do to me; namely, I cannot be required to pay
124
CHAPTER 5
damages to the politician whom I have defamed, even if such a requirement to pay damages could be seen as contributing to the common good (because it would establish counter-incentives to irresponsibility in public expressions). However, someone who disagrees with this articulation of rights – a defamed politician, for instance – might say that his preferred articulation of the right to free speech is that I should pay damages. His preferred articulation gives effect to the idea of rights as limits – it limits what I can say about him, even if my comments could contribute to the common good – and consequently limits the state’s authority to support the defamer’s position. According to the politician, my impunity will reflect a system “uncontaminated by the idea of limits”. So the disagreement is not just about where the proper limits lie, but also about what properly constitutes a limit. This means that, in abstraction from articulations of rights that we actually hold to be valuable, an appeal to rights as limits cannot provide a reason for preferring a regime of rights over a regime of no-rights. 4. PRUDENCE AND JUDICIAL REVIEW Consider a familiar pattern of reasoning. We all want our respective rights articulations to be enforced in our society, but we cannot all have it because we disagree among ourselves, and S-1 and S-2 have areas of incompatibility. The second-best solution would be to have some articulations of rights enforced (even though they will not all be my articulations) on the basis of an expectation of reciprocity, rather than having rights counting only weakly in authoritative decisions. I submit to your articulation regarding some rights because, if I don’t, I might lose the dominance of my articulation of another right. How strong is this argument in the overall configuration of reasons in favour and against judicial review? Not very strong, I would suggest. Some characteristics of rights reasoning render the analogy with standard prudential reasoning less than adequate. Standard prudential reasoning derives much of its attractiveness from an appeal to simple utility maximisation: it makes good sense for me, as a utility maximiser, to concede some value V-1 today in order to gain another value V-2 tomorrow if V-2 is more important to me than V-1. But rights reasoning does not benefit from such an appeal because, in contrast to utility-maximisation, it is not solely agent-oriented. Under the standard prudential argument there is an identity of the subject and the beneficiary of the reasoning process; however, any such identity is not necessarily present in reasoning about rights because we often demand rights that benefit others. If I support affirmative action it is not necessarily (and perhaps not typically) because I am a likely beneficiary of the preferences granted by such a policy. Rather, I support it as a proper articulation of the right to equal treatment because I believe that it is consistent with my understanding of the correct meaning of equality rights, and because I believe that it is just to treat others in this way. This “other-regardedness” of rights reasoning indicates the limits of the prudential argument when applied to judicial review of constitutional rights. Prudential reasoning involves a sacrifice of V-1 today in the expectation of securing V-2 tomorrow. The prudential sacrifice would, however, have a different form with
PROTECTION OF CONSITUTIONAL RIGHTS
125
respect to rights: it would be necessary to sacrifice some rights for some people in order to gain some other rights, possibly for other people, in the future. Therefore, if I decide to “sacrifice” V-1 now in order to protect V-2 in the future, I am in fact choosing the winners and the losers of a particular system of institutional design. It is, however, true that it is often impossible to precisely identify the likely beneficiaries of a specific, authoritative articulation of rights. The rights to free speech, freedom of religion, freedom from unjustified seizure, and the whole set of rights implicated in the judicial process, cannot be characterised, ex ante, as benefiting one class of citizens more than others. I might have a stake in fighting for the strong protection of criminal defenders, even if it is rather unlikely that I will ever be in their position. Still, the possibility of becoming a criminal defender one day cannot be excluded, and I wish to insure myself against weak protection in that case. To the extent that such a motivation for arguing in favour of a particular rights articulation is plausible, the analogy with the standard prudential argument holds. To the extent that it is not plausible (that is, where the disparate impact of the different rights I am advocating is reasonably identifiable with respect to specific classes of people), my support for judicial review must rely on non-prudential arguments. 5. CONCLUSIONS The role of constitutional courts in protecting constitutional rights cannot be captured by any simple formula or defined as “matters of principle”: it can be declared neither that, as a matter of principle, a robust power of constitutional review is necessarily an important ingredient of protection of human rights nor that it inevitably damages such protection. The starting point for an assessment of the role of judicial review must be a careful calculus of gains and losses resulting from a system of judicial review in a given country. The gains and losses resulting from a set of specific decisions, and also those resulting from the very existence of judicial review in that country, must be assessed. (The latter includes the ways in which judicial review affects legislative behaviour – positively or negatively – and the educational role it has in improving rights-consciousness within the community at large.) The outcome of this complex calculation clearly depends upon our preferred articulations of abstract rights, and people who disagree with our articulations will also disagree with the final verdict concerning the role of judicial review in rights protection. This fact of disagreement must also be taken into account in the reckoning. While we might doubt the general net benefit of judicial review, we might have some prudential reasons to support it; that is, it might be rational to support judicial review if the institutional particularities of judicial institutions, compared with those of the political branches, render courts more sensitive to rights considerations in general. This judgement, however, will be contingent on specific institutional comparisons and cannot be made in abstraction from the particular circumstances in a particular country. The review of the activities of constitutional courts in post-communist countries of Central and Eastern Europe in the succeeding chapters may hopefully help us to form such a judgement about how they have contributed to the protection of rights in that region.
CHAPTER 6 PERSONAL, CIVIL AND POLITICAL RIGHTS AND LIBERTIES
This chapter opens a five-chapter survey of constitutional rights in Central and Eastern Europe (CEE), as interpreted and articulated by constitutional courts. Special attention will be paid to those cases in which legislative articulations of rights were set aside, and invalidated, as unconstitutional: this is for the obvious reason that these cases give us the best insight into the role of constitutional courts in shaping the general patterns of protection of rights, irrespective of the choices made by democratically accountable bodies. However, consistent with the conclusions of the previous chapter, I will also note those “missed opportunities”, where constitutional courts had the occasion to review a statutory provision that could have been fairly viewed as objectionable under democratic and liberal standards of rights, and deferred to the legislatures. It is only when we put together these two categories of constitutional courts’ rulings on rights that we can obtain the information necessary to assess the overall contribution of constitutional courts to the understanding of rights in the post-communist states of CEE. In this chapter I will look at some leading court decisions in the areas of personal, civil and political rights and liberties which were arguably the most suppressed under communism, and consequently the most desired after the collapse of the authoritarian system. It is in these areas that most learning had to be done, by the legislatures, lawyers and citizens, and thus where the contribution of constitutional drafters and constitutional courts was the most crucial. In contrast to socioeconomic rights or the anti-discrimination principle, where the old legal doctrine had at least paid lip service to developing criteria of plausible conceptual articulations, civil and political rights had been most infected by the hypocrisy and cynicism of “legal doctrine” in the service of non-democratic political rule. The tasks that the constitutional drafters and interpreters faced were therefore formidable, and the way they discharged these tasks deserves special attention right at the outset of our survey. The classification of rights into personal, civil and political in this chapter is purely conventional; nothing much hinges upon the characterisation of any particular right as belonging to this or that category. I will begin with the rights that can be characterised as most personal: the rights to life and dignity, in the sense that they appertain to an individual herself, with only the minimum significance, if any at all, attached to her involvement in the state and the society. Perhaps religious rights also come close to this category, even though the social context and political
128
CHAPTER 6
embeddedness obviously play a higher role here than in the former case. I will then look at some of the rights that can be better understood as civil (in the sense that they attach to a person’s role as a citizen, in particular in the context of voting, petitions, assembly, etc), and political – in particular, the set of rights related to free expression, the key (although not the only) ingredient of which is based upon the rationale of political participation. Of course, it goes without saying that these categories overlap, as each of the rights in question may have multiple rationales that transcend simple categorization as “personal”, “civil”, and “political”; hence the distinction is made here merely for the purposes of clarity of exposition. 1. A RIGHT TO LIFE AND DIGNITY All CEE constitutions contain provisions on the right to life, with most of them referring also to the right to dignity. Sometimes, these two provisions are spelled out in one breath, as in Hungary, where the Constitution states that “every human being has the inherent right to life and human dignity”.1 In some of the constitutions, “human dignity” is articulated not as a right but as a meta-value: a source of other rights and liberties. This is the case in the Polish Constitution, which states that “the inherent and inalienable dignity of the person shall constitute a source of freedoms and rights of persons and citizens” and that “human dignity is inviolable, and the respect and protection thereof shall be the obligation of public authorities”.2 The wording in the Polish Constitution has its own history: during the long process of constitution drafting, there was a strong (though, in the constitutional commission, clearly minoritarian) pressure to include an invocation to “natural law” and “natural rights” in the Constitution, as a recognition of higher (in particular religious) sources of human rights and freedoms. This pressure was resisted but, in the process, a formulation in terms of “human dignity” was thought to be a compromise that could placate the natural rights proponents. In fact, many commentators of the Constitution interpret this provision as expressing the acknowledgment of a “jus-naturalistic conception of human rights and liberties”.3 Two main areas in which the right to life has been invoked before constitutional courts are, of course, the issues of the death penalty and abortion, and the main bulk of this section will be devoted to these two thorny issues. However, the “right to life” has also been, occasionally, employed in other contexts. For example, the Bulgarian Constitutional Court struck down, in 1997, a number of exceptions to the “excessive self-defence” rule of the Criminal Code, as contrary to the assailant’s constitutional right to life.4 The amendment to the Criminal Code specified a number of circumstances in which self-defence would never be considered excessive, regardless of the nature and danger involved. The Court deemed that most of the following exceptions to the principle of excess of self-defence were contrary to the assailant’s right to life, itself a constitutional right:5 when there are two or more assailants, when the assailant is armed, when the assault takes place in a vehicle, aircraft, ship or train, and when the assault takes place during the night. In all these cases, if the legislator had had its way, the argument that the self-defence employed had been excessive would not have been available to the assailant. The only
CIVIL AND POLITICAL RIGHTS
129
legislative amendment to the law of self-defence that the Court in the end upheld was the suspension of the usual excess rule when self-defence was against an assailant who had resorted to violence or burglary to get into the home. In this case, the Court invoked the constitutional right to inviolability of the home, and clearly gave precedence to that right over the violent assailant’s right to life. This seems like a bizarre conclusion, to put it mildly; when the dilemma is characterised as the conflict between two constitutional rights, to give precedence to inviolability of the home over the protection of human life seems distorted. That this, rather than other personal interests of the attacked person, was decisive for the Court is evidenced by the fact that, at the same time, it found the suspension of usual excess principles in situations where the assault takes place in a countryside property or business facility to be unconstitutional. This is because the principle of inviolability of the home does not apply to premises of these two latter types! The balancing implicit in the Court’s decision is almost perverse. There are many different ways of arguing against the penalisation of excessive self-defence, but to found an argument on the value of the inviolability of the home as prevailing over the value of life is probably the least morally appealing of these. Apart from being invoked (alongside the right to life) in the context of the death penalty, the right to human dignity has also been invoked in some other contexts – although rarely as a single, independent constitutional basis for the invalidation of a statute, usually being accompanied by some other constitutional principles or rights. One interesting example is the invalidation by the Hungarian Constitutional Court, in 1993, of an old (that is, dating from the era of state socialism) rule of civil procedure that allowed a public prosecutor to commence a civil suit if the person entitled to do so was unable to defend his or her rights, or for reasons of important state or social interest.6 Under the old system, this procedural avenue was a consequence of the role of the public prosecutor as the “general safeguard of legality”. The Court decided that, by denying a person full control over whether to initiate a suit to defend their own interests, this rule violates the autonomy of persons and their right to self-determination, and, in constitutional terms, the principle of the rule of law7 and the right to human dignity.8 This is an interesting use of the general principle of non-paternalism (not, however, expressed in these terms in the Court’s decision) understood as the principle that everyone is the best master of their own interests, and should decide for themselves whether, and when, to defend them. In Poland, the principle of human dignity (Art. 30) has not been articulated in the Constitution as a right, but rather as a source of rights,9 and therefore cannot be used as an independent basis for invalidation, although it has been occasionally used in conjunction with rights provisions. As the Constitutional Tribunal explained in one of its decisions: “Art. 30 of the Constitution defines the axiological grounds of rights and liberties guaranteed by the constitution” and, as such, “it is of a general nature and its contents have been further concretised in the specific [constitutional] rules regarding specific rights and liberties”.10 In this decision, the Tribunal considered a constitutional complaint from a citizen, against the law on passports. The complainant claimed that the procedure for refusing (or invalidating) a passport was unconstitutional, because, inter alia, it conflicted with human dignity. The Court considered in detail the possible clash of the law with a number of constitutional
130
CHAPTER 6
rules, but Article 30 was considered only very cursorily, and quickly rejected as a possible ground for invalidity. A year earlier, the Tribunal had invoked the same constitutional “human dignity” principle to invalidate a particular law, but again, only as an auxiliary principle, used to interpret more specific constitutional rights that served as the actual bases for invalidation.11 This case related to a decree by the Minister of Health, which provided for some specific forms of medical certification of temporary incapacity to work. The Ombudsman challenged the requirement of including a statistical code number that identified the illness in such certificates. The Tribunal agreed that this was contrary to the constitutional right of privacy12 and the right to protection of personal data13 and, in the process, referred to the principle of human dignity contained in Art. 30.14 In the structure of the Tribunal’s argument, this appeal to dignity had, however, a purely ornamental function; the law was invalidated not on its merits, but rather on the basis that any restrictions of constitutional rights must be (among other things) provided for by statute and not, as was the case here, by sub-statutory acts. One of the main areas in which the right to life, alongside the right to dignity, had special relevance was the question of the status of capital punishment. During the Communist era, the death penalty was widely used in the countries of the region, and this applies not only to the Stalinist period, when it was an inherent element of state terror, but also to the post-Stalinist days. All criminal codes in the region contained broad catalogues of capital crimes, including “anti-state” crimes, but also murder, crimes against property, “speculation”, etc.15 After the fall of Communism, most CEE countries abolished the death penalty; this tendency was largely driven by their joining of the Council of Europe which implied also (although not as an obvious legal obligation, but as a political consequence) the ratification of the Sixth Protocol to the European Convention on Human Rights (ECHR) concerning the abolition of the death penalty. Currently, all CEE countries except Belarus16 have signed the Protocol, and, of the signatories, all but Russia have ratified it.17 It is clear that, in the face of broad societal support for maintaining the death penalty in these countries,18 the need to comply with Western European norms, especially in the context of accession to the European Union (EU), has been a decisive factor in the abolitionist tendency.19 In some cases, the death penalty is not explicitly mentioned in constitutions, but the criminal statutes were purged of this penalty, as is the case of Poland and Bulgaria. In six constitutions20 it is expressly permitted, usually as “an exceptional penalty for particularly serious crimes”.21 Interestingly, two of the constitutions that allow the death penalty in this manner (Georgia and Moldova) have ratified Protocol 6 of the ECHR; there is, however, no real inconsistency there as both Moldova and Georgia no longer have any crimes in their criminal codes for which the death penalty is a possible punishment. The Amnesty International website states that Moldova abolished the death penalty in 1995, and that Georgia did so in 1997 (thus, before their ratification of Protocol 6), and yet appropriate constitutional amendments to this effect have not been made. However, such amendments would not be necessary because these two Constitutions explicitly only allowed for the death penalty until its abolition.22 It should be added that the only country of those under
CIVIL AND POLITICAL RIGHTS
131
study here that practices the death penalty in fact – rather than simply allowing for it in their laws –is (according to Amnesty International) Belarus.23 Several other constitutions expressly state that the death penalty is forbidden.24 In addition to these, constitutional courts brought about the abolition of the death penalty in Lithuania, Albania, Ukraine and Hungary. The latter decision is of particular importance.25 As the author of the decision, Chief Justice Sólyom was later to remark, this decision “signalled the real power and significance of constitutional review and shocked an unprepared Parliament and the general public as well”.26 This latter “shock” was all the more obvious since, at the time, about seventy percent of Hungarians supported keeping the death penalty.27 I have discussed the peculiar nature of reasoning employed in the Decision earlier,28 so a brief summary should suffice here. The main dilemma the Court faced was that, on one hand, it was dealing with a constitutional text that strongly implied the permissibility of the death penalty; only “arbitrary” deprivation of human life was constitutionally prohibited in Art. 54(1), thus implying acceptability of non-arbitrary deprivation. On the other hand, the majority of the Court was willing to ban the death penalty malgré tout. This has been openly admitted, ex post, by Chief Justice Sólyom, who conceded that “certain general values” upon which the decision rested “were established outside of the Constitution”.29 It is therefore somewhat ironic that, in his own concurrence with the decision of the Court, Sólyom distinguished the role of the Parliament in the abolition of capital punishment (the Parliament being “free to use any arguments” of practical, sociological or political nature) from the role of the Court, which “may only make constitutional arguments to justify its decision”.30 The opinion of the Court (as opposed to the more developed concurring opinions of some of the members of the majority, especially Sólyom) is very succinct by the standards of that court, and consists of assertion rather than arguments. Its key statement is that the Criminal Code’s provisions on the death penalty violate the “essence” of the right to life and human dignity, hence are contrary to Art. 8(2) of the Constitution, which proscribes those regulations that restrict the essence (“the basic meaning and contents”) of fundamental rights. If anyone had any doubts as to how this opinion might be reconciled with the implicit constitutional mandate allowing the death penalty when used in a non-arbitrary fashion (Art. 54(1)), the Court established that these two Articles (8(2) and 54(1)) conflict with each other, and thus that “[i]t is the responsibility of Parliament to harmonize them”.31 Of course, the only “harmonization” that the Court had in mind was to amend Art. 54(1) so as to remove any doubts as to the impermissibility of deprivation of life, arbitrary or otherwise. In this context, the Court emphasised that Article 8(2) was enacted after Article 54(1), and thus that the former should prevail. However, of course, the opposite conclusion was also available to the Court on the basis of the precept that a more general law should not derogate from a more specific one. The Court, naturally, resisted the latter interpretation, and it established the conflict between the two articles by stating that any instances of capital punishment “impose a limitation upon the essential content of the fundamental right to life and human dignity”.32 It is interesting that even the single dissenter, Justice Schmidt, accepted this proposition concerning the conflict between the two articles, and only dissented
132
CHAPTER 6
on the basis of the institutional competence argument, namely, that if a conflict between two constitutional provisions is established, it is up to Parliament to remove it by amending the Constitution.33 Within the majority, only one judge (Zlinszky J.) argued on the basis that the alleged deterrent effect of the death penalty was illusory.34 The remaining authors of concurring opinions agreed largely with the view of Chief Justice Sólyom that the question of the practical effect of the death penalty was irrelevant, and that, as a matter of fundamental moral principle, any exercise of such a penalty offended against the right to life and dignity, which “is from the outset illimitable”.35 So much for the death penalty. The second important problem in which appeals to a “right to life” were implicated in the region was, as one would expect, the question of abortion. None of the CEE constitutions expressly forbid abortion or family planning; they simply state that they protect the life of human beings. However, some constitutions go further, such as the Charter of the Czech Republic, which proclaims: “Everyone has the right to life. Human life is worthy of protection even before birth.”36 On the other hand, the Slovenian Constitution is unique in explicitly endorsing a “pro-choice” position, by proclaiming the freedom “to decide whether to bear children”.37 The question of abortion was prominently dealt with by two constitutional courts in the region: in Hungary and in Poland. The Hungarian Constitutional Court considered the matter of abortion twice, and in each case its decision was slightly more restrictive – but only just – than the choice of the political branches at the time. As Kim Lane Scheppele remarked, “Hungary’s adoption of a restrictive abortion law was not initiated in Parliament”.38 In fact, it had its origins in a Constitutional Court decision in 1991, which restricted the legislature’s discretion over the fashioning of the abortion law.39 The Court struck down the existing, relatively liberal abortion regime, on the grounds that it should have been regulated by parliamentary statutes rather than, as was the case at the time, by regulations promulgated by the Minister of Health (the Court found that the matter concerned fundamental rights and, as such, should be regulated by statute). However, the Court’s intervention did not end there, and it engaged in a lengthy moral and philosophical deliberation that, as Justice Sólyom later put it, “radicalized the question [of the link between the maturity of a foetus and its legal status]: if the foetus is a person, then his/her right to life deserves absolute protection, if not, it enjoys only relative protection”.40 The Opinion of the Court, penned by Chief Justice Sólyom, remarked that the actual status of the foetus cannot be ascertained through the usual means of constitutional interpretation (a point rejected by Justice Labady in his concurring opinion, which, in its conclusions, went further towards a restrictive, anti-abortion position than the decision of the Court),41 and that it was thus a matter for the legislature to decide.42 Despite this, denying the foetus legal capacity altogether (or making it conditional upon birth) was held not to be a satisfactory solution because “[t]he foetus’ right to life – in effect its right to be born – cannot be made dependent on the condition of its being born”.43 Justice Sólyom, in the substantive part of the Opinion, clearly hinted at the “pro-life” option, suggesting for instance that “individual human life is a uniform progression from conception, not from birth, to death”,44 and even that the “extension” of the concept of man [sic] to the pre-birth period, with the concomitant
CIVIL AND POLITICAL RIGHTS
133
extension of legal capacity to a foetus, “is comparable only to the abolition of slavery, but it surpasses even that event in significance”.45 Nevertheless he admitted, grudgingly, that one option open to the legislature is to decide “that only the fully mature foetus is entitled to legal status” or, alternatively, to “create a special legal status for the foetus within which one can distinguish [different stages] on the basis of age, maturity” etc.46 As a result of the substantive guidelines that the Court established (indicating that neither a complete ban nor full availability of abortion would be constitutional), and on the basis of the timetable set by the Court for the legislature to enact a new law, the Parliament passed, in 1992, a statute that added certain additional obstacles to the pre-existing process (such as explicit acknowledgement from the woman concerned that she was in a situation of crisis, a mandatory cooling-off period, and mandatory counselling about the risks of abortion). The Hungarian Constitutional Court then revisited the issue of abortion in November 1998 when it reviewed the 1992 law.47 At issue was a particular provision that permitted abortion within the first 12 weeks of pregnancy if the woman was in a “crisis” (or an “emergency”) situation. Such a situation was defined as those cases in which a pregnant woman was in a desperate mental, physical or social condition, to the extent that the healthy development of the foetus was endangered. The questionable aspect of this provision (from the perspective of the Court) was that the “crisis” or “emergency” situation was to be evidenced simply by the fact that the woman concerned had signed a declaration to that effect. While the Court found that the termination of a pregnancy based on an “emergency situation” was not unconstitutional per se, it objected to the constitutional validity of the regulation if it were based entirely on the pregnant woman’s own statement concerning her “emergency situation”. It recognised that the declaration provision was meant to protect the privacy of the women in question, but it found that an improper balance had been struck between the protection of privacy and the protection of life (which is also a duty of the State). According to the Court, it would not represent a disproportionate burden upon the woman’s privacy (and also her dignity and right to choose) to require her to provide further justification for the abortion. A second, independent, ground for finding fault with these provisions was the fact that, on the one hand, abortion protects the woman’s right to choose as against the foetus’ right to life, while on the other, the “crisis situation” was defined (at least partly) in terms of the best interests of the foetus. Hence, there is a contradiction (the interests of the foetus are invoked in order to deprive the foetus of its life), and such a contradiction (according to the Court) violates the constitutional principle of legal certainty. In conclusion, the Court declared these provisions of the 1992 law unconstitutional, and instructed Parliament to specify more clearly the criteria of what constitutes “emergency circumstances”, and to require the woman to provide justifications for her claim. An alternative course of action suggested by the Court was to counterbalance the law by better protecting the life of foetuses, and, in order to achieve this, to provide pregnant women with the proper psychological, medical, social and financial assistance. It is easy to see that the latter option effectively reduces the availability
134
CHAPTER 6
of abortion to a few extraordinary cases, such as the threat to the life or health of the mother, etc. The right of choice would be, in fact, fully eroded if this course of legislative action were to be taken; no wonder, then, that most legal commentators agreed that this decision would lead to a restriction of the previously liberal practices in this regard.48 Indeed, in 2000, in response to the Court’s decision, the parliament made amendments to the abortion act that included two new requirements: a mandatory waiting period of three days, and rules concerning the contents of counselling to be provided to women seeking abortion.49 The amendments also included a rather narrow definition of “a severe crisis” that might justify termination of pregnancy on a woman’s request. While these amendments, which entered into force in July 2002, restricted the availability of abortion in comparison with the law of 1992, this restriction was not as severe as the Court’s decision might have been read to recommend. In Poland, the Constitutional Tribunal has also dealt with the issue of abortion more than once. The first time was at the beginning of 1991, when the Ombudsman asked the Tribunal to consider a decree by the Minister of Health, of 30 April 1990, which established the right of a doctor to refuse to administer abortions, and also a doctor’s right to refuse to write a health declaration regarding the permissibility of abortion.50 The Ombudsman questioned this ministerial act not on its merits but on the ground that the Minister had no competence to regulate matters that are the subject of statutory regulation (namely, the Statute on Abortion of 27 April 1956), particularly as the statute did not include any delegated powers for the Minister. The statute had not provided for the possibility of a doctor’s refusal to administer legally permissible abortion (or to sign a declaration about the medical permissibility of such an act). In a 2:1 decision, the Tribunal found the ministerial act legal, and consistent with the constitutional right to freedom of conscience.51 It announced that there was no legal duty on doctors to write declarations concerning the acceptability of an abortion, or to perform abortions. The right of a doctor to refuse to perform such actions may be derived, according to the majority of the Tribunal, from the constitutional right to freedom of conscience, as well as from the doctors’ ethical code (incorporated into the legal system at statutory rank), which obliges a doctor to refuse to administer treatment that is, in the doctor’s view, harmful or unethical. The Ombudsman argued that the woman’s right to abortion is the norm, and thus that any restrictions are limited departures therefrom. Ominously, the majority of the Tribunal rejected this view. The Tribunal declared: “The prohibition of abortion is primary, and the statute [of 27 April 1956] only defines the conditions of permissible departures from the prohibition”.52 The single dissenting judge, Czeslaw Bakalarski, objected on the grounds that the Minister of Health exceeded the limits of statutory delegation and established a rule that he had no authority to create.53 In 1992, the Ombudsman asked the Constitutional Tribunal to consider the possible inconsistency between the (already mentioned) “Code of Medical Ethics” (which had been incorporated, wholesale, into the “system of valid law” by the statute on “medical councils”54 of 1989), with the Criminal Code and with the (liberal) law on abortion of 1956.55 The amendments to the Code of Medical Ethics restricted the doctor’s right to perform abortion to cases of pregnancy resulting from rape, or where there was a direct threat to the woman’s life or health, whereas the
CIVIL AND POLITICAL RIGHTS
135
law on abortion provided for the permissibility of abortion also in cases of “difficult life conditions of the pregnant woman”. Echoing the earlier judgment, the Tribunal stated that there is no duty incumbent upon doctors to perform abortions, because the law on abortion of 27 April 1956 “does not impose a duty to perform abortions but only withdraws legal liability for performing abortions under the conditions established in the law”.56 Hence, according to the Tribunal, there was no clash between the law on abortion and the rule of the Code of Medical Ethics. Four judges filed dissenting opinions. Lastly, the most momentous judgment was handed down on 28 May 1997.57 In this decision, the Tribunal struck down the provisions permitting abortion in difficult financial and personal circumstances, as included in the law of 30 August 1996. The central constitutional basis for this invalidation was the principle of Rechtsstaat (the democratic state ruled by law), which, according to the Court, elevated human beings and human life to the level of supreme values. This, the majority announced, was negated by the provision that made abortion permissible on the grounds of difficult financial and personal circumstances. I have discussed this decision before in this book.58 Here it suffices to say that of all the decisions of the Polish Constitutional Tribunal, or indeed of all constitutional courts in CEE, this has been probably the most outrageous case of judicial usurpation of law-making power. Contrary to the explicit will of the legislative majority at the time, with basically no textual constitutional basis (there was no “right to life” in the constitution in force at the time of the decision), and with a clear indication from the constitution-makers that life “from the moment of conception” was not the recognised constitutional understanding of the “right to life” in the new constitution (already adopted but not yet in force),59 the Tribunal handed down a decision of great social significance that invalidated nearly all60 categories of abortion. As one critic noted: “The Tribunal has elevated its own, rather unclearly articulated axiology to the level of a legal principle. I do not think that it is within its legally defined powers”.61 What is particularly disappointing is not only the end-result, but also the reasoning; convoluted, muddled and bearing all the marks of rationalisation of an indefensible conclusion. The decision met with strong social opposition; the Women’s Rights Centre, for example, denounced the decision as “against the dignity, and rights and freedom of women”.62 One point frequently raised by the critics of the decision was to note that “it seems no accident that the decision was made a few days before the Pope’s tenth visit to Poland”.63 2. FREEDOM OF RELIGION The constitutional models of the church-state relationship in post-communist countries situate themselves somewhere between the extremes of, on one end of the spectrum, a system with a formally established state religion and, on the other, the actively atheistic state. Naturally, neither of the extreme positions has been opted for. The latter extreme (an atheistic state) was sufficiently discredited under Communism; the closest to this model are those few constitutions that refer to a “secular” state64 and explicitly renounce the possibility of having a state-sponsored
136
CHAPTER 6
religion.65 However, strict separation, or “the wall of separation” between state and religion, to use the American parlance (which would preclude any form of state assistance to religious groups, any form of religious teaching in public schools, etc), has not been adopted either. This is in line with the more general European constitutional tradition about which one legal scholar has noted that “the constitutional pattern of cooperation (sometimes also understood as enhancing positive state obligations to support churches) should be considered as the most commonly accepted model”.66 The very word “separation” is only rarely used in the constitutional texts to describe the state-church relationship;67 this is usually explained in terms of “negative historical connotations”68 but a more plausible explanation seems to be the influence of churches and their political allies, which resent the very idea of strict separation. This resentment is usually related to such practical matters as the establishment of, and subsidies to, private religious schools or the teaching of religion in public schools. It should be emphasised that this issue was more delicate in some countries than in others. In some, organised religions have played a relatively insignificant political role (as in, for example, the Czech Republic), and so finding the right constitutional formula was a reasonably uncontroversial matter. In others (Poland is perhaps the best example), the dominant church has long been a very powerful political agent and, after the fall of Communism, it managed to establish itself in such a position as to enable it to exercise a high degree of control over the political process, including the constitution-making process, both directly and indirectly, via the parties that benefited from its support. As a result, constitutional debates concerning the best way to characterise the church-state relationship were protracted and difficult. The Catholic Church in Poland resented any use of the very concept of “neutrality” in describing the relationship of the state to religions, fearing that this would in fact mean an ideologically committed, secular state. On the other hand, secular parties and constitutional experts were concerned about the strongly “accommodationist” formulas, discerning in them an attempt by the Church to establish for itself a constitutional ground for demanding various privileges from the state; to the extent that the spectre of a church-state symbiosis and even of a “confessional state” was flagged up in the political rhetoric in these debates. In the end, a compromise was found that used the concept of “impartiality”69 (more palatable to the Church than "neutrality"), and that also referred to the principle of “cooperation [between the State and churches] for the good of the individual and for the common good”.70 In what follows, I will first discuss the interpretations of the principle of the separation of state and religion, and will then move on to the decisions on freedom of religion (with the status of conscientious objection as a special sub-issue). This distinction is based on the view that these two principles are distinguishable from each other and that they raise different dilemmas; indeed, there may be a degree of tension between them. This is because, as I have discussed elsewhere at greater length,71 the principle of separation of state and religion has a distinctly antiaccommodationist logic built into it; it dictates hostility to any form of legally significant recognition of any particular religion, discerning therein a form of “establishment” of religion. In contrast, the principle of religious freedom gives rise
CIVIL AND POLITICAL RIGHTS
137
to demands for the recognition of special claims (for exemptions from common duties, or assistance to observe practices) based on religious beliefs, whenever the general legal burdens might have a differential impact upon the adherents to any particular faith. Hence, the dynamics of the two principles place them on a collision course with each other – something not yet fully appreciated in the religious clauses jurisprudence in CEE. All post-Communist states adopted a form of separation of church and state, although the specific constitutional formulations vary from country to country. Some constitutions state the principle of separation expressly;72 others establish separation without using the concept, but by providing that there shall be no state church or religion;73 others still qualify this by adding an element of linkage between the state and the church. There are a number of ways in which the latter linkage is achieved: by providing that certain religions are considered “traditional” ones;74 by imposing a duty upon the state to support some of the work of churches, namely by promoting their access to prisons, the army, hospitals, etc.;75 by requiring76 or authorising77 the state to offer financial assistance to churches; or by providing for cooperation between state and church, while at the same time making explicit the autonomy of one vis à vis the other.78 In studying the ways in which separation provisions have been interpreted by the constitutional courts of the region, a decision of the Lithuanian Constitutional Court of 13 June 200079 is particularly interesting. The sheer length of the decision and the in-depth character of the reasoning suggest that the Court considered it to be an important foundational decision in the field. Under challenge were various provisions of the 1998 amendments to the Law on Education. In particular, Article 10(4) of this Law stated that, at the request of parents, state schools could be cofounded by a state-recognised traditional religious association. Part 2 of Article 32 stated that the heads of such schools were to be appointed and dismissed on the recommendation of the religious association, and that the religious association were to set the requirements of the “world outlook formation” (a code word for religious beliefs) for the staff. Both of these provisions were found to be unconstitutional. The former provision (Article 10(4)) was found to contradict Article 40(1) of the Constitution, which states that “State and local government educational establishments of teaching and education shall be secular. At the request of parents, they shall offer classes in religious instruction”. The Court found that the religious aspect of the provision applied to all school activities, not just its religious teaching; hence, it conflicted with the constitutional requirement of secularity of schools. The “coordination” of state action with the relevant religious institutions, when setting up educational institutions in the process provided for by Art 10(4), would apply to the secular aspects of these institutions as well as to religious ones. Thus, the powers exercised in the implementation of this provision (founding, reorganisation, closing down, etc., of those educational institutions) would depend, partly, on the will of religious institutions. This, the Court explained, was contrary to the principle of separation of state and religion. The provision allowing religious associations to set requirements as to the teachers’ convictions and views (Article 32(2) of the Law), was found to contradict
138
CHAPTER 6
Article 25(1) of the Constitution, which states that: “Individuals shall have the right to have their own convictions and freely express them”. To adopt the rule of Art. 32(2) would, according to the Court, mean that the religious associations would be able “to interfere with the convictions of the employees [of educational institutions], to exert influence on them, to find out their view on religion, faith, believing or nonbelieving”. This violates the principle of freedom of convictions. It is also contrary to Article 42(1) of the Constitution, which proclaims: “Culture, science, research and teaching shall be unrestricted.” In addition, the provision allowing heads of these schools to be appointed and dismissed by the religious association (also forming part of Art. 32(2) of the Law) contradicts the constitutional principle of separation of state and religion because its implementation would mean that “decisions of state and local government institutions [regarding appointment and dismissal of heads of those educational establishments] are dependent on the will of traditional churches and religious organisations”. Despite these invalidations, the Lithuanian Constitutional Court in its decision also upheld a number of the law’s provisions against the challenge of unconstitutionality, thus drawing a rather fine balance between a moderately secular reading of the Constitution and the petitioners’ (a group of MPs) ideas of strict separation of state and religion. For instance, the petitioners also challenged Art. 1(5) of the Law, which provided that one of the goals of the educational system shall be “to guarantee the same rights and conditions for members of traditional religions” as for everyone else, to bring up their children according to their convictions. According to the petitioners, this conflicted with a number of constitutional provisions, including freedom of expression and of conscience (because, they said, the point of these principles is to provide an opportunity “for people holding different views to live in an open and harmonious civil society”) and, more importantly perhaps, the constitutional principle of non-establishment of a state religion (Art. 43 (7)). The Court rejected this interpretation of non-establishment, and explained that this principle basically entailed that “there shall not be a state religion in Lithuania”. More specifically, the Court announced the following implications of the nonestablishment clause: churches do not interfere with state activity and with state officials; they do not “form the state policy”; the state does not interfere with the internal affairs of churches and religious institutions; the state and its institutions are neutral as regards the matter of religions; the adherents to religions are not discriminated against on the basis of their convictions. Under these standards, Art. 1(5) was held to be constitutional because, according to the Court, the provision simply implemented a constitutional right of parents to bring up their children in accordance with their convictions, including with their religious convictions. However, of course, this is precisely the point at which claims based on religious convictions clash with the principle of separation of state and religion, which, as the petitioners correctly observed, requires that people of different convictions meet and live together in common spaces (including public schools) regulated by religionneutral rules. The Polish Constitutional Tribunal considered questions related to the relationship between the state and religion (defined by the Constitution as that of “impartiality” of public authorities in matters of personal convictions, “whether
CIVIL AND POLITICAL RIGHTS
139
religious or philosophical”)80 and, more often than not, it caved into the pressure exerted by the powerful Catholic Church. (I should add that I do not imply that there was any direct pressure upon the Tribunal, but rather a more general social and political influence exerted by the Church to have its views legally recognised and supported by the force of law). Early in its post-transition activity, in 1990, the Tribunal considered the constitutionality (under the so-called “Little Constitution”) of religion classes in schools. It upheld the decree of the Minister of Education, bringing back the teaching of religion in public schools.81 The ministerial regulation to that effect was made under strong pressure from the Church. The decision of the Tribunal considered both procedural (can the ministerial decree be the source of such a regulation?)82 and substantive matters. The Ombudsman challenged the regulation on several grounds: that the teaching of religion in schools was contrary to the principle of secularity of public education; that this was also violated by putting grades for religious courses in the school report cards; that the right of schools to hang crosses in classrooms may offend the religious sensibilities of nonChristians and of non-believers, etc. All of these objections were rejected by the Tribunal. The Tribunal offered an interpretation of two constitutional provisions: the separation of state and church (Art. 82(2) of the Constitution then in force) and freedom of conscience and religion (Art. 82(1)). It announced that the former should be understood in the light of the principle of “secularity and neutrality” of the state. This means that churches cannot be subsidised, but it does not preclude making public school premises available for the teaching of religion, nor paying the salaries of religion teachers from public funds. The Tribunal also stated that the principle of secularity and neutrality would prohibit the compulsory teaching of religion in public schools or the inclusion of religious tenets in the public school curricula, but not the teaching of religion on a voluntary basis. Rejecting the objections to the introduction of declarations by parents as expressions of their will regarding their children’s participation in those courses, the Tribunal held that these declarations are not a violation of the right to silence on matters of religion because “the right to silence is not a duty of silence”.83 This argument relies upon a highly questionable distinction between “negative” and “positive” declarations (the parents were expected, under this law, to declare that they wanted their children to attend religion classes, not that they did not want them to attend the classes). What was ignored by the Tribunal is that the absence of a positive declaration amounts effectively to a negative one, and so the duty to announce one’s religious conviction is in fact introduced, albeit indirectly. This was not the last word of the Polish Constitutional Tribunal on the matter. In 1992, the Minister of Education issued a new decree on various details of religious teaching in public schools (and related matters, including displaying religious symbols in schools), and the Ombudsman challenged both the constitutionality and the consistency of numerous provisions with the relevant statutes.84 The only point conceded by the Tribunal concerned the “declarations of will” by parents regarding their children’s (non-) attendance in religion classes. The Ombudsman accepted that “a right to silence” is not equivalent to a “duty of silence”, as established in the decision discussed above, but nevertheless proposed that no public body or institu-
140
CHAPTER 6
tion may demand any declaration from citizens regarding their personal religious convictions. The declarations by parents regarding whether or not they wanted their children to attend religious classes amounted to just such statements, and they were thus, according to the Ombudsman, discriminatory. However, the Tribunal explicitly resisted the invitation to engage in reasoning at this level, and instead struck down this particular provision on the narrowest of grounds. It held the law unconstitutional on the basis that the statute upon which the ministerial decree was based had not delegated to the Minister the power to establish a procedure for the parents to withdraw their children from religious education classes; it merely provided for the declaration containing “the wish” to have one’s children participate in such classes. In other words, the Tribunal stood by its formalistic and highly illusory distinction between “positive” and “negative” declarations of will by parents, and failed to use this opportunity to restate the principle that the duty to make any declarations whatsoever that can identify citizens’ religious convictions have no place in a secular state. The remaining parts of this decision are an unmitigated disaster in terms of the principle of neutrality of the state towards religion. First, the Tribunal rejected the challenge to the provisions regarding the employment of teachers of religion (appointed by churches) in schools and paying them salaries from public funds. The Ombudsman argued that this was a form of subsidy to churches, and therefore was both unconstitutional and explicitly prohibited by a statute on religious freedom. To this, the Tribunal replied that salaries paid to teachers couldn’t be seen as a grant to churches, because “the teacher of religion cannot be viewed as a church or a religious organisation”. This, obviously, simply avoids the issue of whether a payment to a church official, or to a priest nominated to teach religion by a church, is or is not a form of subsidising the church; and the argument that the “teacher” (even a priest-teacher) is not “a church” is formalistic in the extreme. Second, the Tribunal rejected the challenge to the provision compelling schools to include the grades obtained by a student in religion classes (if taken) on school certificates. The Tribunal thought the inclusion of such grades to be a simple consequence of having non-compulsory religion classes in schools, because the law on education provides that school certificates should record all school teaching activities – both compulsory and optional. Third, the Tribunal rejected the Ombudsman’s challenge to the provision that entailed religion teachers (nominated by churches) becoming members of “teachers’ councils” in schools (a body partly deciding, and partly advising the school principal, on pedagogical matters). Again, the Tribunal’s response to hold any such presence of religion teachers on school bodies to be a consequence of the school’s legal duty “to organise teaching of religion” should parents want it. Finally, the Tribunal rejected the Ombudsman’s challenge to the provision that allowed public schools to display religious symbols (crucifixes) and to have prayers in schools. The Ombudsman argued that symbols and prayers outside religion classes amounted to a public exercise of religion and was inconsistent with the ban on compelling anyone to participate or not to participate in religious activities. While it is not physical coercion, public prayer before or after classes compels students to reveal their religious convictions and so undermines their right to silence,
CIVIL AND POLITICAL RIGHTS
141
according to the challenge. To this, the Tribunal responded that the provision concerning religious symbols and prayers confers an opportunity rather than a duty to hang crosses and hold prayers – an explanation that surely sounds hollow in a country heavily influenced by one numerically dominant and politically powerful Church. Further, the Tribunal also rejected the complaint pertaining to students who do not want to participate in prayers but feel compelled to do so, emphasising that such prayers must be the result of a common desire of students and teachers – conveniently disregarding the psychological situation of a few “dissidents” outnumbered by the participants in common prayers. The Tribunal also argued that if there were no prayers, this would actually violate the constitutional and statutory principle that no-one can be prevented from participating in religious activities – as if there was no room for prayers outside public schools. Additionally, the Tribunal used the occasion to announce its more general views on religious freedom and the separation of state and religion. It argued that these principles meant that both the state and the church should be “autonomous” in their activities, but that this does not imply isolation from each other; “just to the contrary, it should mean the possibility of collaboration in those areas that serve the common good and the development of human personality”. Thus understood, secularity and neutrality are not inconsistent, according to the Tribunal, with religious teaching in public schools, especially as this teaching occurs only at the request of parents and/or students. The Tribunal also announced that schools should play a role in the “maintenance of the religious identity” of students, as proclaimed in the law on education, and “the possibilities to place crosses [in the classrooms] and to hold prayers are among the ways to achieve this maintenance of religious identity”. All in all, the Tribunal established standards of religious freedom and separation of state and religion that pay lip service to the principle of neutrality of state towards religion but, in all practical respects, deny the spirit of such neutrality. I will now move on to discuss the right to free exercise of religious beliefs. All CEE constitutions, without exception, contain clauses guaranteeing the freedom of religion, in most cases85 describing this freedom as the right of citizens to manifest, profess, worship, etc. their religion. Beyond that, some of these constitutions also contain the following: an express right to educate one’s children in accordance with one's own religious convictions;86 a right to religious privacy, meaning that persons need not state their official affiliation to organs of the State;87 as well as provisions asserting the right not to hold a religious belief or be coerced into holding one.88 In addition to these essentially individual rights (although a number of constitutions state that the rights to manifest one’s religious beliefs may be exercised individually or in a group),89 some of the constitutions also proclaim group religious rights, such as the rights of religious groups and bodies to organise themselves,90 or to maintain schools and charitable institutions.91 With the exception of three constitutions,92 all of the statements of freedom of religion are accompanied by clauses limiting that right (either directly accompanying the statement of the right, or in general provisos concerning the limitations of any constitutional right). Almost all of the constitutions envisage restrictions on religious freedom, based on public order and public health and morals,93 and also some cases of conflict with the rights of others.94 Some constitutions restrict the right to freedom of religion where this could threaten
142
CHAPTER 6
national security,95 while others contain a restriction relating to any incitement or expression of hatred or enmity.96 As an example of a constitutional court decision regarding the articulation of religious freedom, consider a decision by the Polish Constitutional Tribunal in 1999 concerning a complaint raised by a professional soldier and student of a military technical academy who, as a convert Jehovah’s Witness, had decided to abandon the army on religious grounds.97 He lodged a constitutional complaint against a ministerial regulation that required persons in such a situation to repay the costs of training and education incurred by the state so far. The complaint was based on the principle of equality (and, in this respect, he was successful), and also on the grounds of constitutional freedom of religion, arguing that such a heavy burden renders it excessively difficult for him to reconcile this duty to repay with his religious convictions. The Tribunal did not accept this latter claim, basically arguing that the duty to repay had nothing to do with the religious convictions of the soldier: “The burdens resulting from the provision that is under challenge here and which are related to a particular way of executing the debt towards the State, apply to every professional soldier who is in the same situation, and religious beliefs are not a criterion that defines this category of persons”.98 The Tribunal also observed, in this context, that the special status of soldiers affected a number of their constitutional rights, including privacy, the right to travel freely, the right to freely choose their occupation, etc. Hence, it found no reason to analyse the provision in question “from the angle of freedom of religion”.99 This is true only insofar as we limit our attention (as the Tribunal did) to the direct, intentional impact of a regulation upon the free exercise of religion. If, however, we include within the ambit of our concerns of the potential for indirect effects upon the right to free exercise of religious freedom, the decision of the Tribunal appears to be much more problematic and insensitive to the religious concerns of the claimant. After all, regulations that restrict the capacity of various people to avoid religion-burdening measures may have a different impact upon believers and non-believers (or upon adherents to different religious faiths, depending on their attitudes to military service); a matter conveniently overlooked by the Tribunal. Finally, two interesting decisions by the Macedonian Constitutional Court regarding freedom of religion should be mentioned here. In the first, the Court struck down the law on religious communities and religious groups.100 The law stated that only registered religious groups could conduct religious activities. To be registered, there had to be at least fifty founders of the group, who had to provide numerous personal details about themselves, the seat and scope of activities of the group, its meeting places and information about persons in charge of the work of the group. The Court found that these requirements violated the constitutional guarantee of freedom of religious convictions, in that they enabled the punishment of persons who perform religious activities outside of a registered religious community or group. Also, the requirement of a minimum number of founders, and the information that they must provide, was found to be unconstitutional because such requirements do not apply to the setting up of other associations. Therefore, the law was also in conflict with the constitutional right of freedom of association, as it restricted the right to form associations of religious character.
CIVIL AND POLITICAL RIGHTS
143
A second decision, handed down a year later, struck down certain requirements regarding public religious ceremonies.101 The provision under challenge stated that “non-traditional” religious actions could only be performed in public places (that is, not in churches, mosques or other temples, their gardens and graveyards, and believers’ homes) with the prior authorisation of the Ministry of the Interior. In addition, a cumbersome procedure for applying for such authorisation had been set up, including a notice period of at least fifteen days, a description of the type of ritual, disclosure of the identity of those who would perform it, its aim, and when and where it would take place. In contrast, for “traditional” rituals and religious activities, only prior notification to the state administrative body was necessary. The Court found the authorisation requirement to be contrary to the principle of free religious expression (Art. 19(2) of the Constitution), although it chose not to declare the law unconstitutional on the basis of another ground cited by petitioners, namely, that the law treated traditional and non-traditional churches unequally. The question of a right to conscientious objection to military service is an important test case for the understanding of religious freedom: it implicates both the issue of separation of state and religion (as the recognition of religion-specific objections may be seen as making special accommodations of religion in defining the legal rights of citizens) and, of course, free exercise of religion (as nonrecognition of religion-based objections is a burden for a sincere religious believer that affects his ability to comply with the religious precepts). An overwhelming majority of the constitutions of the countries of CEE mention the duty to perform military service, and slightly less than half of these include an explicit guarantee of exemption from such service on the basis of conscientious objection, usually formulated as based on “religious convictions or moral principles”102 or one’s “conscience or religious faith or conviction”.103 In addition, two constitutions proclaim a citizen’s duty to perform military service or an alternative service, without however articulating explicitly the latter possibility as a citizen’s right based on their conscientious objection.104 Further, some constitutions mention the mere possibility of being exempted from military service, whilst stating that when this shall occur will be regulated by law.105 Both the Czech and the Slovak Constitutional Courts have dealt with the issue of conscientious objection. The matter appeared, although not as a central issue, in the decision of the Czech Court,106 arising out of a constitutional complaint against a criminal court decision that sentenced a man to imprisonment for refusing to begin his military service. The basis upon which the criminal court’s decision was held unconstitutional by the Constitutional Court was the constitutional principle of “ne bis in idem”.107 The same person had been already sentenced before for the same offence, the only difference between the two offences being that, the second time, he was asked to join the regiment in a different town. The second conviction was, therefore, overturned by the Court who also examined a second basis upon which the complaint of unconstitutionality was made, namely, the petitioner’s constitutional freedom of conscience. On this, the Court was less favourable towards the complaint. It established that the complainant, a Jehovah’s Witness, rejected not only military service but also the alternative civilian service that was offered. The Constitutional Charter provides for a right not to be forced to perform military
144
CHAPTER 6
service against one’s conscience, and it delegates the task of implementing this guarantee to the legislature.108 The Court held that the legal duty to perform either military or civilian service, as established in the Act on Civilian Service, was not antithetical to the constitutional guarantee of freedom of conscience because to allow people to avoid both of these burdens would violate the principle of equality as such person would gain an unjustified advantage over others. As the Court said: If the Act on Civilian Service provides that, by performing it, a citizen may not obtain any unjustified advantage over those who have performed basic or substitute military service or training, then it must follow logically that if we were to tolerate conduct leading to the avoidance of any sort of service whatsoever, then this would result in the violation of the principle of equality.109
This equality-based argument was not, in the Court’s reasoning, supported by any reference to a specific equality provision of the Charter; instead, a broad intuitive moral notion of equality seemed to be at work here. The Czech Court considered the matter of conscientious objection on two other occasions. In a 1999 decision, it upheld the time limit for lodging a declaration of refusal to perform military service, but struck down, as disproportionate and negating the essence of the right, a strict rule that after the deadline no requests of conversion of military service to an alternative service would be considered.110 It also considered the constitutionality of the article of the Criminal Code that provides for the offence of refusing to perform military service if a person has not requested alternative service,111 and found it to be unconstitutional for contradicting the principle of ne bis in idem. This was due to the fact that the defence authorities were obliged to keep calling up a person until he has performed military service, thus making him vulnerable to consecutive criminal punishments with regard to what is, essentially, one and the same crime. In this case, the Constitutional Court applied the technique of interpreting the law in such a way as to be consistent with the Constitution, whenever possible. The Court found that a constitutionally permissible interpretation of the Criminal Code provisions was possible, and that both the military authorities and the courts should find a way of carrying out their respective legal duties in a way other than conscripting or convicting a person who has already been convicted. The Slovak Constitutional Court considered a case on conscientious objection in 1995.112 It was asked to examine the constitutionality of a provision of the 1992 Statute on Civilian Service which established some conditions under which a person could be exempted from military service. Under challenge was a time limit for the filing of an application to be exempted from military service on the grounds of conscience. The argument was made that this effectively precluded persons from changing their mind after having begun their military service. The petitioners claimed that this was contrary to the constitutional provision on exemptions.113 The Constitutional Court considered the conscience-based exemption from military service to be simply a specific case of the more general right to freedom of religion.114 Indeed, it said, the refusal to carry out military service on the grounds of conscience or religious faith is essentially an external expression of thought, conscience, religion and faith. Hence, the constitutionally approved limitations upon the general right to freedom of religion (i.e. those necessary in order to “protect
CIVIL AND POLITICAL RIGHTS
145
public order, health, morality, or the rights and liberties of others”) apply also to the exemption from military service. Applying this test to the provision under challenge, the Court acknowledged that the time limits for making the exemption application do act as a limitation on the right to freedom of conscience. However, an important public interest is at stake; the government needs to know the correct number of persons carrying out their military service, and this knowledge is needed by a certain point in time. Such a limit aimed at the protection of public order is, according to the Court, “necessary in a democratic society”, and so meets the constitutional criteria of acceptable limits on constitutional rights. The opposite result was reached in a very similar case, decided by the Croatian Constitutional Court.115 At stake was also a provision of the law that established a specific deadline for lodging a request for the conversion of military service into an alternative civilian service (namely, no later than ninety days after the date that a person entered the military register). The Court found this provision to be unconstitutional because, it argued, the constitutionally guaranteed freedom of conscience implies the right to change one’s convictions. Thus, the exercise of the right to conscientious objection cannot be subject to time limits. The Court also struck down another provision of the same statute (the Defence Law) which stated that a conscientious objector who did not make an effort to perform his civilian service well and who violated the disciplinary rules could be ordered to quit civilian service and serve in the armed forces instead. The Court held that the right to conscientious objection couldn’t be linked in such a way to the manner in which civilian service is performed. The question concerning the time period in which a declaration regarding conscientious objection can be validly made has also been raised in a Slovenian Constitutional Court decision in 1995.116 This case related to a provision of the statute on military service which allowed for conscientious objection (therefore, alternative service) at the moment of conscription. The Constitutional Court interpreted the constitutional provisions regarding military service and conscientious objection117 to mean that the right must belong to all soldiers, at all times that they are obliged to take part in the defence of the state. The permanence of this right means that it cannot be limited in time; therefore, soldiers already having started military service, and also those having completed it but remaining on the reserve list and under the obligation to serve in the armed forces in time of war, must also have this opportunity. They must be allowed to change any (further) military service into alternative service. Thus, the law’s restriction of this possibility, to the moment of conscription only, was declared unconstitutional. More specifically, the Court characterised the situation as “an unconstitutional legal gap” because the statute did “not make possible the asserting of the right of conscientious objection to military service also in the case of soldiers in the course of military service and soldiers [who had] completed their military service already”.118 The Court, however, did not abrogate the provision of the law, as that would only make matters worse; it would mean that no soldiers (even at the moment of conscription) would have the ability to choose alternative civil service. Thus, the Constitutional Court simply stated that the legislator must change the law so as to allow all conscripts, at whatever stage of their military duty, to exercise this right.
146
CHAPTER 6 3. THE RIGHT TO PRIVACY
Most of the CEE constitutions include general privacy clauses,119 typical examples being provisions such as: “Inviolability of the person and of privacy is guaranteed”,120 or “the State shall respect and protect private and family life”.121 However, an even higher number of constitutions include more specific protection for the privacy of communication,122 of which this provision from the Czech Charter is representative: “No one may violate the confidentiality of letters or other papers or records, whether privately kept or sent by post or by some other means, except in the cases and in the manner designated by law. The confidentiality of communications sent by telephone, telegraph, or by any other similar devices is guaranteed in the same way.”123 Most of the constitutions of the region phrase the right in this negative way. However, in some cases also positive actions are mandated, such as: “The State shall ensure the privacy of letters, telegrams, other postal dispatches, of telephone conversations and of using other means of communication.”124 The guarantee of data integrity listed in a handful of the constitutions125 can be subsumed within this category of privacy rights. As an example of that right, consider this article from the Croatian Constitution: “Everyone is guaranteed the safety and secrecy of personal data. Without consent from the person concerned, personal data may be collected, processed, and used only under conditions specified by law”.126 Also, the right not to have to state one’s ethnic affiliation,127 or national affiliation,128 could perhaps be placed within this category of rights. In the jurisprudence of the constitutional courts in the region, the concept of privacy has appeared most often in the context of the inviolability of one’s home, and the courts have invariably adopted a pro-privacy stance in these cases. One example is the Bulgarian decision, discussed in greater detail above,129 in which the court struck down most of the clauses of a criminal code that would exculpate the self-defender from the use of excessive force against an assailant but, characteristically, affirmed the validity of such provisions when the assault took place in the attacked person’s home.130 In other words, the fact that the inviolability of the home was at stake in this particular case of self-defence rendered any excessive force used against the assailant tolerable to the Court! Some provisions of tax law have been invalidated by the Constitutional Review Chamber in Estonia on the basis of inconsistency with the constitutionally guaranteed inviolability of the home.131 The Taxation Act of 1993 authorised tax inspectors to “inspect the property, buildings, enterprises etc” without prior warning, to measure and seal containers and equipment, and to place cameras in enterprises. The Chamber determined that the entry into another person’s property without their consent that this provision entailed was unconstitutional; the same also applied to the placing of cameras, insofar as the surveillance is secret. The purpose of tax collection was not found to be among the permissible constitutional grounds for restriction of the inviolability of the home, namely the protection of public order or health, the rights and liberties of others, or the prevention of criminal acts.132 A very similar result was reached by the Constitutional Court of Macedonia; the law provided that a tax official could enter into a taxpayer’s house against his will, if the taxpayer refused to let him in.133 The Court found this inconsistent with the
CIVIL AND POLITICAL RIGHTS
147
constitutional right to inviolability of the home, which can only be restricted for the prevention of crime or the protection of people’s health. The Slovenian Constitutional Court has recently dealt twice with a privacy-based challenge to laws related to crime control and, more specifically, to the surveillance of suspected criminals. Both times it struck down the central provisions of these laws on privacy grounds. In 1997 it invalidated the provisions of the Code of Criminal Procedure that would allow the secret bugging of people in order to catch criminals.134 Although the court order envisaged by the statute authorising such bugging would have had to have shown the necessity of the measures, the Constitutional Court concluded that the law did not differentiate between different methods of bugging and surveillance, some of which are much more intrusive than others. For example, the Court found it improper to apply the same standards of “necessity” to methods as intrusive as “listening in premises”, on one hand, as to “secret observation and following, in which infringement of the communication part of privacy is essentially less intensive, or is entirely non-existent”, on the other.135 According to the Court, the lack of a clear differentiation of different measures and of clear definitions of various surveillance methods led to a lack of legal certainty. For instance, in discussing the distinction between “reasons for suspicion”, “sound reasons for suspicion” and “well-founded suspicion”, the Court noted that the statute did not contain any definitions of these terms, thus implicitly allowing for surveillance based upon the weakest of these grounds.136 The law was found to violate three constitutional rights: the right to privacy, the right to inviolability of dwellings, and the right to the protection of personal data. It is important to state that the Court relied strongly in its opinion upon the rhetoric of breaking with the bad old days when constitutional restrictions upon individual rights were broadly permitted, and when the state and the League of Communists of Yugoslavia made extensive use of such methods as “listening, secret control of telephone and other telecommunication means, secret control of written and other consignments, secret recording and documenting” etc.137 No doubt the evocation of this recent history, from before the collapse of Yugoslavia, added a powerful persuasive force to the Court’s hostility towards the use of such secretive means, even for the perfectly justifiable ends of fighting crime, particularly organised crime. A year later, the same Court struck down certain provisions of the Internal Affairs Act, which actually did what the Court had asked for in this earlier case. That is, the latter Act distinguished between the two types of measures (those more and those less intrusive) that could be used to aid the police in catching suspected criminals, and it provided for the different levels of “necessity” required to be shown in order to make use of these measures.138 However, the Court decided that, within these categories, there were only very general and vague provisions concerning the specific procedures to be followed, and therefore they were not sufficient to prevent arbitrariness and abuse. For example, any acceptable law would have to prescribe the procedure by which recorded conversations, etc. would be dealt with, and, more importantly, it would have to explain exactly what degree of suspicion would be required in order to justify the use of any such measure. In addition to these cases, three privacy-related decisions of the Hungarian Constitutional Court are worth mentioning. On 13 April 1991, the Court declared
148
CHAPTER 6
that uncontrolled collection of personal data was unconstitutional, and it struck down a law introducing personal identification numbers.139 The rejection by the Court of the statute that contained the PIN provisions but whose broader goal was to set up a population register, was based not only on the “right to informational selfdetermination” (as defined by the Constitutional Court, and deduced from Art. 59 of the Constitution, which proclaims, inter alia, “the protection of secrecy in private affairs and personal data”) but also on the right to dignity because “a data user with an undefined scope for data collecting” could become acquainted with the totality of data about a person and thus create a “personality profile”.140 More specifically, the rejection of the PIN provisions was based on the argument that the law did not contain any clear limitations as to the uses of these numbers, and the universal use of such numbers was contrary to “informational self-determination”.141 Compliance with this decision was only partial; in fact, the tax and social security institutions continued to use PINs, and the government resorted to them in its fight against tax fraud and corruption.142 Eventually, the government decided to restore the system of PINs (or, more precisely, to extend the deadline for a temporary use of identification numbers) as part of the “Bokros package” of austerity measures aimed at reducing tax fraud. This law has, however, again been invalidated by the Court.143 A year earlier, the Constitutional Court had held that government regulations compelling Hungarians to declare their assets contravened the right to individual privacy and to the protection of personal information. Also in this case, the government hoped to use compulsory property declarations as a means of combating tax evasion, but the Court found that this measure did not meet the test of being unavoidable and proportionate to the achievable aim.144 The decisions discussed above concerned privacy in the narrow sense of the word, as a “territorial” or “spatial” control by people over their living and working environment, and also over the data and information concerning them. This was therefore privacy in the sense of intrusion upon a person’s seclusion, solitude or private affairs – the first of four meanings of the tort of “privacy” distinguished in the classic article by William Prosser (the other three being public disclosure of embarrassing private facts, false light publicity, and appropriation of a person’s name or likeness).145 However, “privacy” is sometimes understood in a broader sense, as being almost synonymous with a person’s control over his or her own life (and not just over the information about themselves). Hence, it becomes synonymous with individual liberty and is used as an anti-paternalistic ideal. It is in these terms that the issue of privacy can be understood in a decision of the Slovak Constitutional Court which struck down provisions of the Code of Civil Procedure that allowed the attorney-general to enter, if requested to do so by the court or a party to the dispute, into civil proceedings.146 It was the possibility of the public authority’s interference in a civil dispute, regardless of the wish and consent of one party thereto, that the Constitutional Court found to be inconsistent with the right to privacy, defined in the Slovak Constitution as “the right to protection against unwarranted interference in [a person’s] private and family life”.147 Clearly, the concept of privacy that is at work here goes much further than merely protecting a person’s “territorial” control over his or her house, office or over private information; it is a right of non-interference, even if the interference might be motivated
CIVIL AND POLITICAL RIGHTS
149
by the best interests of that person. This shows that the concept of “privacy” sometimes plays an anti-paternalistic role, and is thus synonymous with individual autonomy from unjustified interference. 4. FREEDOM OF MOVEMENT AND THE RIGHT TO CHOICE OF RESIDENCE All of the CEE constitutions contain provisions guaranteeing freedom of movement and the right to choose one’s place of residence. In some of the constitutions, this right is conferred upon everyone; 148 in others, it applies to all persons staying lawfully in the territory of the State;149 or, in others still, it is only enjoyed by citizens of the State.150 Typical factors that can give rise to legitimate limitations of the right include national security, public health, rights of other persons and the facilitation of criminal investigations. Almost all of these constitutions contain a provision guaranteeing the right to freely leave and (re)enter the country.151 As far as the beneficiaries of this right are concerned, in some constitutions, again, only citizens enjoy the right to both enter and leave the country,152 whereas in others everyone is granted the right to leave, with only the right of (re)entry being reserved to the relevant citizenry.153 In an important decision concerning freedom of movement and residence, the Czech Constitutional Court made it clear that the second aspect of the right (freedom of residence)154 applies also to non-citizens, and that any restrictions thereon must meet the standards of proportionality.155 The Court considered a provision of the law156 that stated that a foreign national who has violated any duties laid down by law could be forbidden from residing in the Czech Republic. This law was challenged by foreign nationals who were barred from residing in the Czech Republic for a period of three years, after they had been convicted of selling unmarked goods and impeding customs inspections. In striking down this provision, the Constitutional Court announced that freedom of movement and residence is guaranteed also to foreign nationals, if they have a residence permit. While this right can be limited by statute in the interest of, amongst other things, public order, in this case the restriction of the right was disproportionate. The negative impact of the restriction on the constitutional right outweighed the benefit to the public interest. In addition, the restriction was, according to the Court, impermissibly vague, and the explanation by the representatives of Parliament, that the administrative bodies that apply the law could reduce the vagueness in this legal provision, was not accepted. The specific limitations must be precisely established in the statute itself, in order to make its legal consequences foreseeable to the general public. The Estonian Constitutional Review Chamber examined, in 1998, a law on police service that would make it possible to transfer a police officer, without his consent, to another position, even if this resulted in a change of his residence.157 The Chamber acknowledged that a police officer must be at the disposal of his superiors for the performance of urgent police tasks, even in his rest time; therefore, he must live near his place of work. However, Article 34 of the Constitution states: “All persons legally present in Estonia shall have the right to freedom of movement and
150
CHAPTER 6
choice of residence”. Article 34 goes on to allow for restriction of the right to freedom of movement; however, no permissible restrictions on the right to freedom of choice of residence are envisaged in this article. Freedom of choice of residence can only be restricted, under Chapter 10 of the Constitution, for reasons of national defence, and this can apply only to members of the armed forces, not to the police service. Thus, the Chamber concluded that this section of the law was in conflict with Art. 34 of the Constitution. The Court further remarked that the involuntary transfer of a police officer “may also prejudice police officers’ and their family members’ right to family life”.158 One of the few decisions in which the Romanian Constitutional Court has struck down a parliamentary bill was based upon the constitutional right to freedom of movement.159 This was related to the parliamentary approval of a government ordinance160 on the imposition of a tax for passing the frontier, with a view to generating some resources that would be destined for social protection. As the Court noted, the law clearly created an impediment to the exercise of the constitutional right of free movement;161 thus, in order to be valid, it must conform to the constitutional principles controlling the acceptability of limitations on rights.162 This was not the first time that the Constitutional Court of Romania had faced a similar issue. An earlier Court decision relating to the constitutionality of a tax for passing the frontier with a view to funding tenement-heating assistance, during November 1993 – April 1994, was held to be in conformity with the Constitution.163 This was based on the exceptional character of the situation: as a lawyer working for the government at the time the case went before the Constitutional Court has later explained, the government had been desperately searching for any sources of budgetary income, in order to subsidize heating for the poorest in what turned out to be a particularly harsh winter in 1994.164 At this time the tax, which was expected to be in operation for only two months, seemed like a good idea and met with hardly any protest. This encouraged the government to try to win parliamentary endorsement of a more permanent rule that would permit the government, in the case of harsh winter conditions, to enforce a similar taxation measure. This quasi-permanent arrangement was rejected by the Court. In the process of quashing this bill, the Constitutional Court established a test for when such a tax would be constitutionally acceptable. To pass constitutional muster, it must be only: (a) introduced exceptionally, and (b) imposed “to protect a certain right of social protection belonging to citizens”. It must be the case that, without this limitation of one constitutional right, the enjoyment of another would be gravely restricted. In other words, it cannot be a general protection of other vague “rights”; it must be very specific, such as to provide heating in tenements at a certain specified time. Only then would it comply with the constitutional demand that “The restriction [on a constitutional right] shall be proportional to the extent of the situation that determined it and may not infringe upon the existence of the respective right or freedom”.165 Using this test, the Court found the law in question to be constitutionally defective. It was not specific at all; rather, it simply claimed to enhance “social protection rights” generally. Thus, there could be no evaluation of what particular right this measure would aid, or of the extent to which the restriction of
CIVIL AND POLITICAL RIGHTS
151
freedom of movement would be proportional to the proclaimed rights-protective aim. The Constitutional Court of Russia has also dealt with the issue of freedom of movement. In a 1998 decision, the Court deemed unconstitutional a provision that stated that passports could only be issued at the applicant’s registered place of residence.166 The proceedings in this case were initiated by a Russian citizen who had been registered in Tbilisi, Georgia and who had no place of permanent residence in Moscow, where he applied for a passport. He argued that he had been discriminated against, and prevented from enjoying the constitutional right to freedom of movement. The Court agreed with this point of view; it argued that, if the procedure of applying in one’s place of permanent residence cannot be followed for whatever reason, as in the case of forced migrants and persons with no officially registered address, then their right to obtain a passport – and, in consequence, to travel – is violated. The law was therefore struck down, based both on the constitutional right to equality (Art. 19) and on the right to leave the country (Art. 27). In this context, the struggle of the Russian Constitutional Court against the infamous system of “propiska” should also be mentioned. This is the term for the old Soviet system of residence registration, which effectively rendered it impossible or extremely difficult for citizens to move to the place of their choice. A person without a propiska could not find accommodation or a job, and could be arrested by the police and fined. This remains one of the most invidious legacies of the old Soviet days, particularly immune to change, perhaps because it is such a convenient device of corruption and the abuse of power. As a Russian scholar recently observed: “The fact that propiska still exists and is protected by the Russian government and regional leaders starkly reveals the disturbing reality behind its legal reforms – Russia with its new Constitution has once again created an elaborately decorated showcase for rights and freedoms which is empty inside”.167 Although, in its most drastic form, “propiska” has been abolished, its remnants persist in the form of high registration fees in the capital, for non-Moscow residents. Propiska is no longer a privilege that the authorities can confer or refuse to citizens at their own discretion, but has been transformed into an instrument of economic discrimination against out-of-towners. The Constitutional Court has found the registration laws to be unconstitutional on several occasions,168 and in these cases met with non-compliance, most recently by Moscow authorities. Although the Moscow Duma’s law was declared unconstitutional by the Court on 2 February 1998, it remained unchanged until at least late 2001.169 5. CITIZENSHIP AND VOTING RIGHTS Almost all constitutions of CEE contain some provisions that seem to guarantee the equality of constitutional rights and freedoms to citizens and non-citizens alike, thus calling into question the importance of citizenship status for the enjoyment of these rights and freedoms.170 However, I have deliberately used the words “seem to guarantee”, as this guarantee is often undermined by allowing statutory exceptions to the principle that rights apply to all human beings. As an example, consider the
152
CHAPTER 6
Constitution of Poland, which states: “Anyone under the authority of the Polish State, shall enjoy the freedoms and rights ensured by the Constitution”,171 but adds immediately that “[e]xemptions from this principle with respect to foreigners shall be specified by statute”.172 While the generality of rights as applying to everyone clearly has the status of a norm, and the restriction of certain rights to citizens only is a departure from this norm, the fact that ordinary laws can establish such exceptions is a potentially dangerous carte blanche to give to the legislator. A different construction has been adopted in those constitutions that do not contain a general proviso of the sort mentioned in the previous paragraph, but which instead frame some rights generally and others more restrictively in reference only to citizens. Thus, it should be assumed (in the absence of an article such as that present in the constitutions already described above) that the more general rights must apply to all people on the state’s territory, and that statutes cannot restrict the applicability of a more general right to citizens alone. This, for instance, is the case with the Lithuanian Constitution which makes no mention at all of who is the subject of the rights. Some provisions are phrased generally (e.g. “The right to life of individuals shall be protected by law”),173 and others as pertaining only to citizens (“Citizens may move and choose their place of residence in Lithuania freely, and may leave Lithuania at their own will”).174 Finally, there is a group of constitutions that frame the equality of rights of citizens and non-citizens explicitly, without any authorisation of statutory exceptions.175 These are therefore, on the face of it, the most rights-protective for non-citizens: it is clear that only those specific constitutional rights that are explicitly confined to citizens have a limited applicability, and also that statutes cannot make any inroads into any other rights. It is interesting to consider which rights are generally considered as being confined to citizens only, as opposed to those of universal applicability. Naturally, in all the constitutions of CEE, the right to vote is given to citizens only. Further, whenever there is an explicit right of free entry into the state,176 or one granting protection from expulsion or extradition, these are without exception restricted to citizens. Two other rights are uniformly confined to citizens whenever they appear in constitutions: the right to publicly criticise state authorities,177 and the right to adequate housing.178 In addition, in relation to duties, all those constitutions that proclaim the duty to perform military service understandably restrict this solely to citizens. Beyond these most obvious cases where there is a uniform attribution of a right (whenever such a right is mentioned) to citizens, there is a certain variability of the range of people to whom the other rights are attributed. For example, among those constitutions that contain the right to political participation, some restrict this to citizens179, while others do not.180 Similarly, other traditional political rights, such as the right of assembly and of association, are so restricted in roughly half of the constitutions that list them.181 Furthermore, the rights to property and freedom of movement182 are citizen-only rights in roughly half of the constitutions that mention them. There is no uniformity among socio-economic rights. Generally, provisions for aid to the weak in society (such as the old, families where the bread-winner has died, those unable to work, those unemployed through no fault of their own etc.) are
CIVIL AND POLITICAL RIGHTS
153
restricted in scope to citizens only; this is the case in nine of the eleven states in which such a right is acknowledged.183 Health care provided by the state is likewise, though less often, restricted to citizens; five of the eleven constitutions granting this right limit it in this manner.184 Also, five of the thirteen constitutions granting free education do so for citizens only185 (some grant no right to education at all to noncitizens,186 while others provide a general right to education, but only guarantee free education to citizens.)187 The right to own land is a special case, considering the widespread fear of a “sell-out of the country”, so deeply ingrained in many CEE states; hence, a number of the constitutions explicitly exclude foreigners from the right to acquire land.188 If one were to draw up a continuum of constitutions, ranging from those that are the most, to those that are the least “generous” in conferring rights on non-citizens, then it is likely that nearest the generous end of the spectrum would come the Albanian and Hungarian Constitutions, which only reserve the rights to vote, to nonexpulsion and to some forms of social security to their citizens. The spectrum moves on to include a broader selection of “citizen-only” rights, with many countries adding their own, idiosyncratic ideas. Thus, the Croatian Constitution, with an already high number of rights restricted in this way, adds, somewhat bizarrely: “The Republic ensures citizens the right to a healthy environment.”189 The Montenegrin Constitution includes an unusual right and therefore an unusual restriction: “Citizens shall have the right to participate in regional and international non-governmental organisations. Citizens shall have the right to address international institutions for the purpose of protection of their freedoms and rights guaranteed under the Constitution”.190 The most paradigmatic civil right, the right to vote, is phrased very similarly in all191 of the constitutions. Admittedly, some offer minor exceptions, such as for those “placed under judicial interdiction or serving a prison sentence”,192 or those “declared legally incapable by court”.193 These exceptions apart, however, the right remains similar in all constitutions of the region. The right to vote, and the accompanying principles regarding democratic elections, has triggered a number of constitutional court cases in which the courts had an opportunity to spell out their visions of democracy and, in some cases, replace the visions of the legislature with their own substantive conception. Perhaps the most striking examples of such a displacement of a legislative choice regarding a very fundamental feature of the democratic system was the decision of the Slovak Constitution Court: the Court rendered impossible the introduction of a quota system, whereby ethnic minorities would have guaranteed representation in local self-government authorities.194 The Court found such a system of representation, which would mirror the ethnic composition of a local community, to be contrary to the principle of equal rights to elect one’s representatives. The Court’s individualist interpretation of political equality is one, certainly respectable, understanding of the democratic, “one-person-one-vote” principle, but it is not the only possible one. Reasonable persons may espouse conflicting interpretations of this principle, and the fact that the Court made its own vision prevail over that of the legislator raises some thorny questions concerning its democratic legitimacy in shaping the instruments of democracy itself.
154
CHAPTER 6
Other decisions in the region cover a broad range of aspects of the electoral system that gave rise to challenges based on democratic principles, differently understood and interpreted. One issue that raised its head in the constitutional courts was that of the setting of a threshold level of support necessary for a political party to gain seats in parliament. In a case before the Czech Constitutional Court195 it was claimed that a five percent threshold violates two constitutional principles: firstly, that of proportional representation, as it disturbs the correspondence between the number of votes given for particular parties and the allocation of seats in the parliament; and secondly, the rule that all have equal access to electoral office, because some votes, all those that for parties that failed to reach the five percent threshold, would in fact benefit candidates to whom they were not given. The Court rejected both of these arguments. In reference to the first argument it explained that certain measures (such as thresholds) are introduced in many countries in order to prevent splintering, and cannot be seen as contradicting the principle of proportional representation. As to the equality of each vote, the Court stated that as long as voters know in advance who may be allocated the seats should their preferred parties not meet the five-percent requirement, the principles of the equal power of every vote and of direct voting, are not contradicted. However, the bottom-line argument was that of political expediency, which requires some departures from the principles of proportionality and equality strictly understood; whilst elections aim to express voters’ preferences as accurately as possible, they also have the aim of creating a parliament that has the capacity to adopt decisions. Strict proportional representation would lead to over-fragmentation, and perhaps the impossibility of forming a majority. Therefore, the Court concluded, it was reasonable to incorporate integrative factors. It shared the lawmakers’ view about the right balance between the principle of proportionality and that of the cohesion of parliament. It added that any increase in the threshold might not be unlimited, but that five percent was considered reasonable. Similarly, an electoral threshold was upheld against constitutional challenge in Romania. The threshold was three percent for a political party, with a variable, sliding-scale threshold for coalitions (with one percent being added for each political party, up to a maximum of eight percent).196 As for the threshold in general, the Constitutional Court explained, similarly to its Czech counterpart, that the electoral process is not only a means to allow for the exercise of the individual rights of citizens but that it also represents a method of establishing democratic institutions with a coherent capacity for expression, and of creating efficient centres of political decision making. Too many small parties in the parliament would hamper the democratic process, and thus a threshold is reasonable and constitutional. Concerning the thresholds for coalitions of parties, the Court referred to the constitutional provisions on pluralism197 and on the functions of political parties.198 From this latter provision, the Court derived (arguably, from a somewhat dubious inference) the principle of equality of opportunity of all political parties. However, the Court stated that the position of parties is weaker than the position of coalitions; hence, it is not unreasonable to require coalitions to face higher electoral thresholds than single parties do.
CIVIL AND POLITICAL RIGHTS
155
Another structural feature of the electoral system, the so-called “national list”, was challenged, and eventually upheld, by the Constitutional Court in Slovenia.199 One of the results of the system was that the possibility existed that a person high on the national list set up by a party could be elected even though fewer people in the district had voted for him or her than for an unsuccessful candidate. According to the challengers, this violated the principles of democracy, directness of elections, and equality of voting rights; the Court, however, disagreed. According to the Court, the creation of a national list did not frustrate the principle of equality of voting rights, because no voter’s vote has a lower weight than another’s; it may be taken into account either at the district, or at the national level. As for the democratic principle, it is not violated by the fact that some candidates with poorer election results may be elected over those on the same list, even if the latter had better results. The Court’s view was that the voters vote primarily for the parties, not for individual candidates; therefore, the greater success of a particular candidate is mainly due to the success of the list. Thus, as the vote is primarily for the list, it is not unfair if it assists a candidate other than the one whose name is actually circled on the ballot paper. Finally, the national list system was held not to violate the principle of directness of voting, because the structure of the list is known in advance to the voters. This simply gives the party the opportunity to shape the composition of elected parliamentary representatives. In order to strengthen this feature, the Court required the lawmaker to amend the law by prohibiting any last-minute changes in the composition of the national list. Subject to this minor correction, the whole structure of the national list method was found to be in compliance with the constitutional principles of the electoral system. Various legislative limitations upon passive electoral rights – the right to stand in elections – have been subject to the critical scrutiny of constitutional courts in the region. In 1994, the Hungarian Constitutional Court struck down a proposed amendment to the electoral law which stated that elected representatives of the so-called “social security self-governments” could not be put forward as candidates at the parliamentary elections.200 As the holding of these two offices was not one of those listed as “incompatible” in the Constitution, the amendment was considered to be contrary to the right of all citizens to be elected. Similarly, the Macedonian Constitutional Court invalidated an analogous exclusion from election to local councils or mayoral office, addressed to members of the armed forces, police officers and the officers of the ministry of internal affairs, and the intelligence agency.201 The Ukrainian Constitutional Court, in a wide-ranging decision that struck down a number of provisions of the 1997 election law, invalidated the exclusion of a number of categories of people (judges, public prosecutors and state employees, etc) from running in elections. According to the statute, they could register as candidates only on the condition that they promised to terminate their employment whilst campaigning.202 This decision also struck down a provision concerning the active electoral right – the right to vote – that withheld the right from incarcerated in prisons. As the Court stated, the Constitution envisaged that only those declared by a court to be incompetent may not vote; this does not automatically encompass all those sentenced to imprisonment.
156
CHAPTER 6
One example of a case in which a particular limitation on passive electoral rights was upheld by a constitutional court is afforded by a Romanian decision. In this case, the Constitutional Court approved a statutory ban on the nomination for election to the parliament of persons who had acted as prefects and heads of public services in the given territorial unit, within the six months immediately prior to the election date.203 Such exclusion is not listed in the Constitution, and yet the Court found that it did not contradict constitutionally guaranteed voting rights. The Court looked at the constitutional list of permissible statutory restrictions of rights (“the defence of national security, public order, health or morals”)204 and concluded that this restriction was necessary to protect public morals. The goal of the restriction, as interpreted by the Court, was to prevent the nomination of persons whose position meant that they could have an important influence on electoral opinions. This, one may object, is precisely what should motivate a person to stand in parliamentary elections! The Court's reasoning seems, therefore, a convoluted way of defending a rule that is indefensible from the point of view of the plain constitutional text. To be fair to the law, and to the Court’s decision, it should be added that the law did not prevent those persons running in a constituency other than that in which they had held those positions. 6. FREEDOM OF PETITION, ASSEMBLY AND ASSOCIATION The right of petition is present in all but four of the constitutions of the region.205 It is formulated in a number of different ways. At one end of the scale is the bare right to simply be able to lodge a petition.206 Other constitutions then expand this to include the right to receive an answer207 or, more frequently, both this and the right not to suffer negative repercussions from having put forward the petition, especially when the contents thereof are explicitly linked to criticism of the State or its agencies.208 The Romanian Constitution has a more detailed provision for petition rights, saying that not only can one petition and have the right to an answer, but also that “legally established organisations have the right to forward petitions exclusively on behalf of the collective body they represent”209 and that “the exercise of the right to petition shall be tax-exempted”.210 In turn, the Czech and Slovak Constitutions limit the right to petition, by saying that such petitions “must not be used to call for the violation of basic rights and liberties”211 and “must not interfere with the independence of a court”.212 The right to assembly is also proclaimed in all of the constitutions. Its basic component is phrased in almost the same way everywhere: “All citizens are granted the right to peaceful assembly and protest.”213 After this, the constitutions differ in the way in which this right is to be put into practice. Some state that no prior permission is required from the authorities for assemblies,214 whereas others assert that there must at least be prior notification to the authorities of any public meetings.215 A big distinction in the effect and scope of the right exists due to the exceptions that are made to it. This ranges between mild limitations, such as restrictions during a national emergency or war,216 and very broad limitation clauses, permitting statutory restrictions “that are necessary in a democratic society for the
CIVIL AND POLITICAL RIGHTS
157
protection of the rights and freedoms of others, public order, health, morals, property, or the security of the state”.217 Likewise, the general right to association exists in basically the same form in all of the constitutions. Two constitutions seem to depart from the general wording, by proclaiming the right to associate for certain stated purposes only.218 However, this is merely a semantic difference, the purposes being stated so broadly and vaguely as to cover basically anything. The only real additions to this right in some constitutions are that one cannot be forced to join an association,219 and that the state will provide positive assistance to certain types of associations.220 Once again, the difference comes with the way in which valid grounds for potential exceptions to the right are phrased. Some limit the right should it threaten the constitutional order, be promoting an unconstitutional aim, or threatening the sovereignty of the state.221 Others preclude associations that attempt to induce national and ethnic hatred or unrest,222 or those that establish clandestine or paramilitary structures.223 Perhaps the broadest mandate for restricting the right to association is in the Polish Constitution, which provides simply: “Associations whose purposes or activities are contrary to the Constitution or statutes shall be prohibited”.224 The right to association was subject to interpretation by the Estonian Constitutional Review Chamber (CRC) in 1996, as a result of a petition by the President, whose initial veto to the law had been overridden by Parliament. The Chamber struck down as unconstitutional the Non-Profit Associations Act of 1996.225 One ground of invalidation was the right of children to associate. The Act confined this right to associate to “capable natural or legal persons only”, which excluded minors, while the Constitution gives this right to “everyone”, hence, the CRC concluded, also to minors.226 The second questionable provision of the Act was one that stated that a member may be excluded from an association, regardless of what had been provided for in the statutes of that association, if he failed to comply with the rules of the association, caused substantive damage to the association, or “upon other weighty reasons”. The Chamber determined that this contravened Article 14 of the Constitution which provides that it is the duty of the legislature to guarantee rights and freedoms. Under the Act, the legislature could not discharge this duty, because the regulation about the possibility of exclusion was “inconsistent with the nature of the right [to association] in a democratic society”. In addition, the Chamber found one other faulty provision, even though it had not been challenged in the presidential petition to the Chamber. The Act provided that an association that possesses arms or performs military exercises may be formed only on the basis of specific legislation; however, the Constitution provided that the conditions for the establishment of such association shall be provided by law.227 In other words, according to the Constitution, all that is necessary to form such an association is a single administrative act, while the statute implied that, e.g., each sports association practicing target shooting should be authorised by Parliament with a pertinent law. As the CRC declared: “A law as a legal act of general character can not replace a permission as a single administrative act”. It is interesting that the CRC preceded this section of its scrutiny by stating that it found it necessary to go beyond the bounds of the petition on the following grounds: its own function “as the executive of the judicial power”; the public-law nature of constitutional review; the
158
CHAPTER 6
principle of autonomy of administration of justice; and the requirement of the integrity of legal analysis. A case on the right not to associate, as a corollary of freedom of association, was considered by the Constitutional Court of Lithuania in 2000.228 Under challenge were provisions of the Law on Associations of Apartment House Owners, which stated that all persons who become owners of a house should enter the association of apartment house owners, and that this membership shall terminate when they lose their ownership rights. The Constitutional Court held these provisions to be contrary to the constitutional right to freedom of association, which also proclaims that no one can be forced to belong to any society or association.229 The provision under challenge did not give owners a free choice about entry into the association, therefore was clearly contrary to the relevant article of the Constitution. As the Court stated, “the Constitution guarantees the right to decide, of one’s own free will, whether to belong or not to belong to a certain society, political party or association”.230 It should be added, however, that the Court’s decision was remarkable in that it considered very seriously the arguments of the defenders of the statute, even though, at first blush, the conflict between the automatic membership rule and the constitutional principle of free association seems evident. Despite this, the statutory rule was defended (by the representative of Parliament) on the basis that ownership in apartment blocks gives rise to a number of obligations (for example, to take care of the communal spaces), and that the execution of these obligations is best implemented by an association of all the owners. As to free will, the representative of the Parliament argued that “a person has the right to choose as to what real property to acquire”;231 once this choice has been freely made, certain consequences simply follow. In its decision, the Court put quite a strong emphasis on the obligations following from “the social function[s] of ownership”, and conceded that public interest dictates that certain obligations stemming from common ownership of communal spaces must be regulated. It conceded also that “[t]he establishment of the association of owners is one of possible ways to implement the right of common shared ownership”.232 However, it held that the principle that joining (and leaving) associations must be voluntary need not be compromised in order to attain this goal. The point was to not allow the unwillingness of an individual owner to join an association to hinder the implementation of common obligations;233 the Court merely implied that this end could be achieved without forced membership in an association of owners. 7. FREEDOM OF EXPRESSION Even though Communist constitutions in CEE paid lip service to freedom of speech and freedom of the press,234 it is obvious that these rights were among those most widely abused. The Communist parties exercised their control over the societies largely through their tight grip on what could be said, read, listened to and taught. Small wonder, then, that the demands for guarantees of freedom of expression, especially freedom of the media, were amongst the main claims of the democratic
CIVIL AND POLITICAL RIGHTS
159
groups and movements in the region, and that these rights have been given special prominence in the new constitutional charters of rights. In all of the CEE constitutions, there are extensive provisions on freedom of expression. Often freedom of the press is proclaimed separately; either as a separate right to that of freedom of expression,235 or as a general principle underlying the political system.236 In some constitutions, the possible restrictions on freedom of expression are closely modelled on Article 10 of the ECHR.237 Often there are express prohibitions on censorship,238 which is understandable, taking into account the bad memories of state socialism. Constitutional provisions on freedom of expression and freedom of the press differ in various CEE states when it comes to the extent of “positive” provisions about the right of access to the media, and the constitutional guarantees of media pluralism. In some constitutions, there are clauses expressly providing for free access to the media. For example, the Romanian Constitution promises that “freedom of the press also involves the free setting up of publications”,239 and also provides that public broadcasters “must guarantee, for any important social and political group, the exercise of the right to be on the air…”.240 These types of provision (free access, prohibition of monopoly) are rather exceptional. However, even if they are not textually present in other constitutions, they have, at times, been established by constitutional courts. For instance, the Lithuanian Constitutional Court, in its decision (described below) on the powers of the Governing Board of state radio and TV to confirm the allocation of private broadcasting facilities, emphasised that it was not only freedom of information but also freedom and diversity of the mass media that needed protection. Hence, it demanded institutional guarantees for pluralism by prohibiting the “monopolisation of means of mass media”.241 Perhaps the most significant interpretation of the general right to freedom of expression given so far in CEE constitutional jurisprudence was in the ruling of the Bulgarian Constitutional Court, issued as a response to a request by President Zhelev in 1996.242 It concerned the binding interpretation of three related constitutional provisions: on freedom of expression,243 freedom of the press,244 and freedom of access to information.245 The Court declared that these three rights are systematically and functionally related, and that they can only be restricted in exceptional circumstances, as provided in the Constitution. These circumstances include cases in which expression is used to the detriment of the rights and reputation of others, for the incitement of a forcible change of the constitutional order, for the perpetration of crime, or for the incitement of hostility or violence against others. The Court further explained that, of these categories of legitimate grounds for restriction, the most important is the protection of the rights, reputation and dignity of other persons, but that this must not be construed so as to restrict public (as opposed to personal) criticism of politicians, public officials and governmental institutions. Concerning the restriction of speech instigating hostility, this ruling of the Bulgarian Constitutional Court stated that any valid limitations falling under this category must be based on the values of mutual respect, but must not deny the possibility of expressing diverse and opposing opinions, which is the essence of the notion of freedom of expression. Regarding freedom of the press, the Court explained that the express constitutional prohibition of censorship includes both
160
CHAPTER 6
formal and informal means of governmental interference with the media. While various organisational and financial aspects of the broadcasting media can be regulated by law, these regulations must guarantee the independence of that media. Therefore, the supervisory and governing bodies of public media must be set up in a way that frustrates any attempts at undue interference by governmental institutions. The only acceptable way of direct interference in media activities is through the judicial system, which may act only in order to protect the goals defined by the Constitution. Overall, this is a very libertarian position and, although, its specific implications are yet to be discovered as the decision was an abstract interpretation of the law, it certainly lays out a libertarian philosophy of freedom of speech in an enlightened and helpful manner. I will now move on to focus on five issues that have turned out to be particularly controversial and that have given rise to important court rulings: the question of the legal status of “hate speech”, the protection of religious sensibilities, the defamation of high officials, the freedom of the media, and the regulation of commercial speech. To begin with, all CEE legal systems contain restrictions upon (variously defined) incitement to hatred on racial and ethnic grounds. In this willingness to control racist or other hate speech, CEE countries are in line with the European tradition (and contrasting with that of the United States) of readiness to restrict speech in the interests of social peace, or the protection of the dignity of the victims of racism. This is consistent with the International Convention on the Elimination of all Forms of Racial Discrimination of 1969, to which all of the CEE countries are party; and none lodged any reservation regarding Art. 4(a) thereof, which obliges States to declare, as an offence punishable by law, all dissemination of and incitement to racial hatred and discrimination.246 In some countries, such restrictions on freedom of speech are pronounced constitutionally. For example, the Romanian Constitution prohibits, amongst other things, “instigation of … national … hatred” and any “incitement to discrimination, territorial separatism…” etc247, and the Bulgarian Constitution provides that freedom of expression cannot be used for “the incitement of enmity or violence against anyone”,248 etc. Significantly, the basic propriety of such restrictions, whether constitutional or statutory, has never been subject to any fundamental challenge in CEE, and the responses of constitutional courts in those cases in which such laws have been questioned, have been by and large supportive of the idea of such restrictions. As an example, consider the 1992 decision of the Czech Constitutional Court, which examined the Criminal Code’s prohibitions on hate speech.249 At issue were two articles: one prohibiting support for a political or social movement aimed at the suppression of the rights and freedoms of citizens, or a movement declaring national, racial, class or religious hatred (Art. 260), and the other prohibiting public expressions of sympathy for the movements described in Art. 260 (Art. 261). In principle, the Court found no conflict between these criminal prohibitions and the constitutional principle of freedom of expression.250 This was because they complied with the constitutional grounds for restrictions of the freedom of expression, namely being necessary for the protection of the rights and freedoms of others, for the security of the state, or for public safety. In particular, the Court rejected the argument that, by prohibiting propaganda for these movements, the State commits
CIVIL AND POLITICAL RIGHTS
161
itself to an official ideological orthodoxy: “By no means is it the case that only a certain ‘exclusive’ ideology remains permitted merely due to the fact that the law criminalizes the support for and propagation of ideologies . . . demonstrably directed at the suppression of civil rights”.251 The only constitutional defect that the Court found was in the second of the challenged articles. In a newly amended version of the criminal prohibition in question, there was a bracketed reference to “fascist or communist movements”. The Court said that it could have been interpreted in two ways. The first, unobjectionable, interpretation would be that these two types of movements were just examples of some, among presumably many other, movements for which expressions of sympathy were illegal, insofar as these movements met the substantive criteria listed in Art. 260 (the movement being aimed at a suppression of the rights and freedoms of others, declaring racial hatred, etc). However, the bracketed reference was open also to a second interpretation: that a public expression of sympathy for communism or fascism is illegal, regardless of whether the movement supported by the person propagates racial or other forms of hatred. The Court held that the second interpretation ran counter to the constitutional protection of freedom of expression. For instance, teaching about the ideal of a classless society (as long as no violent means for bringing it about are propagated) is not criminal. As such an ambiguity of interpretation is inconsistent with the principle of legal certainty, which the Court read into the general constitutional proclamation that the Republic is a law-based state, the bracketed reference to “fascist or communist movements” was deemed unconstitutional. The prohibition of incitement to hatred was the basis of the first important freedom-of-expression decision of the Hungarian Constitutional Court; the one that, as Chief Justice Sólyom later boasted, “opened the ‘Hungarian First Amendment jurisdiction’ [sic] of the Constitutional Court, laying down at the start a liberal, extensive interpretation of the right to the freedom of expression”.252 In this decision, the Court struck down Article 269(2) of the Penal Code (the offence of denigration of the Hungarian nation), while upholding the constitutionality of Article 269(1) (the offence of incitement to hatred).253 As to the former, we are told by one critical commentator that the invalidation met with criticism in some circles and “remains highly controversial both in the legal-political community and in society at large. . . . Public criticism of the [invalidation of Article 269 (2)] is based on the assumption that democracy is not yet well enough developed and that extremist speech may therefore have real consequences and even endanger the democratic order”.254 The Court stated, among other things, that “[f]or the maintenance of public peace the application of penal sanctions for the public utterance, or similar conduct, offending, disparaging or denigrating the Hungarian nation, other nationalities, peoples, creed or race is nor unavoidably necessary”.255 However, on a closer reading, the decision appears overall more libertarian in its rhetoric than in its actual argument, as far as the reasoning regarding incitement to hatred is concerned. At the level of rhetoric, the Court (in the opinion written by Justices Sólyom and Szabo) presented an impressive array of constructions supporting a very robust conception of freedom of expression: it is granted “a special place among the constitutional rights”, 256 it is described as a condition “of a truly vibrant society capable of development”,257 the laws restricting that right must
162
CHAPTER 6
be strictly construed,258 and it should be interpreted in a content-neutral manner, the right protecting opinions “irrespective of the value or veracity of [their] content”.259 This is an expansive, strongly libertarian theoretical construction. Yet, in the part of the decision that supports the constitutionality of a prohibition of incitement to hatred, one is struck by how one-sided, pro-restriction the argument is. Obviously, the incitement to hatred is an extremely controversial matter, which generates a strong clash of values; the value of freedom of public speech (including speech hurtful to some groups, and having the potential to promote hostility towards those groups) and the value of maintaining equal dignity and social peace among the members of different groups. The opinion by Sólyom and Szabo emphasizes the latter value, but gives very short shrift to the former. While the opinion expressly cites the “clear and present danger” test for such legitimately restricted incitement,260 the test merely plays an ornamental role, because there is no hint that the prohibition in Art. 269 (1) is activated only when the incitement to hatred is likely to lead to violence or discrimination. Rather, the very fact of such incitement is seen as violating the rights of the victims of such speech, whether actual consequences in the form of violence or discrimination are likely to follow or not. As some commentators have noted, under the provision upheld by the Court “[i]t is not necessary . . . that the incitement to racial hatred result in any clear and present danger”.261 It is one thing to incite to “hatred”, and another to incite to “violence”, and the distinction marks an important point at which the speech in question ceases being merely a carrier of ideas and becomes dangerous conduct. However, this distinction is not evident in the judgment; indeed, incitement to hatred is characterised as “the emotional preparation for the use of violence” and, as such, “an abuse of the freedom of expression”.262 This was not the end of judicial scrutiny of anti-hate speech prohibition in Hungary. After Art. 269 (2) was declared invalid by the Court as described above, the parliament modified the surviving art. 269 (1), that is, the provision on incitement to hatred, by adding a provision prohibiting any act “capable of triggering incitement to hatred”. The Court invalidated this modification263 and, in effect, art. 269 (1) was reduced to a prohibition on incitement to hatred against the Hungarian nation or against any national, ethnic, racial or religious group, taking place in front of a large audience. Following this decision, there was no attempt, on the part of the parliament, to reintroduce a wider prohibition, perhaps because it was not in the interest of the then ruling Young Democrats to alienate the extreme right-wing, xenophobic Hungarian Justice and Life Party, which had entered the parliament in 1998. Although this party was not formally a member of the governing coalition, they supported the government in most cases.264 In May 2000, the Hungarian Court considered a challenge to Art. 269(B) of the Criminal Code, which punished the public display of a swastika, an Arrow Cross (the symbol of the Hungarian fascists during World War Two), or of traditional Communist symbols (a hammer and sickle, or a red star).265 The use of these symbols for educational, scientific, artistic, historical, or informational purposes was specifically exempted from the ban. The petitioners argued that the provision violated freedom of speech266 and that the list of symbols was arbitrary and, therefore, violated the constitutional prohibition against discrimination.267 The Court
CIVIL AND POLITICAL RIGHTS
163
placed these provisions within the framework of crimes against public peace, and argued that these symbols can induce fear in those who have suffered under past authoritarian regimes. A prohibition on such symbols is, therefore, a necessary and proportionate limitation on freedom of speech. Regarding the charge of the discriminatory nature of these prohibitions, the Court retorted that the list of prohibited symbols was not discriminatory, since the law did not discriminate against any persons on any grounds, it was merely concerned with the symbols. As one can see, the Court took a different approach from that of the United States Supreme Court’s ruling on a similar subject. In a 1992 decision, the Supreme Court struck down a city ordinance that made it a misdemeanour for anyone to place on public or private property a symbol that might arouse anger or alarm and, to provide an example, a number of such prohibited symbols were listed, including the burning cross and swastikas.268 The reason for the invalidation was that the prohibition was found “viewpoint based”, in that it selected some grounds for “anger or alarm” for special protection: race, colour, creed, religion and gender. The Supreme Court found that the ordinance picked for special (and thus, discriminatory) treatment only some “distinctive ideas”, conveyed by “distinctive messages”. In a separate decision, the Hungarian Court also upheld the criminalisation of the desecration of national symbols (the national anthem, flag and coat-of-arms of Hungary), as provided for in Art. 269(A) of the Criminal Code.269 In its decision, the Court stated that the legal protection of national symbols was the general practice in Europe, and that it may be particularly necessary in states undergoing the transition to democracy. These national symbols, the Court emphasised, are linked in Hungary to the achievements of democratic transition, and embody the departure from communism. The protection of national symbols cannot be achieved in any other way, hence, the restrictions meet the requirements of constitutionality. However, it is clear that the relationship between the means and the purpose of protecting constitutional values is illusory here. If the value lies in the communicative nature of a symbol, then a ban on desecration of the symbol is, by definition, a necessary means of protecting this value. The real question should be whether the state should protect the values of symbols that hinge upon their purely communicative impact, and prohibit behaviour that is exclusively, or predominantly, expressive in character. Again, this decision shows the difference between the European approach (which the Hungarian Court evidently adopted as its own) and the US approach, where any statutory attempts to punish the desecration of the flag have so far been found unconstitutional.270 This is precisely because any negative impact of desecration is deemed in the United States to be attached to the expressive aspect of the conduct in question, and this places such conduct at a very high level of constitutional protection, under the First Amendment. While the suppression of hate speech raises the issue of balancing the values of free expression the protection of groups subjected (at least in the past) to persecution and harassment, the notion of protecting of religious sensibilities raises in CEE a quite different balance, as the claims to protection have typically been made by majoritarian, dominant religious groups. Hence, the demands for legal protection lead to something resembling more of a censorship than an anti-harassment policy. This was true, in any event, in the most important (if not the only) case of that type
164
CHAPTER 6
that has come before a constitutional court in the region. Mixing the issues of freedom of expression in the media with religion, one of the most controversial decisions of the Polish Constitutional Tribunal was handed down on 7 June 1994,271 upholding a law of 29 December 1992 requiring respect for “Christian values” in broadcasting.272 The challengers (a group of Left-wing deputies) argued that the provisions were inconsistent with freedom of expression (at the time, this freedom was deduced from Art. 1 of the “Little Constitution”, which proclaimed the principle of a “democratic state based on law”, in connection with international human rights laws) and also with the principle of equality before the law (by privileging a particular system of religious values). Two provisions of the broadcasting law were subject to challenge: a ban (applying to all electronic media) on violating the “religious feelings” of the viewers and listeners (Art. 18(2)), and a requirement (applying only to public radio and TV) to respect “those Christian values that coincide with universal moral principles” (Art. 21(2)). As is readily evident, the latter rule, potentially much more questionable because formulated as a positive requirement, was quite ingeniously worded so as to dispel possible charges of discrimination; the mention of “coinciding with universal principles” was meant to inject a non-discriminatory and non-religious character into this provision. However, one may oneself, why mention “Christian values” in the first place, if their ambit is relevant to the provision only insofar as they overlap with “universal” values? The only possible reason for it was that the “universality” clause was a fig leaf, inserted to disguise the pro-Christian character of the provision. In trying to diffuse the objection to this use of “Christian values” terminology, the Tribunal accepted that “[t]he order to ‘respect’ [Christian values] cannot … be interpreted as an order to propagate the Christian value system. The directive [in the statute] identifies the sphere of values that should not be omitted in the program of public radio and television treated in its entirety”.273 This must be seen as an attempt to restrict the impact of the provision, particularly as earlier the Catholic Church’s Conference of Bishops had issued its own authoritative interpretation of “Christian values” (explicitly for the purpose of the interpretation of this law), in which it attacked such phenomena as homosexuality, divorce, abortion etc. as inconsistent with “Christian values”.274 The Tribunal explicitly referred to the principle of “religious neutrality” and, on this basis, announced that the Art. 21(2) requirement was meant to identify “those values belonging to the realm of Christian culture that are at the same time fundamental, universal moral principles”. As far as the other, less problematic provision was concerned, the Tribunal announced that the prohibition against violating religious sensibilities is related to the constitutional principle of freedom of religion, and that “religious feelings, due to their nature, are subject to special legal protection”. Responding to objections regarding the discriminatory nature of this special protection for Christian values, the Tribunal defended the law by saying that it did not differentiate between the protection of religious feelings depending on what religion it is, and that the mention of Christian values figured there only as an example. The Tribunal reinterpreted the law in a way that injected it with a non-discriminatory character as between various religions. However, in the context of Poland, heavily dominated by Roman Catholicism, this is perhaps not as important as is the broader question of the problems
CIVIL AND POLITICAL RIGHTS
165
related to the elevation of religion as such, as an object of special protection in the media. This, as was argued in the Polish press at the time, might have a chilling effect on any public debates that touch upon religious issues. However, with the benefit of hindsight one should say, in all fairness, that these provisions have turned out to be relatively toothless, so far. Nevertheless, the “chilling” potential of the statute cannot be totally disregarded.275 It should also be noted that, even before the June 1994 decision, the same law had been subject to an “interpretative decision” handed down as a response to a question (but not challenge) from the Chief Justice of the Supreme Court. In a decision of 2 March 1994,276 the Constitutional Tribunal explained that the provisions of Art. 21 cannot be understood as mandating prior restraint because this would amount to an impermissible restriction of freedom of speech. Also, further limiting the possible negative impact of the law, the Constitutional Tribunal explained that the programs broadcast by public radio and TV (to which Art. 21 applied) should be considered in their “entirety”, and so whether they properly respect religious sensibilities and, in particular, the “Christian system of values”, is not a matter that can be ascertained with respect to any particular broadcast taken in isolation from others. In conclusion, while the Constitutional Tribunal abstained from striking down the law, which is extremely problematic from the point of view of freedom of expression, one should acknowledge that in its two decisions on the matter, it went a long way towards disarming its “chilling” effects. So much for protection of religious sensibilities. Another hot issue that has implications for our understanding of the constitutional protection of freedom of speech concerns the scope of responsibility – and in particular, of criminal liability – for defamation of politicians. Special, high penalties for defamation of high officials were not only the legacy of Communist era, where political criticism was identified with anti-state subversion, but also, more generally, of the long tradition going back to the feudal crimen laese maiestatis: a pattern of increasing the severity of punishment as a function of the relative height of position of the person defamed. Constitutional courts in the Czech and Slovak Republics and in Hungary have played a role in redressing this pattern, which is an aberration in modern democratic system where the high officials should enjoy lower rather than higher protection against potentially defamatory criticism. In the Czech Republic, President Vaclav Havel brought, in 1993, a constitutional challenge against the criminal prohibition of disparagement of the Republic, the Parliament, the government and the Constitutional Court (Art.102 of the Criminal Code). He did so on the grounds of vagueness and conflict with the constitutional freedom of expression.277 As, more recently, a judge of the Constitutional Court has commented, such a prohibition could be seen as a remnant of the old Austro-Hungarian period ban on criticism of the Emperor; as he added “the Communists [after the Second World War] loved it”.278 In its decision, the Court adopted a Salomonic approach, drawing a line between disparagement of “the Republic” (as an abstract concept, symbolizing the State) on the one hand, and disparagement of the specific state bodies protected in this provision, on the other. The Court found the former prohibition reasonable, and not lending itself to any indeterminacy. When it came, however, to the protection of the Parliament, the government and the Constitutional Court against “disparagement”,
166
CHAPTER 6
the Court found that this was contrary to Art. 17 of the Charter of Fundamental Rights and Basic Freedoms (freedom of speech). This was due to the fact that the prohibition was vague and indeterminate; it is unclear where criticism ends and disparagement begins.279 Such a prohibition also does not correspond to the acceptable limitations to rights defined by Art. 17(4) of the Charter (i.e. necessary to protect the rights and freedoms of others, the security of the state, public safety, public health or morals), and the European Convention of Human Rights, Art. 10.280 The Court noted that “Article 10(2) of the [European] Convention allows for the protection of third persons’ reputation, and politicians enjoy this protection as well, but in such a case, the demands of protection must be moderated by the interest in the free discussion of political issues”.281 Much more recently, in January 2002, the Slovak Constitutional Court considered a challenge against similar provisions of the Criminal Code, namely Articles 102 and 103, which penalised the defamation of the Republic, public officials or institutions;282 this was soon after a parliamentary motion to repeal these provisions was narrowly defeated (in November 2001). The urgency of the challenge was all the more evident as Art. 103 explicitly referred to the defamation of the President of the Republic in matters related to the discharge of his official duties, and at the same time President Schuster was suing a journalist in connection with a satirical article. The Court invalidated both of these provisions as inconsistent with freedom of speech, thus making the prosecution of the journalist involved impossible. However, as one of the MPs instrumental in bringing this petition to the Court observed, it still left another article in the Code (Art. 156), that “contained similar language pertaining to speech directed against the country or its public officials”.283 In 1994, the Hungarian Constitutional Court handed down an important and highly libertarian decision regarding protection of public officials against defamation.284 The Court struck down Article 232 of the Criminal Code (modified in 1993, hence with an already post-communist pedigree), which criminalised the defamation of public officials. Invoking the equal protection standard, the Court determined that Article 232 was unconstitutional because it granted more protection from defamation to public officials than to private officials. In fact, the Court stated that the sphere of allowable criticism must be wider for politicians (especially for the government and its officials) than it is for others. This is because the activity of government has to be examined carefully, not only by the legislative and judicial branches, but also by the press and the public. Therefore, the Constitution assumes that public officials must endure possible indignity in the interests of democratic deliberation, while at the same time retaining their right to sue in their private capacity. Further, the Court noted that the article under challenge resulted in someone committing a criminal offence even if the statement made was true. Also, if the statement was false, it was irrelevant under Art. 232 whether it was made in good faith or negligently, rather than in bad faith, an observation clearly influenced by the case New York Times v. Sullivan285 with its “actual malice” test for the defamation of public figures. (The impact of this US case upon that particular decision was later to be explicitly acknowledged by Justice Sólyom).286 For all these
CIVIL AND POLITICAL RIGHTS
167
reasons, the Court found that the article did not pass the “test of necessity”, and that it unnecessarily and disproportionately restricted the right to freedom of expression. In relation to the freedom of the media, the Hungarian Court struck down, in 1992, a provision of the decree on Hungarian Radio and TV, which aimed to subject public radio and TV to the supervision of the Council of Ministers.287 The decision, as Justice Sólyom later explained, “was born out of the ‘media war’ concerning the Government’s influence on the state-run nationwide radio and television”.288 The issue of control over public media was all the more important as commercial electronic media did not exist at the time, so the governments of the day kept arguing with the opposition about which political group was favoured by the public broadcasters, who enjoyed a virtual monopoly on the airwaves. The Constitutional Court based its decision on formal and on substantive grounds. Formally speaking, the amended provision of the governmental decree (Art. 6 of the Decree on radio and TV) was issued after the entry into force of the Act of 1987 on Legislation, which provided that, from 1988 onwards, regulation of the press could only be performed by statute. In addition, the decree was also found to be inconsistent with Art 61(4) of the Constitution, which provides that the supervision of public media may only be regulated by an Act passed by a two-thirds parliamentary majority. Regarding the substance of the law, it was held to be contrary to the constitutional right to freedom of expression,289 the more specific right of freedom of the press and, even more specifically, the need to secure “comprehensive, balanced and accurate reporting” by national public radio and TV.290 The Court also linked the issue of the composition of the public broadcasting supervisory body to journalistic independence and freedom of expression. A quasiconsensus developed at that time, to the effect that such a body must reflect the parliamentary composition, and that the appointment and dismissal of chairpersons of public radio and TV must be backed by a consensus of the parliamentary parties, including the opposition. The Court rejected this opinion, shared at the time both by the President and by the parliamentary opposition.291 As the Court explained, the creation of a joint, parliamentary/governmental board to oversee radio and TV was not acceptable, because consensus between political parties is not sufficient to guarantee the constitutional protection of freedom of opinion. Rather, freedom from the state demands that neither Parliament nor government nor political parties can exert influence over the media’s program content. The Bulgarian Constitutional Court has handed down a number of decisions aimed at protecting the independence of the media. One of the most important was a decision that struck down new procedures for selecting a Media Council, on the basis that they tended to constitute partisan interference with national media.292 This law was adopted by the Bulgarian Socialist Party legislative majority, and immediately challenged by the Union of Democratic Forces opposition. Most of the provisions were struck down by the Court, including Art. 25(2), which required that, in comments appearing on TV and radio, the facts “cannot be presented onesidedly”. The importance of this decision was additionally underlined by the fact that this was the first case in which proceedings were open, not in camera. The majority opinion was drafted by Justice Todorov, who was rapporteur in this case. As he later told me:
168
CHAPTER 6 What guided me when I was writing the draft was the practice of the European Court of Human Rights. I argued [in the text of the decision] that you cannot prevent a journalist from making his own selection of the facts to be communicated to the audience, and that in every case you have to make a distinction between facts and the personal opinion of the journalist.293
In 1994, the Lithuanian Constitutional Court struck down some provisions of the Statute on Radio and TV.294 The statute provided for a competition to lease broadcasting facilities to private broadcasters, and conferred (in Art. 9(3)) upon the Board of the State Television (LRTV) the power to confirm the decision of a special commission regarding the result of such a competition. The Constitutional Court examined the role of this board, and compared its powers to those of broadcasting boards in other countries. It then stated that Art. 9(3) of the statute contradicted Article 25(2) of the Constitution (the right to obtain, seek and disseminate information and ideas). The provision envisaged that the Board could veto the commission’s decision, yet the Board was not bound by any of the legal conditions in the Statute, nor were the criteria for its decision set out. This meant that the board could hinder the activity of the private media and, through procrastinating over making a decision, could thus prevent the dissemination of information and ideas. Art 9(3) of the law was therefore held to be unconstitutional. As one can see, nearly all the constitutional court decisions on the media have concerned the degree and the forms of control of the public and private media by the state. Press law, sensu stricto, has rarely reached the Constitutional Courts. One exception was a recent decision of the Hungarian Court which found unconstitutional a provision on the right of reply in a proposed amendment to the Civil Code.295 Under the proposed amendment, anyone could demand the publication of their opinions in the press if they felt that their rights had been infringed by a publication. The law envisaged sanctions, including court-imposed fines, against the media in the case of refusal to publish the reply. The Court, which considered the case pursuant to a referral by the President, found the provisions for the right of reply unconstitutional; not because they violated freedom of the press per se, but rather because “the language of the amendment might give rise to uncertainties and unforeseeable circumstances”.296 Considering the very contingent nature of the justification, the fact that the Court was split (four justices dissented), and that the judgment was subject to strong criticism,297 this is probably not the last word of Hungarian law on this matter. The question of restrictions on advertisements, or commercial speech, has not thus far generated far much jurisprudence of constitutional courts. In a rare, commercial-speech case the Lithuanian Constitutional Court invalidated a ban on the advertisement of tobacco and alcohol related products.298 This remains one of the more controversial decisions in the history of the Lithuanian Court thus far, and the responses to the decision were far from unanimously approving.299 The two statutes in question banned tobacco and alcohol advertising. The Constitutional Court first held that, generally speaking, this was a justifiable restriction of the freedom to obtain and disseminate information, because of these products’ proven risks to health, and the profit-making nature of advertising. Hence, the very idea of a “lowered scrutiny” (to use the US constitutional parlance) for commercial speech
CIVIL AND POLITICAL RIGHTS
169
restrictions was accepted. However, the Court found that Articles 1(5) of both of these statutes that provided also for a ban on advertising products and articles not directly linked to tobacco or alcohol, did contradict Article 25(3) of the Constitution, which provides for free dissemination of information and ideas. This is because these statutory provisions impermissibly broadened the concept of advertising such products, and thus could lead to the banning of advertisements on a groundless basis, this, therefore, being an illegal restriction of the freedom to information.300 In addition, these two statutes provided that the criteria for determining whether an advertisement is for tobacco or alcohol, and the procedure for control of compliance, would be established by the Government. This was held to contradict Article 25(3) of the Constitution, which provides that any restrictions on the freedom to information will be established by law, as this Law would have given the government a right to restrict advertising, even in its legitimate forms. It is interesting to note that the petition of the challengers of these laws (a group of parliamentarians) wanted to go much further and have any forms of bans on tobacco and alcohol advertising invalidated. The grounds of the challenge included not only freedom of expression but also an anti-discrimination provision: “a person who does not smoke or is a teetotaller [sic] is not by himself of a greater social value than a person whose standpoint and behaviour are opposite”.301 Hence, there was a clear appeal to the argument of viewpoint-neutrality; according to the challenge, the advertising ban discriminated against some members of the population on the basis of an official disapproval for their views regarding smoking or drinking. As the petitioners argued (in the summary provided by the Court in the text of its decision): the Constitution guarantees not only the right of every person to have his convictions as to the consumption of alcoholic beverages or expediency of smoking etc., but also that persons who have a negative standpoint as regards the consumption of alcoholic beverages and smoking may not have any privileges with respect to those whose standpoint concerning analogous matters is indifferent or positive.302
It is a serious and persuasive argument, but the Constitutional Court did not share this view. The main body of the decision (as far as the upholding of the main core of the advertising ban was concerned) is based on the observation that the use of alcohol and tobacco is harmful to human health;303 that “in most cases the content of advertisement of tobacco products is one-sided and biased”304 (in that the behaviour is presented as glamorous and non-harmful); and that a ban on such advertising is not discriminatory because it “concerns the whole society and not particular groups of people”.305 8. CONCLUSIONS Without doubt, the articulation of personal, civil and political rights, as discussed in this chapter, has been a real showcase of the role of constitutional courts in applying democratic and liberal standards to the interpretation of constitutions in post communist states. It has been the most visible and the most obvious area in which the break with the authoritarian past could be made, and largely has been made. New constitutions in CEE have, largely, complied with the criteria of robust
170
CHAPTER 6
protection of personal, civil and political rights, achieving standards broadly similar to modern West European constitutions; the differences are at the margins, and cut across the West/East divide in Europe. What mattered was the articulation of vague constitutional mandates to controversial and politically vexed cases: Does a right to religious freedom allow or prevent employment in public schools of religion teachers nominated by religious groups?; Does freedom of speech protect racist hate speech against criminal persecution?; Does freedom of association extend to minors? etc. Constitutional courts have, on balance, done a good job in providing those vague constitutional mandates with more liberal rather than restrictive interpretations: in four countries it was the task of the constitutional courts to invalidate death penalty; and in a number of states, they have protected media (in particular public radio and TV) against control by political élites. They have made the right to conscientious objection more meaningful; they have cleansed rights to association of excessive statutory restrictions; they have protected personal privacy against governmental intrusion, etc. But the picture is not unambiguously positive. They have been more libertarian when making general, abstract declarations about the scope of political rights (such as freedom of expression) than in applying truly libertarian rationales in interpreting constitutional provisions when striking down specific statutory rules – the Hungarian decision on incitement to hatred, outlined above, may serve as example of this. Furthermore, from time to time, they have missed opportunities to invalidate laws that were very questionable from the point of view of liberal standards: the deference by Polish Constitutional Tribunal to governmental decrees and parliamentary statutes that privileged the Catholic Church in education and broadcasting illustrate this well. Worse than this, the same Constitutional Tribunal dealt a blow to a reasonable, compromise-based law on abortion; less radically, its Hungarian counterpart also pushed the legislator whenever it could into making the law on abortion less rather than more liberal. Therefore, it needs to be stressed: the overall picture is far from unambiguously positive.
CHAPTER 7 SOCIO-ECONOMIC RIGHTS
Regardless of philosophical controversies over whether socio-economic rights properly belong in constitutions, the omission of such rights was never a plausible political option for constitution-makers after the fall of communism. For one thing, the economic legacy of communism was characterized by widespread poverty, deprivation, and a lack of equal economic, cultural, and educational opportunities. While the economic transformation unquestionably brought some benefits, it also caused, at the same time, a sharp increase in inequality, unemployment and poverty in large segments of the population – and this in a region used to the ideas of (relative) economic equality, full employment, and universal social safety nets. Omitting socio-economic rights from the new constitutions would have sent a signal to the “ordinary people” in those societies that the political élites that emerged after the fall of communism were insensitive to the plight of common people who had been so dramatically affected by the dire economic situation in these countries. Socio-economic rights have been a visible, and highly symbolic, expression of numerous claims, demands and pressures upon social policy-making in a novel and, for many, dramatic situation, in which “[s]ocial policy was . . . used to compensate for the immediate social consequences of economic transformation”.1 In addition, the impact of liberal political forces (“liberal” in the European, freemarket sense), which are often reluctant, on philosophical grounds, to constitutionalise broad catalogues of socio-economic rights, has been relatively weak, and the political weight of social democrats and Christian Democrats in the region has been quite strong.2 These parties, as well as non-ideological peasant parties, played an important role in infusing the constitutional charters of rights with symbolic statements of their attachment to the idea of an activist state, protecting the citizens against economic calamities. To some degree, this was also an ideological legacy of communism3 which generated strong welfare expectations. The idea that citizens are entitled to a certain, albeit often miserably low, standard of living, work, recreation and education has proved to be particularly well entrenched in public consciousness.4 This may explain the popularity of views advocating broad constitutionalisation of socio-economic rights. For example, in Poland, a prominent and influential lawyer, Tadeusz ZieliĔski, has argued that, unless socio-economic rights are elevated to the constitutional level, the authorities will have full discretion to further reduce, or even to disregard, citizens’ social entitlements.5 Referring to his own experience as Poland’s second ombudsman, Professor ZieliĔski further claimed that,
172
CHAPTER 7
unless socio-economic rights are constitutionalised, even legislative and executive acts contrary to the European Social Charter and international human rights covenants will be immune to legal challenge.6 Similarly, Herman Schwartz has argued that many socio-economic rights are, or may be, judicially enforceable;7 and that, even if they are not, they nevertheless represent “a way of imposing political and moral obligations on those who operate the state’s governmental apparatus” to take appropriate steps.8 He went on to suggest that at least some of the critics of the constitutionalisation of these rights are, in fact, opposed to their very implementation, constitutionally or otherwise.9 It is also worth adding that the absence of socio-economic rights in the constitution does not necessarily imply the absence of any constitutional anchor for social welfare programmes; terms such as “social justice” (e.g., in Estonia’s Constitution, art. 10), or “social state” (Slovenia art. 2; see also Germany art. 20, or France art. 2, which refers to “social Republic”) may also serve as the basis for constitutional review of government social programs. Two caveats of terminological nature are useful at the outset. While socioeconomic rights are usually treated as “positive” rights, and as identifying “programmatic” goals for the government, such characterisation is not quite accurate. The distinction between policy guidelines and rights sensu stricto does not correspond neatly to a distinction between socio-economic rights and civil-political rights (because the rights that apply to a socio-economic sphere may have a determinate content that imposes clear limits upon state action). Nor does it correspond to a distinction between “positive” and “negative” rights (“positive” rights may impose determinate limits upon state action, with the result that a failure to act may be unconstitutional). The positive/negative distinction, in turn, does not correspond to a distinction between socio-economic and civil-political rights (some civil and political rights may require positive state action10 while some socioeconomic rights may demand state non-interference with individual action).11 It is therefore important to keep these three distinctions (determinate rights v. policy guidelines, socio-economic v. civil-political rights, and positive v. negative rights) separate. Furthermore, even within the same range of rights, either a “minimal” or a “strong” use can be made thereof, both in the process of enforcement and in judicial review. A minimal use would be to view entrenchment as a guarantee against arbitrary and discriminatory limits on access to a given socio-economic program, whatever it may be. In other words, it does not order the state to run any particular program, say, of education, housing or health care; however, once a program is in place, constitutionalisation amounts merely to a guarantee of equal access.12 Such a use of socio-economic rights has been suggested, for example, by Herman Schwartz.13 A strong use of socio-economic rights, on the other hand, would go further than merely prohibiting arbitrary or discriminatory exclusions; it would call for the adoption by the government of effective measures aimed at attaining the programmatic goals as defined by the constitution-maker. While choosing the “minimal” interpretation would probably weaken much of the criticism of the constitutionalisation of socio-economic rights, such interpretation seems somewhat
SOCIO-ECONOMIC RIGHTS
173
implausible, due to its redundancy: discriminatory policy is already prohibited by constitutional rules against discrimination. 1. CONTROVERSY AROUND SOCIO-ECONOMIC RIGHTS In proclaiming broad catalogues of socio-economic rights, CEE constitution-makers were no different from the drafters of the Western European constitutions who, in contrast to the US model of constitutionalism, included these types of rights alongside civil and political ones. These constitutions, the majority of which originated in the post-World War Two wave of constitution-making (or, as was the case of Spain and Portugal, from the fall of authoritarian regimes well after the War), elevated the then dominant model of welfare state into a constitutional structure. Hence, almost all of the Western European constitutions14 proclaimed the general principle of the Social State,15 or enumerated broad catalogues of socioeconomic rights,16 or did both.17 In addition, the predilection of Western European constitutionalism for dispensing broad socio-economic rights (often interspersed with descriptions of the goals of the state in the field of socio-economic policy) is visible in the most recent constitutional-rights product of Western Europe, namely the Charter of Fundamental Rights of the EU. The Charter proclaims a number of social rights, grouped mainly in Chapter IV (“Solidarity”), with a few listed in Chapter II (“Freedoms”) and III (“Equality”). Some of them are formulated in a categorical fashion, suggesting that they impose strict conditions upon the lower laws, and that all EU and national laws must comply with them. These include rights to education (including to free compulsory education);18 rights of children to protection and care, to express their views freely, and to maintain contact with both parents;19 freedom to choose an occupation and the right to engage in work;20 maternity-related rights (against dismissal and to paid maternity and parental leave),21 etc. Other social rights depend for their protection upon other Community and national laws; therefore they presumably cannot be used to declare Community laws contrary to the Charter (although, arguably, a failure to provide for such rights would constitute a violation of the Charter). This category includes the rights to social security and social assistance,22 to health care,23 to services of general economic interest,24 etc. Finally, certain provisions in the Charter are worded in such a way as to suggest that they are not really viewed as “rights” (even though they are contained in the “Charter of Fundamental Rights”) but rather as policy directives regarding certain socio-economic goals to be pursued by the authorities, such as the goal to achieve a high level of health protection,25 to provide legal, economic and social protection of the family,26 to ensure the social integration of persons with disabilities,27 and to ensure the right of dignity and independence for the elderly.28 It is thus clear that, in this respect at least, CEE constitutions follow the pattern of Western European constitutionalism, and contrast starkly with the constitutional tradition of the United States, where all attempts to read welfare rights into the Constitution have been consistently and emphatically resisted by the Supreme Court.29 In the oft-quoted words of Judge Richard A. Posner, the official interpretation of the Bill of Rights is of “a charter of negative rather than positive
174
CHAPTER 7
liberties” motivated not by the concern “that government might do too little for the people but that it might do too much to them.”30 Decidedly, the concern that the governments “might do too little” featured prominently in the minds of the drafters of CEE constitutions. It is, however, one thing to say that the inclusion of socio-economic rights was politically the only plausible option, and quite another to say that it was unproblematic. Far from it; in the constitutional debates within post-communist states and among outside observers, some important objections were raised to the idea of constitutionalising welfare rights. It is important to emphasise that the reasons for the rejection of the idea of constitutional welfare rights were not grounded on a rejection of welfare policies.31 The basic idea was that there is a non sequitur between advocating a welfare policy and advocating the elevation of welfare rights to the constitutional level. It is interesting to note that some of the countries with the most developed and generous welfare policies have no constitutional social rights: the Scandinavian countries,32 Australia and New Zealand belong to this category. This serves to rebut the objection that all those who oppose constitutional socio-economic rights are anti-welfare. Some participants in this debate compare, on the one hand, generous welfare states with no socio-economic rights in their constitutions with, on the other, countries that have an appalling welfare situation despite boasting impressive catalogues of constitutional socioeconomic rights. Some scholars actually go a step further, and assert an inverse relationship between the presence of socio-economic rights in a constitution and the existence of a welfare safety net. For example, Ulrich Preuss, a careful student of post-communist constitutionalism, has noted, with respect to CEE constitutions, that “it is striking that a number of [constitutional] pledges—be they state goals or social rights—increase in inverse proportion to the extent that these countries are able and prepared to establish a welfare state…”.33 As already mentioned, the opponents of constitutional welfare rights may or may not be advocates of broad welfare policies. Practice around the world shows that there is no tight connection between how “generous” the social rights in a constitution are, and how generous social welfare policy actually is. Some opponents of constitutional welfare rights are concerned that, once a welfare right is written into a constitution, even if subject to various provisions relating to non-justiciability, there is nothing that will disable a constitutional court from scrutinising a government policy or a new law in terms of its compatibility with the right in question. As Preuss has noted: Both social rights and state goals [when entrenched in constitutions] increase the power of the executive – which has the resources to design and to implement particular policies – and that of the courts – which make the final decision about the constitutional duties of the government – at the expense of the democratic authority of the parliament.34
Thus, the primary reason for disapproving of constitutional welfare rights is that they produce an unfortunate institutional shift in the separation of powers and allow (indeed, require) constitutional judges to decide matters in which they have neither qualifications nor political authority.35
SOCIO-ECONOMIC RIGHTS
175
Perhaps the most eloquent criticism of this transfer of powers from the legislature and the executive to the judiciary has been raised by Andras Sajo, in his article on the impact of decisions by the Hungarian Constitutional Court upon governmental attempts to restructure the welfare system.36 Sajo’s passionate critique was occasioned by a series of Hungarian Constitutional Court decisions striking down several laws that together made up an austerity policy package (the “Bokros package”, named after the Hungarian Minister of Finance). The Court invalidated, among others, changes to maternity and family support, reductions to household allowances and state subsidised sick leave, and a raise in the interest rate on state loans to homebuilders.37 Significantly, the journal in which Sajo’s article appeared provided it with the following subtitle: “Welfare rights + constitutional court = state socialism redivivus”. Also in Poland, the prospect of review by the Constitutional Tribunal under the new Constitution, which contains a broad array of “programmatic” socio-economic rights, has led some commentators to fear that the Tribunal would get embroiled in policy-making. As a leading Polish constitutional scholar (who later became a judge of the Constitutional Tribunal) Jerzy Ciemniewski warned, if the Tribunal gets to wield the power of review of the compatibility of laws with certain socio-economic constitutional norms, “we will embark upon a very dangerous path by combining the roles and functions of different categories of branches of state and by confusing the scope and nature of the responsibilities carried by these bodies”.38 It is clear that interpreting the scope and limits of socio-economic rights provides the constitutional courts with the most obvious opportunity to engage in making economic policy judgments. This fact has been recognised, for example, by a leading Polish constitutional scholar, supportive of the Constitutional Court and of social-economic rights, later to become a judge of the Constitutional Tribunal and now a judge of the European Court of Human Rights, Professor Lech Garlicki. He acknowledged (approvingly) that “if one of the criteria of limiting the social rights of citizens is the difficult situation of the state budget . . . then it becomes necessary for the [Polish Constitutional] Tribunal to form economic judgments”.39 The same author later stated, in describing the current doctrine of the Tribunal, that despite the fact that the Tribunal recognised the broad discretion of legislators in the area of social rights, nevertheless “appeals to the principle of social justice and to the notion of the ‘essence’ of social rights . . . leave a large margin of intervention to the Constitutional Tribunal should it consider it justified”.40 A second reason for not including socio-economic rights in the constitutions was the fear of “contaminating” the entire charter of rights by the under-enforcement of this one particular category.41 These rights are, by their nature, under-enforceable. The fear was that a habit of tolerance for under-enforcement of some rights could erode a rigid commitment to the enforcement of all other rights, including civilpolitical ones. The plausibility of this argument has been strongly disputed by, among other people, Herman Schwartz: “This notion that if some rights turn out not to be effective, others will be in some way degraded in value, is utterly complete nonsense”.42 For my part, I believe that Schwartz’s point may well be correct with regard to systems in which the values of constitutionalism, the rule of law and the protection of rights are well established, and where disagreements about rights
176
CHAPTER 7
pertain to the margins rather than to their core meaning. However, in a system in which a nihilist tradition of treating a constitution as a purely decorative instrument is strongly embedded, and where the fundamental notions of constitutionalism and the rule of law have a weak hold upon the collective consciousness, anything that undermines a strict construction of constitutional limits upon discretionary governmental action is to be regarded with concern. Finally, it has been claimed that, while statutory welfare rights may be a good thing, writing them into the constitution is wrong because, by its very nature, a constitution is meant to restrain legislators (and indirectly, the electorate) from giving in to potentially pernicious temptations.43 Constitutional rights are seen primarily as restraints upon certain actions that may arise from human nature under some particular socio-political circumstances. For example, in the context of CEE constitutions, Cass Sunstein has claimed that elevating welfare rights to a constitutional level may promote attitudes of welfare-dependency and become a counter-incentive to self-reliance and individual initiative.44 However, as soon as this rationale is explicitly spelled out, it becomes immediately evident that it is extremely unlikely that any actual constitution-making process will follow its logic. Politically speaking, it is an almost impossible proposition, as it would require the constitution-makers to propose, and enact, ideas directly contrary to conventional societal norms. 2. CONSTITUTIONAL CATALOGUES OF SOCIO-ECONOMIC RIGHTS As a result of these theoretical and political discussions, and of differences in local circumstances, there is some variety in the catalogues of socio-economic rights and in their status within the post communist constitutions of CEE countries. The three fundamental socio-economic rights that figure most prominently in these constitutions are social security, health care, and education. Nearly all of the constitutions of the region contain broad provisions for rights to social security, either for all those unable to work or for all in material need.45 The latter group is sometimes defined as those that have no other means of support.46 Many of these social security provisions go on to delineate specific subcategories of people who can legitimately be subsumed under the general notion of persons unable to work. These subcategories include the old,47 the ill,48 the disabled,49 those that have lost their breadwinner,50 and widows and orphans.51 In contrast to these constitutions with “generous” provisions, a small minority of the constitutions in the region52 has only narrowly drafted provisions for social security, while two have no such provision at all.53 A right to health care54 is present in all of the constitutions of the region: it has been proclaimed very “generously” (with free health care for all) in twelve;55 in some, however, health care is only a right conferred upon some categories of people, such as the elderly, children, and pregnant women.56 Four constitutions delegate to lawmakers the task of determining who will obtain free health care.57 Three constitutions cautiously provide a right for all to health insurance rather than to actual health care.58 The Hungarian Constitution proclaims a right for all those living in the
SOCIO-ECONOMIC RIGHTS
177
territory of Hungary “to the highest possible level of physical and mental health”59 – a guarantee that deftly evades any possible challenges due to the vagueness and indeterminacy of the term “highest possible level”. Often, health is linked in constitutional provisions with the individual’s environment, again not in terms of a state objective to care for the latter, but rather in terms of citizens’ right to a “healthy environment”60 or “healthy life . . . [in] a healthy environment”.61 The third right, that to free education, is recognised universally, although the level of constitutionally mandated free education varies; a free education is guaranteed up to university level in eight constitutions,62 up to secondary level in five,63 and up to primary level in six constitutions in the region.64 Among other socio-economic rights, the most frequently mentioned are those that relate to working conditions, including a right to choose one’s own profession, a right to safe conditions at work, to “adequate pay,” to guaranteed leisure time, and special protection for certain specified categories of employees (women, the young, the old) in the workplace. Without going into detail, the analysis shows that nine constitutions have a very broad list of such work-related rights.65 Other assorted socio-economic rights include the protection of the family, motherhood, and/or childhood (14 constitutions), training for the disabled (8), protection of culture (13), and a right to a good environment (13).66 Considering the catastrophic housing situation in most of these countries, it is not surprising that only four constitutions explicitly proclaim a right to adequate housing,67 with two others listing it as an aim of the state rather than an enforceable right.68 To sum up the textual analysis of the catalogues of socio-economic rights, one can establish a simple taxonomy of the constitutions as falling into the following categories: 1.
Nine constitutions list comprehensive social security, education, health care, work protection rights, and other socio-economic rights; these nine are the most “generous” constitutions;69
2.
Six constitutions have limited social security, education, and health care rights, but good work protection guarantees, and many other socio-economic rights;70
3.
Three constitutions provide for good social security, education, and health care rights, but only a limited number of the other rights;71 and
4.
Two constitutions have very few socio-economic rights at all.72
At this stage, three preliminary conclusions can be drawn. Firstly, postcommunist constitutions are, overall, “rich” in socio-economic rights. If one imagines a continuum in world constitutionalism, based upon the “generosity” of
178
CHAPTER 7
dispensing socio-economic rights, post communist constitutions are approaching the pole that provides the maximum number. Secondly, the range of local variety is not particularly wide. If one ignores two non-typical cases (Bosnia and Herzegovina and Georgia), and perhaps also the three Baltic states, the degree of diversity is relatively small. Thirdly, and most importantly, there is no discernible variable that would significantly account for this “generosity”. Not a single significant factor can persuasively explain the taxonomy suggested above – not the level of economic growth, not the adopted strategy of development, not the strength of post communist political forces, not the legacy of the former USSR, not the speed with which the constitution was created, not the realistic prospect of admission to the EU, and so on. For instance, category (1) above (the most “generous” constitutions) includes both the relatively affluent (Czech Republic, Poland) and the poorest (Moldova, Ukraine) countries of the region; those countries that adopted economic “shock therapy” in the transition to a free market economy (Poland) and those that failed to adopt freemarket measures (Belarus); those countries where post communist parties have been relatively marginalised (Czech Republic) and those where they have maintained their grip on power for a reasonably long time (Belarus, Slovakia); those that adopted constitutions soon after the transition (Slovakia) and those that took a long time to do so (Poland). Similar points can be made about category (2). In other words, there seems to be no meaningful correlation between the “generosity” of the catalogues of socio-economic rights in a given constitution, and the objective circumstances of that country. Apart from everything else, this would confirm the hypothesis that the constitutionalisation of welfare rights has little or no effect upon the actual welfare policy of the government, although it certainly may have an effect upon the institutional system of separation of powers. 3. THE STATUS OF SOCIO-ECONOMIC RIGHTS The effect of enhancing the powers of constitutional courts by bringing them into the social policy-making process is as much a matter of the constitutional status of socio-economic rights vis à vis the “traditional” rights as of the content of the actual catalogue of the former group of rights itself. This is why it is important to study not only the catalogues of rights, but also the ways in which post-communist constitutions handle the constitutional weight of socio-economic rights compared to other constitutional rights. Particularly because, as one keen observer of the constitutional scene of the region has remarked, there is “a growing sensitivity in East-Central Europe that social and economic rights should be treated differently from political rights and citizens’ freedoms”.73 How has this “sensitivity” penetrated the actual structures of constitutional texts? Looked at from this angle, one can distinguish three categories of constitutions. The first group, by far the largest, contains those constitutions that do not draw any meaningful distinctions between socio-economic and all other rights. In the fourteen constitutions belonging to this category,74 no differentiation is made as to the enforceability of socio-economic versus civil-political rights. In some of these constitutions, the two types of rights are even lumped together in the same
SOCIO-ECONOMIC RIGHTS
179
subdivision of the constitutional text.75 The second category contains two constitutions: those of the Czech Republic and Slovakia (not surprisingly, considering that their respective bills of rights originate from one and the same text).76 Here, a clear separation of socio-economic from other rights is achieved by a general clause stating that a number of specifically enumerated rights (including most socio-economic rights) can be claimed only within the limits of the laws enacted to implement these rights-provisions.77 Hence, in contrast to all other rights unaffected by this general limiting clause (which can only be restricted in accordance with constitutionally established criteria), socio-economic rights are subject to legislative restrictions over which the (ordinary) legislator has wide discretion. This severely limits the possibility of mounting constitutional challenges to laws and policies in terms of these rights. Also, it effectively, though not formally, reduces the weight of these rights by putting them on par with constitutionally established state targets or aims – politically binding upon the legislature and the executive, but not judicially enforceable. It should be noted that, in the first category of constitutions (those with no distinction between socio-economic and other rights), one can also find particular provisions that establish that practical details of certain rights shall be decided by law. For example, the Russian constitutional provision guaranteeing the right to social security is accompanied by a proviso that the details of state pensions and social benefits shall be established by law.78 This, however, is different from a general clause (in the Czech and Slovak fashion) that conditions enforceability of a right upon a legislative choice. Indeed, both the Czech and Slovak constitutional documents contain some (not socio-economic) rights that provide that the details of a right are to be established by law, and yet are not covered by the general limiting clause.79 The third group is a hybrid category that combines the first and second approaches. In four constitutions in the region,80 we find a mixture; some socioeconomic rights are directly enforceable, and some are left to legislative discretion (and can thus be viewed as targets of the state). The Polish Constitution serves as an example. It contains a general limiting clause, similar to the one in the Czech Charter and the Slovak Constitution. This applies only to a select number of socioeconomic rights. These rights (which include the rights to a minimum wage, full employment, and aid to disabled persons) “may be asserted subject to limitations specified by law”.81 On the other hand, the constitution lists a number of socioeconomic rights to which the general limiting clause does not apply, even if they have their own clauses attached that delegate the duty to determine the scope and form of implementation to the legislature (the right to social security82 is an example). The fact that these particular socio-economic rights were deliberately left outside the scope of the general limiting clause suggests that they are seen as fully enforceable rights. In fact, the Constitutional Tribunal’s jurisprudence and authoritative doctrine extend enforceability also to those socio-economic rights to which the limiting clause apparently applies. According to the dominant doctrine, the constitutional rules that define the state’s tasks in the field of socio-economic policy have the full normative power of constitutional provisions, and hence can serve as a basis for
180
CHAPTER 7
constitutional review. The authoritative commentary on the statute of the Constitutional Tribunal, when discussing the provisions to which article 81 applies,83 states that the “limitation” stemming from article 81 “applies only to the procedural sphere and restricts the range of means of protection of rights and liberties that belong to an individual. It does not, therefore, undermine the normative character of rights and liberties enumerated in Art. 81, and so there are no obstacles against these rights and principles serving as an independent basis of constitutional review if a proper motion is addressed to the Constitutional Tribunal by an authorised subject”.84 Overall, the above survey shows that nearly all post communist constitutions ignore the distinction in status between civil and political rights on the one hand, and socio-economic rights (either all constitutional socio-economic rights or at least a significant number of them) on the other. This is a messy arrangement. Pretending that socio-economic rights may be enforceable in exactly the same way as the rights to freedom of speech or to vote creates expectations that cannot be fulfilled. It also brings the courts into complex policy-making, and threatens to dilute the enforceability of civil and political rights. As an example of how to reconcile socio-economic constitutional commitments with a clear separation of socio-economic rights and the objectives of the state in the field of socio-economic policy, the constitution-makers of the region could have followed the example of some Western European constitutions, notably those of Spain and Ireland. The Spanish Constitution draws a distinction between “Rights and Freedoms” (Chapter II) and “The Guiding Principles of Economic and Social Policy” (Chapter III). Similarly, the Constitution of Ireland distinguishes between “Fundamental Rights” (arts 40-45) and “Directive Principles of Social Policy” (art. 45), with a provision in the latter to the effect that “they shall not be cognisable by any Court” (art. 45). There were some attempts at similar constitutional design (other than in the Czech Republic and Slovakia where, as just shown, this model was adopted). One such heroic attempt deserves particular acknowledgement: the 1992 “Presidential draft” (so called because it was formally proposed by the then President, Lech WaáĊsa) of the constitutional Charter of Rights and Freedoms, in Poland.85 It clearly distinguished “Social and Economic Rights and Freedoms” (including the right to education, to safety in the workplace, to medical protection, to social welfare, and to freedom to work) from “Economic, Social and Cultural Obligations of Public Authorities” (including, amongst other things, improvement of working conditions, full employment, aid to families, and medical care beyond the basic level).86 There was also an explicit statement that the latter “obligations” were to be performed by public authorities “depending upon their economic resources.”87 This was meant to convey the idea that provisions on “socio-economic tasks” applied to governmental actions and aspirations, rather than to determinate results. As a result, no pretence was made that these tasks and aspirations described a range of constitutional “rights.” This project, however, never became law.
SOCIO-ECONOMIC RIGHTS
181
4. THE DRAWING OF DISTINCTIONS BETWEEN DIFFERENT TYPES OF RIGHTS BY THE COURTS: SOCIAL SECURITY CASES In some of the countries of the region, the task of drawing the necessary distinctions between various categories of rights has been undertaken by constitutional courts. For example, the Hungarian Constitutional Court established in 1990 that the right to social security (art. 70E of the Constitution) does not entitle any citizen to any specific benefit: “social security means neither guaranteed income, nor that the achieved living standard could not deteriorate”.88 As one Hungarian constitutional expert commenting on this decision has suggested, “the interpretation of Chief Justice Sólyom clearly states that social and economic rights are not raised to the rank of subjective rights that can be enforced by the judiciary against the state”.89 In fact, the Court effectively converted the “rights” provisions into targets for the state to pursue. However, this was not an obvious path to choose, and as Justice Sólyom later commented, “On no other question was the Court so divided…”.90 In his expost description of the doctrine that emerged from this decision and throughout the following years, “a method became established . . . according to which an infringement upon the social rights provision of the Constitution could only be sustained if the benefits sank below a minimum level determined by the Constitutional Court”.91 This, somewhat surprisingly, Justice Sólyom understood to be equivalent “to a silent acceptance by the whole Court of the understanding of social rights as being similar to ‘state goals’”. There appears to be a conflict here – clearly not recognised as such by Sólyom himself – between the “minimum level” standard (as determined by the Court) and the view of social rights as merely determining the goals to be pursued by the state. A more recent exposition of this doctrine can be found in a decision on the rate of pension increase at the end of 1999.92 In this case, it was held that the state has a constitutional obligation to maintain a social security system, but that the detailed rules of this cannot be derived from the Constitution itself. In the situation of pension increase, pensioners have no right to a specific rate, although it would be unconstitutional if the state did not provide any increase at all, or if the rate of increase was established arbitrarily. The original rule of the pension law of 1997 provided that social security pensions would be increased in January 1999, to adjust their level to the net growth of income projected in 1998. However, the 1999 budget amended this increased rate, and established that pensions would be raised by eleven percent above the inflation; this would have resulted, according to the petitioners, in a smaller rate of increase than if the original rule had been maintained. The petitioners argued that the amendment violated vested rights. The Court explained, however, that the state has a constitutional obligation to maintain a social security system (Art 70/E of the Constitution), but the pensioners have no vested rights in the rate of increase. Hence, the reduction of increase rate did not violate a vested right. However, the fact that socio-economic rights are not directly enforceable by the courts93 does not prevent these rights from becoming grounds for constitutional challenges to laws and policies through the process of abstract judicial review. It is one thing for the court to say that a specific individual has a legitimate claim to a
182
CHAPTER 7
particular good, and another to find a particular governmental policy unconstitutional. Indeed, the dominant opinion in post- communist constitutional doctrine is that all constitutional provisions, including those that contain socioeconomic rights, can be used as yardsticks to assess the constitutionality of statutes.94 As a result, constitutional courts have been quite active in reviewing, and at times invalidating, statutes under the standards of socio-economic rights, even though this often calls for judgments on social and economic policies, in which judges have no expertise to provide adequate review. A view that these rights are “merely programmatic,” and thus non-justiciable,95 has never become a dominant, recognised doctrine. For instance, in Poland the Constitutional Tribunal began striking down some laws on the basis of violation of certain socio-economic rights even before the democratic transition, and thus before amendments to the “socialist” constitution had occurred. As an example, consider the decision of 30 November 1988, in which the Tribunal reviewed and partly invalidated the law of 14 December 1982, which revised the rules relating to acquiring disability pensions for the handicapped.96 The petitioners (trade unions) objected to the fact that the law limited the eligibility for those benefits (by, among other things, increasing from five to ten years the required period of earlier employment for those employees who took up their first job after 40 years of age). The Tribunal considered the compatibility of this reduction of benefits with art. 70 of the Constitution which was valid at the time (social security). In particular, it considered the petitioners’ argument that the constitutional provision for an “even fuller implementation” of the right to social security prevents the legislator from ever reducing benefits; in other words, a “ratchet” theory (without using this word). The Tribunal rejected this argument by saying that such violation would occur only if several consecutive laws issued over a “longer period of time” denied the principle of steady improvement of citizens’ rights; it was therefore not the case that art. 70 was violated if a particular provision, taken in isolation from others, caused deterioration in the benefits derived from a particular entitlement.97 However, in this case, the Tribunal found that this particular provision “has deprived a certain group of people of their right to a disability pension, which has been formed during many years of development of the pension system in Poland, without at the same time establishing any other benefit entitlement”. For this reason, the new rule violated art. 70 because this constitutional provision entrenched a “ratchet” requirement upon the legislator.98 (It should be added that, in the years to come, the Tribunal consistently resisted invitations to adopt a “ratchet” theory of socioeconomic rights; in a 1997 decision on family benefits, it observed that it would not be rational for the law-maker to adopt the new economic system – meaning, market economy – and at the same time enact rules that would render impossible the reforms necessary to bring about this new system).99 The Tribunal, in this pretransition period, also found a violation of the constitutional principles of social justice and non-retroactivity of the law. Despite this, at this early stage of its activity, the Tribunal still emphasised the by-and-large “programmatic” character of the socio-economic provisions of the Constitution.100 With time, however, it abandoned the doctrine of “programmatic” rights, and unequivocally accepted that those rights
SOCIO-ECONOMIC RIGHTS
183
may serve as a basis for decisions invalidating statutes on the grounds of unconstitutionality.101 This approach adopted by the Constitutional Tribunal applies not only to the abstract review of statutes but also to concrete review. In Poland, the view that “programmatic constitutional norms” can be the basis of citizens’ claims and complaints has become dominant; the authors of the Commentary on the Law of the Constitutional Tribunal explicitly reject the view that the so-called “programmatic norms” of the Polish Constitution cannot serve as the basis for “constitutional complaint”. As they say, a breach of a “programmatic norm” happens when “the legislator incorrectly interprets a provision of the Constitution that defines a particular goal or task of the public authorities, and, in particular, has enacted a statute that provides for measures that cannot lead to that goal and thus breached constitutional liberties or rights”.102 This view (which has the authoritative backing of three judges of the Polish Constitutional Tribunal, who jointly authored the Komentarz) effectively blurs the borderline between “programmatic norms” and “claim rights”. However, this is not the unanimously accepted doctrinal position. There is also an influential view that the issue of the enforceability of socioeconomic rights should be distinguished from the issue of socio-economic rights as a ground for judicial review, and that “programmatic” rights are essentially different from claim rights. This is meant to suggest that the constitutionalisation of the right to work, housing, health care etc. does not authorise citizens to press any specific claims against the government in court, but merely imposes a duty upon the government to conduct an effective policy aimed at the fulfilment of these programmatic goals. In this sense, these rights are not directly enforceable, or selfexecuting. A leading Polish proponent of this view, the late Tadeusz ZieliĔski, distinguished between “claim rights” and “programmatic rights”, with the latter “defin[ing] the tasks of public authorities in the area of welfare rights of citizens.” In addition, “[a] right to work means only that a citizen has a right to assistance in finding a job by the public authorities. A right to lodging means only that a citizen is provided the opportunity to make use of policies leading to satisfying citizens’ needs for lodging.”103 However, if all there is to a right is an opportunity to benefit from whatever state policy is in operation, it is redundant to call it a “right”; it is, rather, another way of urging the government to have a policy in this field. For another example of enforcement of the right to a pension, consider a decision of the Bulgarian Constitutional Court on pension entitlements, which held that restricting this right when another constitutional right, namely the right to work, was being exercised was unconstitutional.104 In this decision, articles 50(1) and 59(2) of the Pension Act (a law adopted by the left wing government of the Bulgarian Socialist Party) were challenged. According to those provisions, pension entitlement would be withdrawn from all pensioners that had an earned income (i.e. those that worked and got an income from this). The Constitutional Court held that this was contrary to the right to social security (art. 51(1) of the Constitution), because pensions are a type of social security entitlement covered by Art. 51(1). Rights can only be restricted in accordance with the Constitution, but the Constitution provided for no restrictions to this right. The linking of the entitlement to a pension to having no earned income was unreasonable, according to the Court; the right to social
184
CHAPTER 7
security is a totally separate right from the right to work, and the former cannot be made dependent upon whether the latter is being exercised or not. For this reason, the challenged section of the law was held to be unconstitutional.105 As Justice Todorov explained recently in an interview, two conflicting theories were raised in the Court’s deliberations.106 One – which gained support of the majority – was more “civilistic” (in Todorov’s words), and assumed that the state has an obligation to pay a pension to any person who has been working and “making his contributions to the budget”. This obligation has a “contractual” nature; it is a contract between the person concerned and the state. As Justice Todorov acknowledged, he supported this theory because, in his words, “I am inclined to reason like a civil lawyer”.107 The second theory (espoused by two dissenting justices, Milcho Kostov and Aleksandr Arabadjiev, the rapporteur in this case) presupposed, in the words of Justice Todorov, that “the money for pensions, as a whole sum, is produced by the generation living in the moment of paying the pension”108 (therefore, a “pay-as-you-go” model of pensions). As the financial and budgetary situation may change, the obligation of the state to pay the pension depends not upon the contribution received from each individual, but rather on the ability of society to provide it. The Court, according to Justice Todorov, chose the former theory, and decided that the state cannot free itself from paying the full amount of pensions, even if this creates financial difficulties. “Thinking about it again, I am not so convinced about this position any more”, Justice Todorov admitted a few years after the decision had been taken, but he stressed that the original view of the majority of the Court had been affected by two considerations. Firstly, the majority judges were concerned that “if you say that the state can determine the pension, you give a very broad scope of discretion to the state”; after all, the government always argues that the financial situation is very difficult.109 Secondly, there was strong pressure from public opinion; the situation of Bulgarian pensioners was very difficult, the pension stipend being around 70 Leva per month.110 In general, according to Justice Todorov, there was a conflict between two constitutional principles: the “liberal principle” and the social state, the latter demanding a degree of welfare and distributive justice. However, as a result of the dire economic situation, the principle of the “social state” remains “a dead letter”.111 An interesting feature of this case is that, while the Court had no qualms about invalidating a government policy on the grounds of a socio-economic right, it chose to rest its decision upon a more “legalistic” philosophy of acquired benefits and contractual obligation rather than upon a theory of social justice of a distributive kind. Ironically, the choice of the former led, in this particular case, to a less deferential attitude towards state policy. This is because the austerity measure under challenge here might well have been defensible based on the theory that, in the context of drastic economic scarcity, those fortunate enough to have paid employment should forsake their pensions in order to benefit those who cannot work. On the other hand, it is significant that, when constitutional courts in the region have had a choice between striking down a law under a general constitutional clause such as “social justice” or “equality” on the one hand, or under a specific welfare
SOCIO-ECONOMIC RIGHTS
185
right on the other, they usually have opted for the former solution.112 This is a symptom of a certain malaise over the direct enforcement of socio-economic rights. As an example of such a strategy, consider a 1996 Slovenian decision on the reduction of extra benefits for disabled soldiers.113 In February 1994, a new assessment basis for such benefits came into effect, and as a result the payments were significantly reduced (by about a quarter). The Constitution provides that war veterans and civilian casualties of war are guaranteed special benefits from the state (Art. 50 (3)), which means that they should receive more than the basic level of social security. The Court admitted that amending a law related to benefits was not unconstitutional in itself. However, the principle of trust means that the state will not worsen a person’s position without real cause grounded in the legitimate public interest. The extent of the reduction of the benefits, and the absence of any transitional period in which they were lowered gradually, meant that the State had behaved in a way inconsistent with the general criteria of a “law-based State”, or Rechtsstaat (Art. 2 of the Constitution). Thus, it was unnecessary – according to the Constitutional Court – to analyse whether there was a violation of the specific constitutional provision on benefits to disabled soldiers.114 The Court clearly preferred such a strategy to that of a direct appeal to a specific welfare right. At first blush, this strategy may seem surprising. After all, general clauses such as those guaranteeing a “law-based state”, the principle of trust or of fidelity to vested rights, seem to be more vague than specific socio-economic rights. The fact that constitutional courts usually prefer the former, notwithstanding their lofty character, may be significant. It shows that the courts themselves feel that they are on rather shaky grounds when telling governments or legislatures what, to whom, and how much, should be paid or supplied to citizens as a result of their socioeconomic constitutional rights. This perhaps confirms the opinion that, had constitution-makers opted for a solution under which the welfare interests of citizens belonged to the category of constitutional “targets” (with the clear implication that, as such, they are not cognisable by the courts), much clarity could have been gained.115 5. THE RIGHT TO WORK The “right to work” has undergone a dramatic transformation since the fall of Communism. The advent and growth of private industry (for example, in Slovakia, 75 % of the GDP was produced by the private sector by 1998; in Slovenia, this figure was 55 %)116 and the generalised experience of unemployment,117 rendered hollow the idea that the state has a duty to provide all those able and willing to work with meaningful employment. Within a market economy, even one that provides a significant role to an “activist” state, the “right to work” may at most mean (as is the case of the European Social Charter, and of some Western European Constitutions, e.g. the Constitution of Italy, Art. 4) that the state has a duty to create a policy aiming at full employment, to protect the opportunities of every worker to earn their living in the occupation of their choice, to run vocational training and employment services, etc, and to provide social assistance to those unable to find work. However,
186
CHAPTER 7
as Bob Hepple has argued, the “right to work” cannot be meaningfully understood as giving birth to a state duty to provide employment, because “such a ‘duty’ would simply beg the question how that employment was to be created”. As he adds: “We cannot expect ambiguous legal ‘rights’ of this kind to succeed where social and economic policies of full employment have failed”.118 This has been reflected in the constitutional treatment of the right to work. Even though roughly a half of the CEE constitutions contain simple “right to work” provisions,119 these have nevertheless come to be understood not as proclaiming a right to be provided with employment by the state. Constitutional courts have played an important role in this re-interpretation; for example, in Hungary, the right to work was identified by the Constitutional Court with the right to free enterprise; further, in negative terms, the Court stated that this right secured no “subjective right” to obtain a given job.120 In the remaining constitutions, no express “right to work” is specified; rather, the right to freely choose one’s own occupation has been proclaimed.121 It should be added that, in addition to those provisions, most constitutions impose upon the State some positive duties with regard to employment, such as to provide training,122 to create “conditions for the exercising” of the right to work,123 or to pursue policies aimed at full employment.124 The difference between the “right to work” in some constitutions and the mere right to free choice of occupation in others has not resulted in different constitutional (or, indeed, socio-political) practices. In Poland, the former right was inscribed in the Constitution until 1997;125 by then, however, the “right to work” had come to be interpreted by the Constitutional Tribunal not as an individual’s claim to be employed, but only as freedom to choose work, and – in the jurisprudence developed still under the old Constitution – a right to be rewarded in accordance with the “amount and quality” of the work, but only in those areas where the state directly controlled the salary of an employee.126 Further, there exists the correlated right to unemployment benefits. For those already employed, however, the Tribunal held that a requirement to pay at least a certain, minimal salary was a constitutive part of the right to work itself. Consequently, paying employees at a level lower than the minimum rate (as determined by the Minister of Labour) is a breach of the constitutional right to work. This rule was announced in a decision that concerned a somewhat specific situation, namely, the employment of prisoners.127 While, in general, the Tribunal determined that the general rules of labour law do not apply to the employment of prisoners (because this is not based on contracts of employment), nevertheless the rates of minimum pay do apply to the prisoners’ labour, on the basis of the constitutional right to work. This, as one can see, is the interpretation that stretches the very concept of the right to work quite far. The most charitable justification for this decision is that the Tribunal wanted, properly, to grant a degree of protection to a particularly vulnerable and defenceless category of “workers”, as far as payments for labour are concerned, and that the best constitutional provision that could serve this purpose was the right to work. However, it is ironic because the most straightforward reading of this “right” suggests the voluntary element of employment; and it is precisely this that is missing in the case of prisoners. Under the new Polish Constitution, in which the “right to work” has been finally dropped and replaced by the right to freedom of
SOCIO-ECONOMIC RIGHTS
187
employment and to minimum pay,128 the focus of the interpretation by the Constitutional Tribunal has shifted; it now centres on the qualifications sought and the conditions of hiring or of conducting business activity. Are the qualifications required for certain types of jobs or business activity a limitation on the freedom to work? The answer, of course, depends on the relevance of those qualifications to the type of job or business in question; such assessments, however, are eminently controversial, as are the legislative choices of parliaments and the decisions of constitutional courts on this matter. One such decision was the Polish Constitutional Tribunal’s invalidation of a statutory requirement forbidding the employment as taxi drivers of (or granting licenses for other road transport businesses to) those with a criminal record.129 The Tribunal found no proper relationship between this condition and any legitimate purpose served by such a regulation. It therefore found the regulation to be an excessive restriction of the constitutional right to work. It remarked, sensibly, that the “no criminal record” requirement may be relevant in a number of positions, particularly when they are connected to the exercise of public authority, and/or require a degree of public trust. However, allowing such a condition in this particular case would lead to the possibility that the “no criminal record” condition could be extended to a great number of positions and businesses, with the negative side-effect of making it excessively difficult for people with a criminal past to return to a law-abiding way of life in society.130 As I have remarked, this was a sensible judgment but, at the same time, it was controversial. It nicely illustrated the fact that, by combining constitutional socio-economic rights with active judicial review, the door is opened for a non-representative body to displace the controversial judgments of the legislature, on which reasonable people may, and do, disagree. The arguments of the Tribunal are serious and reasonable; one may, however, oppose them with the (equally reasonable) motives that the legislators had (or might have had) in enacting this barrier to obtaining a position as a taxi-driver (or other transport business). If one looks at actual reality, and notes the declining standards of personal behaviour and honesty of taxi drivers, the alarming conditions of road safety, and the high supply of candidates for these jobs, then the inclusion of the “no criminal record” requirement as one of the screening devices for hiring in this field ceases to look so unreasonable. Incidentally, the right to choose one’s profession was discussed by the Hungarian Constitutional Court, also in relation to taxi drivers’ work. In a 1993 decision, the Court struck down a provision of the statute enabling local governments to limit the number of taxi licenses.131 The main constitutional provisions under which the Court considered the matter were the right to free enterprise (Art. 9(2)), and the principle of market economy (art. 9(1)). While the Court stated that the setting of conditions for obtaining taxi licenses was not unconstitutional per se, it nevertheless held that the placing of restrictions on the number of taxi licences available conflicted with these constitutional provisions. In addition to this, the Court found the limits to be unnecessary and disproportionate restrictions of the right to choose one’s profession (Art. 70(b)(1)). The question of occupational qualifications, in terms of potential infringements of the right to free choice of occupation, has also been raised before the Slovenian
188
CHAPTER 7
Constitutional Court. In a 1995 decision, the Court considered a provision of the Notaries Act that specified the conditions that must be met by all persons wishing to become notaries.132 One of these conditions was that the person must be worthy of public trust; and it further stated that a person is not worthy of such trust if he has been the subject of criminal proceedings for a crime that would make him morally unworthy to perform such work, has been convicted of such a crime, or behaves in such a way that it can be concluded that he will not honestly and conscientiously perform his duties as a notary. In its discussion of these conditions, the Court declared that the office of notary is a public service, and therefore a restriction of the constitutional right to freely carry on work or a profession (Article 49 of the Constitution) serves an important public good; the protection of the rights of others to legal security in legal business, which a notary guarantees. (The Court did not fail to mention that this limiting criterion was different from that which ruled under the Communist system, which “allowed privilege or discrimination in relation to ideological or political convictions and activity”).133 Given that the Constitution allows the limitation of Constitutional rights to protect the rights and freedoms of others (Article 15), this limitation of the right to obtain the position of notary was found to be constitutional. As the Court noted, “a guarantee of [the candidates’] honestly and conscientiously performing the notary profession must . . . be established prior to the appointment as a notary”.134 Interestingly, however, the Court struck down one part of the challenged provision; namely, the section that stipulated that people who have been subject to criminal proceedings will be deemed morally unworthy. The Court held that this contradicts the presumption of innocence in that, prior to a court judgement, no negative consequences should be created for the accused and, most certainly, their constitutional rights should not be restricted.135 The Lithuanian Constitutional Court has dealt more than once with the right to free choice of occupation. In a lengthy and complex decision in 1999, it examined the situation of former employees of the Soviet security forces being barred from holding certain posts in the Republic of Lithuania.136 I discuss this decision elsewhere in this book;137 the only provision of the challenged law that the Court actually invalidated was the scheme for setting up a presidential commission empowered to waive the application of the general rule of the law: this was found to offend the principle that matters regarding restrictions of constitutional rights must be decided by statute rather than delegated to the executive branch. However, on the central substantive issue, the Court found no constitutional defect in the law that imposed checks on persons holding influential positions in public life, in order to determine whether they had had any ties with the secret services of the former communist regime. In a much more recent decision, the right to choose one’s employment was also raised, albeit marginally and indirectly.138 The provisions under challenge were two articles of the Laws on the Bar establishing the rules of incompatibility for the practising lawyer (advocate) in a given court; namely, if they had worked in that same court as a judge within the last three years, or if a close relative was working as a judge in that same court. The central constitutional provision upon which these rules were challenged was the right to legal counsel in criminal cases (art. 31 (6)), which implies also the right to choose one’s advocate. The Constitutional Court rejected this challenge, finding that although the
SOCIO-ECONOMIC RIGHTS
189
incompatibility rules did indeed restrict the right to the choice of one’s lawyers, they did so properly, in order to ensure the impartiality and independence of courts and judges, without at the same time removing altogether people’s ability to choose a lawyer. Another ground of the challenge was art. 48 (2) of the Constitution, which provides that “Every person may freely choose an occupation or business…”. However, according to the Court, the law contained no restrictions on a person becoming an advocate, as they did not relate to the possibility of attaining this post (which is a choice of occupation) but merely restricted some of the functions of advocates in certain courts. 6. RIGHTS TO HEALTH AND EDUCATION The right-to-health provisions have only rarely been applied in the jurisprudence of the constitutional courts of CEE thus far. One example was a Croatian Constitutional Court decision of 1998, in which a provision of the Law on Health Insurance was successfully challenged.139 The provision denied health protection services (except for urgent medical help) to those patients whose contributions to medical insurance had not been made. Considering that such payments were the obligation of the employers, not the patients themselves, the failure to make such payments may not have been the fault of the employees, and thus the denial of health protection to them was found to be in violation of the constitutional right to health care (Art. 58). This is an example showing how an old right, with a state socialism pedigree, became quite crucial in the new economic situation in which the state has ceased to be the main employer and, in consequence, is no longer the provider of quasiautomatic health insurance to all those employed. With the introduction of health insurance funds, and a partial privatisation of health services, the constitutional right to health takes on the function of protecting patients against arbitrary discrimination and exclusion from the benefits of health care, rather than of supporting a claim for universal provision of such services. As mentioned earlier, some CEE constitutions provided for the right to free public education. This principle ran into some practical difficulties, and, again, it was the job of the constitutional courts to narrow down its meaning. Consider the example of Poland. Here, the constitutional text was rather ambiguous: Art. 70 para. 2 provided, in the first sentence, that “Education in public schools is free”, while the second sentence allowed a statute to provide for “rendering some educational services by public universities for fee”. Many understood that those “educational services” applied to mature and distance-education students who attend separate courses, and that the first sentence of this provision prevented the imposition of fees upon day students admitted in the regular selection process. Despite this, the Constitutional Tribunal upheld statutory regulations that provided for tuition fees to be paid by regular students in public universities, as long as the universities also provide regular education to a certain number of students for free.140 The argument of the Tribunal went as follows: the right to education (art. 70) is, in its essence, a guarantee of “availability and universality” of education, not of it being free of cost. To be sure, the second paragraph of the article provides for “free education in public
190
CHAPTER 7
schools”,141 but this is only one of the elements of that right and should be interpreted as being derivative from and instrumental to the principle of “availability and universality”. The guarantee of free university education is not absolute and unlimited. Article 70 paragraph 2 cannot be interpreted as conferring a right to free university education to everyone who meets the formal requirements to be a student. Only those who meet additional criteria, established in the selection procedures of particular universities, can be beneficiaries of this right. Public universities have a duty to implement the principle of universal access to education through different types of study programmes; both free (as the fundamental form) and also other forms of studies with respect to which students themselves would be expected to contribute financially. In deciding on the tuition fee, the actual costs of study play the central role, and the fees cannot apply to that part of the educational work of the university that is fully financed by public funds. The Tribunal also rejected an argument that the introduction of fees in relation to only some students violates the principle of non-discrimination (Art. 32 of the Constitution) “on the condition that the introduction of tuition fees was brought about in order to assure as wide access to education as possible”. As can be seen, the Tribunal tried to reconcile the general principle of free education with a realistic understanding of the dire economic situation of universities, starved for funds. The argument proceeds on the basis of the unspoken assumption that to introduce full fees for at least some regular students (but not to all, as this would be unconstitutional) is a Pareto-optimal solution because the only other alternative option would be to reduce the student intake; thus, those who now have to pay would not have been admitted at all if the universities had been compelled to offer education to non-paying students only. This seems like a common-sense solution; the downside, however, is, of course, that it creates incentives for universities to steadily reduce the number of places offered for free, and thus, in practice, to endanger the very principle of universal access that is posited as the essential meaning of the Art. 70 guarantee. 7. CONCLUSIONS For various reasons mentioned at the beginning of this chapter, constitution-makers in CEE did not really face the choice of whether to constitutionalise socio-economic rights (notwithstanding the invitations by some foreign, especially American, experts to leave these rights outside the new constitutions). They rather had to decide how many rights, which rights and how these rights should be constitutionalised. As a result, CEE constitutions present reasonably broad catalogues (with some variants, identified earlier in this chapter) of these rights, which correspond to the concerns, basic needs and legitimate claims of people in the field of social policy in an age of economic transformation resulting in the collapse of some old certainties, such as full employment (even if often accompanied by socalled hidden unemployment, hiding a de-facto unnecessary employment), paternalistic social services offered by state-owned enterprises (including child-care, leisure etc), state-provided free education, health and social services (even if often at a
SOCIO-ECONOMIC RIGHTS
191
lamentably low level, forcing the beneficiaries to pay the difference in order to obtain services of satisfactory quality), etc. The end of these certainties focused the attention of the general public upon the socio-economic rights written into their constitutions. The contrast between a plain reading of certain “promises” contained in the constitution and the grim reality of a state unable to fulfil these promises made socio-economic rights the most hotly contested area in the field of constitutional rights after the fall of communism. The sheer number of people having a stake in the vindication and implementation of these rights has contributed to the size of the problem; in Poland, for example, there are nine million people claiming old-age pensions or disability benefits in a total population of thirty-nine million. In a situation in which governments, struggling with budget deficits, were forced to restrict spending on social services, the potential for challenge centred on the socioeconomic rights written into the various constitutions was obvious. These challenges naturally found their expression in the case law of the constitutional courts in the region. Whether they wanted it or not, these courts have been forced into the domain of socio-economic policy-making under the guise of defending constitutional rights. This is particularly so in those countries (such as Hungary or Poland) in which the constitutions themselves have not drawn any meaningful boundaries between those rights that are and those that are not enforceable by the courts in the process of constitutional review. The courts have found themselves, as a result, subject to conflicting temptations, incentives and pressures. On the one hand, the combination of societal pressure to invalidate a “heartless” policy, the pursuit of popularity, an understanding of socio-economic rights as giving the courts a legitimate entry into the domain of socio-economic policymaking, and the probably sincere belief that all they were doing was applying constitutional strictures has encouraged the courts to displace decisions and policies of governments and parliaments in this area. On the other hand, the rhetoric of judicial restraint, awareness of the complexity of the issues involved (including an awareness of the possibility of negative side-effects of their decisions upon those whom they purported to benefit), a sense of their own low institutional competence and technical expertise in the field, and a “minimalist” understanding of the function of socio-economic rights have often combined to constrain the courts from entering this minefield. In the face of such conflicting pressures and incentives, it is unsurprising that the overall picture of the output of constitutional courts in this field is ambiguous. Regardless of whether one welcomes or deplores the intrusion of these courts into the arena of socio-economic rights, one must admit that, at times, some decisions of great economic importance have been handed down. The decisions on pensions in Poland142 and Croatia,143 to mention only two countries, frustrated to a very high degree the plans of the respective governments for reform of the social security system. On the other hand, there were cases in which the constitutional courts did not intervene, even though there was a textual basis in the constitution for striking down a law or policy. For example, when the parliament in Lithuania adopted, in 1997, a national health insurance scheme based on employee contributions, it was not invalidated by the Constitutional Court, even though it might have been seen to contradict the constitutional right to free health care.144
192
CHAPTER 7
There were also numerous instances in which the constitutional courts did intervene (as they were constitutionally compelled to), and reinterpreted the right in question in such a manner as to make it compatible with the new reality of a socioeconomic system in which the state no longer dispensed all of the benefits once covered by the meaning of a constitutional right. They thus rendered the right more realistic, more relevant, and more appropriate considering the systemic changes to, and the fiscal realities of, the post-communist state. The decision of the Polish Constitutional Tribunal on the right to “free” education145 is a good example of such a re-interpretation in light of a clash between the text of the constitutional right (itself a product of populist constitution drafting) and the realities of starved-forfunds public universities. To prevent those universities from charging some students for tuition would, in effect, harm the very people whom the constitutional right in question was meant to protect. If a university has to make a choice between restricting the number of enrolments (because of scarcity of public funds and inability to charge tuition) or charging some students for tuition (on the condition that the rules are clear and non-arbitrary), and given the strong persistence of formal and informal mechanisms of social reproduction that ensure that the student body is largely composed of children whose parents themselves went through higher education (and are thus, typically, more affluent than others), the less advantaged may actually be better off in terms of access to education if they can buy their way into a public university rather than if its doors would remain closed to them. While there is no one obvious answer to the dilemma faced by the Constitutional Tribunal in this decision, the outcome that it reached is not obviously unreasonable, and led to a partial reconciliation of the constitutional text with prevailing socio-economic realities. For this reason, I am unable to agree with a leading expert on post-communist constitutionalism, Professor Andras Sajo, who deplores the fact that post-communist constitutional courts have perpetuated the wrong welfare policies.146 Sajo’s claim is that, upon entering this domain, the courts have supported the status quo in the field of welfare, rather than transformative welfare policies, and that they have consistently opted for an egalitarian-distributive, rather than a merely corrective, conception of justice. For one thing, I would not characterise all policies that look as if they are based on distributive justice (e.g., universal as opposed to individualised free services, the provision of some services in kind rather than in cash, pay-as-yougo social insurance schemes as opposed to fully funded, contribution based schemes, etc.) as necessarily irrational, in the same way that State-sponsored, mechanical egalitarianism is irrational. I would see some of them at least as being a kind of insurance scheme (though based on Dworkinian hypothetical insurance, because there have not been conditions for real insurance).147 Others, I would classify as examples of paternalism, but cases in which paternalism is not necessarily an objectionable policy but rather a rational response to the lack of knowledge or rationality of the purported beneficiaries, as in the right to compulsory and free education up to a certain stage. However, and more importantly from the point of view of the discussion of CEE post-communist social policy, I do not think that the distinction between corrective justice (safety net) and distributive justice is as sharp as Sajo implies.148 I would
SOCIO-ECONOMIC RIGHTS
193
venture a thesis that the lower the overall level of general welfare in society, the less significant is the distinction between corrective and distributive justice. This is because the absolute level of welfare is so low, and the number of people in need so high, that distributive transfers through state provision of welfare services are likely to benefit those in need. There will always be, of course, a degree of over-inclusion (in terms of some benefits going to those who do not “deserve” them in terms of corrective justice) but its cost has to be balanced against the costs of administering a highly individualised system of corrective justice. These include, to begin with, the obvious administrative costs, for example of monitoring and means testing (corrective justice prefers means tests to universal eligibility criteria). Further, there is a cost in the form of moral hazard – incentives to cheat – and the deterrence of this. Finally, there is the possibility of “perverse incentive” effects. For example, a decision to refuse a pension to those pensioners who undertake paid work – a typical corrective justice device – may create a counter-incentive against pursuit of employment by retirees who are able to work. This was the subject matter of the decision by the Bulgarian Constitutional Court in 1997, discussed above.149 The law in question would make the right to a pension contingent upon not having any other income. It is worth recalling that the Constitutional Court struck down this law as making an impermissible link between the exercise of a property right (pension) with the surrender of the exercise of another right (the right to work). The Court’s action in this case indeed resembles a “distributive” rather than “corrective” measure, in Sajo’s distinction. A purely and properly-termed “corrective” measure would be to confine pensions only to those who truly need them, and therefore who have no other income. However, in terms of incentives, the Court’s action makes good sense, because to uphold the law would be to produce counterincentives to retired job seekers. Has the activity of the constitutional courts – in interpreting welfare rights and in articulating the meaning of other provisions with consequences for welfare – been uniquely status-quo-maintaining? This is Sajo’s thesis,150 and, in his article, he supplies important evidence to illustrate this pro-status-quo effects of their judgments (the doctrine of vested rights, etc). He additionally explains that the logic of representative democracy reinforces the incentives of the governing elites to cater to the interests of groups interested in the maintenance of the status quo.151 Then again, it is certainly arguable that democracy and civil and political rights can also play a role (although, perhaps, a weaker one) in influencing socio-economic policy in the direction of change; and some credit for protecting these must be given to constitutional courts. János Mátyas Kovács has coined the notion of “invisible welfare”, that is, welfare promoted by the new liberties.152 Democracy and the rule of law, claims Kovács, promote the establishment of new welfare institutions, from trade unions to private kindergartens, and the introduction of new social policies, from openly acknowledging poverty to allowing patients to be cared for by the doctor of their choice. These are examples of the potentially transformative impact of civil liberties on welfare. It has to be admitted that, by and large, constitutional courts in the region have a positive score card when it comes to the protection of constitutional civil and political rights, regardless of the occasional bad decisions in
194
CHAPTER 7
which rights-restrictive laws have not been struck down or rights-protective laws have been invalidated, as shown in Chapter 6 of this book.
CHAPTER 8 EQUALITY AND MINORITY RIGHTS
Clauses on equality and anti-discrimination have been among the most frequently used constitutional provisions in the process of constitutional review in CEE. This is understandable. Firstly, these provisions are eminently malleable, “open-ended”, and lend themselves to application in virtually all spheres of law and policy. Secondly, when constitutional courts are identifying the substantive rights provision under which they will consider a particular challenge, they will often have a choice between a socio-economic right and an equality provision. The selection of the equality provision as a basis for review has the advantage of appearing less controversial, less “activist” than if a specific socio-economic right were to be appealed to directly.1 Further, the rhetoric of equality has been particularly popular in the post-communist states of CEE due to the emergence of greater inequality associated with the introduction of the market economy, coupled with deeply ingrained egalitarian societal attitudes. This explains the inclination of constitutional courts to use the equality provisions generously, a tendency that will be discussed in sections 1-3 below. In contrast, however, to the general equality and antidiscrimination provisions, the constitutional rules on minority rights, and, in particular, on the rights of ethnic and national minorities, became something of a problem for constitutional courts: they were, by and large, neither intellectually equipped nor morally and politically prepared to interpret them in an expansive, generous manner. This will be discussed in sections 4-7 of this chapter. 1. EQUALITY AND CONSTITUTIONAL REVIEW It is useful to sketch at the outset a brief account of the design of equality provisions in the constitutions of the region. Two techniques are used: either a general statement on equality (or a prohibition of discrimination; the two concepts are treated interchangeably), or a general provision on equality with a list of specific grounds of prohibited discrimination. The first technique is used in a small minority of constitutions in the region: Poland, Latvia and Belarus belong to this category. For example, the Polish Constitution proclaims: “All persons shall be equal before the law. All shall have the right to equal treatment by public authorities. No one shall be discriminated against in political, social or economic life for any reason whatsoever”.2 As is readily evident, the prohibition of discrimination is worded more broadly than the right to equal protection (the former being applicable also “horizontally”, not only vis-à-vis public authorities), but no grounds for unlawful
196
CHAPTER 8
discrimination are spelled out. In consequence, the reach of the anti-discrimination rule is theoretically unlimited, and therefore its vagueness is obvious: what important phenomena cannot be subsumed under the notions of “political, social or economic life”? Similarly, the Latvian Constitution states: “All persons in Latvia shall be equal before the law and the courts. Human rights shall be realised without discrimination of any kind”. 3 Again, no hints as to what constitutes “discrimination” are provided, but the vagueness of the prohibition is somewhat reduced by limiting it to the protection of “human rights” only, and not to any and all legitimate interests of individuals. In contrast, a great majority of the CEE constitutions attach a list of specific grounds of prohibited discrimination to the general provisions on equality. These characteristics are strikingly similar: of the seventeen constitutions belonging to this category, all prohibit discrimination based on race, gender and religion; all but one mention language, political opinion and social status; and all but two mention nationality. Other frequently listed grounds of prohibited discrimination include nationality (in fifteen constitutions), property (thirteen) and ethnic or other origin (eleven). Less than half of those constitutional provisions mention birth (eight) and education (six) as grounds of prohibited discrimination. It is interesting to note that there is one great and obvious absentee, namely, sexual orientation. It is significant that not one of the constitutional charters of rights in CEE that make an effort to list the grounds of prohibited discrimination mentions sexual orientation. This clearly reflects the widespread ambiguity – often, outright hostility – in attitudes towards homosexuality in CEE.4 There were powerful counter-incentives at work, preventing constitution drafters in the region proposing, or supporting, the inclusion of an explicit prohibition of discrimination based on sexual orientation: such a proposal was unlikely to find favour among the majority of voters, and certainly would have been ridiculed and rejected by powerful traditionalist groups (such as the Catholic church and related political forces). It does not, however, necessarily follow that anti-homosexual legislation automatically passes constitutional muster in those countries: indeed, decisions by the Romanian5 and Hungarian6 Constitutional Courts have already shown this not to be the case. But in order to find constitutional arguments against such proposed or actual legislation, its opponents must look to the general equality provisions rather than to the list of specific prohibitions of certain grounds of discrimination. In fact, this omission is not surprising, given that a number of the constitutions that do list specific grounds of prohibited discrimination provide, at the end of these lists, that “any other reasons” for discrimination are also prohibited, thus explicitly acknowledging that the list provided is non-exhaustive.7 Anti-discrimination provisions in these constitutions are therefore “open” in two ways: firstly, by allowing any other, unspecified reasons, to count as impermissible grounds for discrimination; and secondly by attaching the lists of specific grounds for prohibited discrimination to the more general provisions on equality and discrimination. As the principle of equality is violated by any classification that relates a specific treatment to personal characteristics that are irrelevant to that treatment, or where they may be relevant but there is no proportionality between the characteristic and the different
EQUALITY AND MINORITY RIGHTS
197
treatment,8 the lists of prohibited grounds for discrimination may be seen as redundant. What role, then, do they actually play? One can venture two hypotheses. The first is that, notwithstanding the open character of the prohibition of discrimination, some types of discrimination are viewed as being worse than others, with those that treat people unfairly on grounds listed belonging to this group. Such an explanation would make sense provided that there was a clear and obvious common denominator between those grounds of prohibited discrimination that would point at the common rationale for our particular hostility towards classifications based on that rationale. But there is none. Clearly, the immutability of the listed personal characteristics that form the basis for discrimination is not such a common denominator: among the three universally listed grounds for prohibited discrimination, two are immutable (race and gender) and the third is not (religion; the same applies to political opinion which is listed quasi-universally). The second hypothesis is perhaps more cynical but likely to be more realistic: the catalogues of prohibited grounds serve a rhetorical purpose of symbolically conveying a message to the population, to the effect that the constitutional drafters are not misogynist, not racist, not bigots etc. This is harmless enough; however, it shows that the actual constitutional value of these lists is extremely limited. The Polish Constitutional Tribunal has worked out a relatively elaborate theory of equality-based scrutiny of legislation, focusing on the idea that equality before the law is compatible with the fact of legal differentiation, as long as the differences in treatment are related to relevant differences in those involved. Early on in its judgments, it had adopted the idea that equality before the law does not imply that all rights must be equal for everyone, and that “for the law to be just, it cannot avoid making certain differentiations in the form of particular rules addressed to some groups and classes of citizens. . . . reasoning on equality before the law therefore collapses into an evaluation of adopting a particular classificatory criterion as justified and as just”.9 Perhaps the most developed definition of this position thus far provided by the Constitutional Tribunal is as follows: If the differential treatment of similar subjects introduced by a regulation is one of the purposes pursued by the legislator; if the implementation of these purposes finds its justification in other [i.e. than equality] constitutional rules, principles or values (and in particular, in the principle of social justice); and if the departure from the principle of equality is proportionate to the importance of this purpose, then a different treatment of similar situations cannot be viewed as discriminatory (or privileging).10
As is clear, the identification of the “relevance” of a subject’s particular characteristic to the goal pursued is the cornerstone of this analysis. The question of what characteristics are relevant to particular purposes is however, an essentially open-ended one, inevitably implicating issues of either morals or policy; thus, scrutiny of the constitutionality or otherwise of legislation through the lens of the equality principle necessarily grants a wide range of discretion to any court exercising judicial review of statutes. In this respect, it is significant that the Constitutional Tribunal described the principle of equality as derivative from the principle of social justice.11 It shows that, as views on social justice are essentially contestable and largely indeterminate, so are conclusions as to whether a given
198
CHAPTER 8
regulation complies with the constitutional principle of equality. The Tribunal has used this conceptual device to strike down a number of socio-economic regulations, such as the laws on “indexation” of pensions (i.e., adjustment of pensions to rises in the cost of living): upper limits imposed upon pension payments were found to be inconsistent with the principle of social justice, and hence with equality.12 There have been, however, also some decisions in which the Constitutional Tribunal has attempted to narrow down the scope of those characteristics that may be deemed “relevant”, by appealing at times to the concept of immutability (albeit tacitly) as an indicium of a forbidden basis for discrimination. For example, in one early decision, in 1987, which struck down gender quotas in admissions of students to the medical academy, the Tribunal pronounced the principle that equality “in the field of law” is respected “when every citizen may become an addressee of each [legal] rule conferring a certain civil right”.13 The upshot of this is that it is improper to differentiate among citizens in terms of criteria (such as gender) that lead to the creation of closed (caste-like) categories of legal subjects.14 This argument has not, however, been developed or even repeated in later decisions of the Tribunal; the preferred theory in all subsequent cases has been based upon the more open-ended concept of the relevance of the contentious distinction to the purpose of the rule in question. More generally, it should be observed that the great majority of decisions of constitutional courts taken under equality provisions do not appeal directly to any of the prohibited grounds of invidious discrimination, but rather challenge the rationality of the legal classifications employed in various areas of socio-economic policy: in taxation, pensions, unemployment and welfare benefits, etc. They can be considered to be the more routine, less obvious cases of discrimination, in which no suspicion of the intention of the legislator to act to the prejudice of a specific group – women, ethnic minorities, people of unpopular sexual orientation – is justified, but rather where a socio-economic choice is questioned by the constitutional courts on the basis of its perceived irrationality or arbitrariness. An interesting case on pensions for war veterans was brought, in 1994, before the Romanian Constitutional Court.15 Under challenge was a statutory provision that excluded those who had fought against the Romanian Army from receiving veteran benefits. The Constitutional Court ascertained that this section related to those persons who had been compulsorily conscripted into the Magyar army (as they lived in territory that was temporarily occupied). It was thus impossible for the Romanian army to have conscripted them at that date. According to the government, who defended the legislation before the Court, whether the Romanians fighting in enemy armies were volunteers or forcibly drafted was irrelevant: the argument against granting them war veteran benefits was presented as a “moral” one, as “it would be inconceivable that the Romanian state should give rights to the people who infringed upon its independence and integrity”.16 The Court, however, rejected this argument, finding that, given that the obligation (to fight for, and not against, one’s own army) was impossible for some to fulfil, this law unlawfully discriminated against those who had been conscripted into the Magyar army. It was thus held to be in violation of Article 16(1) of the Constitution, which states: “Citizens are equal before the law and public authorities, without any privileges or discriminations”. However, it is
EQUALITY AND MINORITY RIGHTS
199
interesting that the Court felt compelled to make the following statement in its opinion, probably as a concession to the Government and to rebut possible accusations of insufficient patriotism: The arguments [by the government] involved in the combat of claims [sic] examined by the Court, referring to the events that took place during 1940-1945 in temporarily occupied territories . . . are in no way annulled by the present decision, which was imposed in order to insure the concordance of the provisions of the law with the Constitution. The constitutional juridical [sic] solution given for the problem does not distort in any way the historic reality, nationally and internationally known.17
A decision of the Slovenian Constitutional Court on a law dealing with the reprivatisation of land illustrates how the equality principle can be used to interfere with a policy – in this case, a policy designed to control and restrict the return of large pieces of land to private owners (and, in this case in particular, to the Church).18 This case related to Article 1 of the Act on Partial Suspension of the Return of Property. This act put a moratorium of three years on the operation of another law, which regulated the return, through denationalisation proceedings, of agricultural lands and forests that had been seized by the former regime. This moratorium applied in all cases where claimants requested the return of more than 200 hectares of agricultural lands and forest. The Court conducted a proportionality review of this new statute, and found that the measures adopted were not proportionate to any constitutionally permissible goals; for example that the legislator was motivated by a groundless (in the Court’s opinion) fear of a return of “feudal ecclesiastical lords”.19 More relevant to our current discussion, however, is the Court’s objection to the fact that the moratorium only applied to land over 200 hectares. This, according to the Court, amounted to discrimination between denationalisation claimants (as between those entitled to receive over 200 hectares as opposed to under 200 hectares, and also as between those who had not yet had their claims processed and those who had already received their land over 200 hectares before this moratorium came into force);20 thus, it violated the principle of equality (Article 14 of the Constitution). However, it is clear that, in cases such as these, any quantifiable line of distinction between different categories of claimants can easily be attacked for its discriminatory character, because it draws a line between claimants; any such line can be attacked as arbitrary. 2. GENDER AND SEXUAL ORIENTATION EQUALITY Decisions related to gender equality belong to the most important in the equality jurisprudence of the constitutional courts in CEE. Small wonder: old Communist laws and rhetoric paid lip service to the ideal of equality of the sexes, while at the same time maintaining, and often petrifying, traditional social norms of inequality and discrimination against women. Modern ideas concerning gender equality, especially in the workforce and in political life, collide with traditional attitudes and prejudices. Ken Jowitt is correct when he observes: “Antifeminism is palpable throughout Eastern Europe”.21 As a result, various forms of discrimination against women, both in the legal system and in society in general, have persisted. One
200
CHAPTER 8
obvious area is in the sphere of employment, with a disproportionately small number of women in higher managerial positions, the average remuneration of women well below that of men (e.g. in Poland the average remuneration of women constitutes 75 percent of men’s; in the Czech Republic, that figure is 70 percent), no equal pay for equal work, and a widespread tendency among employers to follow discriminatory practices, such as demanding pregnancy tests prior to hiring a woman.22 Often the discrimination results not so much from formal legal rules, but from societal norms and prejudices, such as, for example, the strong conviction persisting in the countries of the that there exist “feminine” (secretaries, cleaners, etc) and “masculine” (engineers, managers) professions and positions.23 There is also a dramatic underrepresentation of women in all areas of politics: in parliamentary and local legislative bodies, governments, top judicial bodies, etc.24 One example of such discrimination, traditionally taken as reflecting self-evident differences between men and women, is provided by the differences in compulsory retirement age, which, in most countries in the region, has usually been about five years less for women than for men. The most progressive court in this regard turned out to be the Polish Constitutional Tribunal. Beginning in 1991, when it abolished a differential retirement age in academic positions,25 it subsequently took several decisions, striking down particular statutes that provided for a lower compulsory retirement age for women than for men, e.g. for civil servants,26 for employees of pharmacies,27 and for teachers.28 In this series of decisions, the most elaborate and interesting is the one that struck down the earlier retirement age for women than men in civil service, and at the same time established the grounds for permissible positive discrimination.29 Under challenge was a rule in a 1996 statute regulating the civil service, which allowed the discharge of a female civil servant at the age of 60, against her will – that is, five years earlier than the permissible non-voluntary discharge of a male civil servant. In this decision, the Constitutional Tribunal criticised an earlier decision of the Supreme Court, in which the latter had found no discrimination in an earlier discharge age for women on the basis of the theory that such a disadvantage for women is counter-balanced by the respective advantages that they receive, namely, an earlier acquisition of retirement pension benefits.30 The Supreme Court had then argued that no discrimination exists when the different treatment was relevant to the different positions of the addressees of a rule, and the relevant differences here applied not only to “biological and social differences” but also to those established by the law, namely, the legal privileges that only women enjoy (such as earlier pension benefits). According to the Supreme Court, the benefits for women outweighed the disadvantages, and so “the reasons which support a more advantageous status vis-à-vis pension benefits for women . . . argue for a differential regulation of the situation of both these categories of persons [women and men] with regard to the possibility of their discharge, and in any event can be sufficient evidence to deny the charge of discrimination against women”.31 The Constitutional Tribunal categorically rejected this theory. It began by reasserting its earlier pronouncements on gender equality, which had established the principle that departures from equal treatment can only be legitimate where they can be justified by “a desire to achieve actual social equality [between men and women]”.32 It is constitutionally acceptable,
EQUALITY AND MINORITY RIGHTS
201
the Tribunal stated, to establish a different legal status for men and women insofar as it is based on the principle of social justice, which demands that women are offered equal positions compared to men. This is an explicit articulation of the principle of affirmative action: “as in social reality women, as a rule, have weaker positions . . . there is a constitutional justification for enacting rules that confer certain benefits upon women because this is an instrument leading to a de facto equality for women”.33. The Tribunal even went a step further and held that, when social and biological differences between men and women are particularly pronounced, the enactment of such “compensatory privileges” is a duty of the legislator.34 Having so defined the principle of positive discrimination, the Tribunal then rejected the Supreme Court’s approach, based as it was on the mutual balancing of the privileges and disadvantages for women. It stated: “there are no constitutional grounds to accept the thesis that if the legal position of a female employee displays in a particular respect a privilege vis-à-vis the position of males, then the principle of equality allows (or perhaps even, demands) a balancing of this privilege by an imposition upon women of certain duties that do not apply to men”.35 Thus, it was held that there is no requirement of an equal overall balance of benefits and duties of employees of both sexes.36 A parallel argument cannot be applied to men, the Tribunal added, because “in Poland these days there are no grounds for treating men as a weaker social group”.37 That is why the rule of an earlier discharge is discriminatory and this discriminatory character is not redeemed by a benefit consisting in an earlier acquisition of old-age pension rights.38 It is important to emphasise the particularly enlightened character of this (relatively unknown) decision. The Constitutional Tribunal rejected, although not in so many words, the spurious doctrine of the “equivalence” of compensatory privileges for women and men, thus adopting a context-sensitive approach in which lawmakers are sensitive to the actual pattern of disadvantages in their society. Further, it based its theory justifying affirmative action (albeit, again, implicitly) upon a goal of “genuine” (or “fair”, in Rawlsian terms) equality of opportunity, which sees actual material inequalities as relevant to inequalities of opportunity in a society of systemic inequalities. In a somewhat different context, the Hungarian Constitutional Court appealed explicitly to the notion of “positive discrimination” in favour of women when rejecting a complaint by a man who challenged a gender distinction with regard to military service: as women were not compelled to serve in the army, he claimed, they received an unfair legal benefit.39 The Court explained that this distinction amounted to a “positive discrimination” aimed at achieving eventually greater equality; although, as Kim Lane Scheppele (who is otherwise very sympathetic to the Hungarian Court) caustically observed, the Court “did not explain why excluding women from military ranks constitutes ‘positive discrimination’, or how exclusion would lead to greater equality in the long-term…”.40 As mentioned above, anti-homosexual prejudice is quite widespread in the CEE region, with, in particular, those inspired by the teaching of Catholic Church displaying a hostile approach to same-sex relationships, considering them a deviation from and a threat to the moral fabric of society.41 Furthermore, on top of the general societal aversion, formal legal rules generally discriminate against
202
CHAPTER 8
homosexuals, especially in the spheres of family and succession law: it is, for example, impossible to legally register same-sex marriages; and those living in homosexual relationships are denied “family member” status for the purposes of taxation, inheritance, social assistance, etc. In the most important decision on the question of sexual orientation in the region of CEE, the Hungarian Constitutional Court struck down the legal non-recognition of a de facto homosexual relationship qua “a common household”, but at the same time upheld the rule allowing only heterosexual marriages.42 Under challenge were two provisions: a family law provision that defined marriage as a union between a man and a woman, and a civil code provision that defined a “domestic partnership” as a woman and man living in a common household outside marriage. Regarding the question of the legal definition of “marriage”, the Court based its decision on the basis of the traditional understanding “both in our culture and in law” of the institution of marriage as a heterosexual union.43 One might, of course, object that the traditional understanding of marriage in law is precisely what is at issue here, and so cannot figure both as evidence and as a conclusion. But it surely could not be expected that the Hungarian Constitutional Court would be the first legal authority in the world at that time to recognise same-sex marriages; further, it should be noted that the Court’s relative conservatism with regard to the notion of marriage is counterbalanced by its liberalism concerning the second issue in this decision, namely, the de facto relationship. In this respect, the Court linked its reasoning to the principle of equal personal dignity which must apply to any union of two persons living together, regardless of their gender. To deny legal recognition to same-sex couples is a case of “negative discrimination” because “[t]he cohabitation of persons of the same sex . . . [is] in all respects . . . very similar to the cohabitation of [heterosexual] partners in a domestic partnership – involving a common household, as well as an emotional, economic and sexual relationship, and taking on all aspects against third persons . . .”.44 Such a legal exclusion “is arbitrary and thus violates human dignity; therefore it is discriminatory. . . “.45 This compelled the Parliament to amend certain laws, including the Civil Code, in order to allow succession of property within de facto homosexual couples. Further, a relatively minor but symbolically meaningful decision of the Hungarian Constitutional Court of 1999 struck down as unconstitutional a provision of the Criminal Code that penalised “normal” heterosexual intercourse between siblings and “unnatural” homosexual intercourse between siblings.46 “Unnatural” heterosexual intercourse between siblings was not singled out in the code, and, on this basis, the Court struck down the provision, saying (with two judges offering dissenting opinions) that it differentiated arbitrarily between “unnatural” sexual intercourse between siblings of the same, and of different, sex. A much more fundamental, and more objectionable, form of discrimination was invalidated by the Romanian Constitutional Court which struck down, in 1994, a Criminal Code provision (art. 200) prohibiting homosexuality even in private.47 The matter came to the Court in the process of concrete review, at the insistence of two indicted men charged with the offence under art. 200. They claimed that this provision violated several articles of the Romanian Constitution48 as well as art. 8 [right to privacy] of the European Convention on Human Rights (ECHR). The trial
EQUALITY AND MINORITY RIGHTS
203
court, for its part, held that the challenged penal law provision did not violate the Constitution, because it defended public order and “good morals”, hence, the values protected by the Constitution (art. 26). In accordance with the usual procedure in such cases, the Constitutional Court asked the government and both chambers for their views; the only substantive response came from the government, which defended the penal prohibition under the constitutional exception to the right of privacy that refers to the “rules of conduct of other members of society” as well as “the general moral sense”. The government also offered a similar reading of the European Convention on Human Rights (ECHR): the limits on the right to privacy in art. 8 include interference based on the protection of public order and morals. The Constitutional Court was therefore confronted with a traditionalist viewpoint whereby the right to act in accordance with one’s sexual orientation has to surrender to the views of the majority concerning the proper rules of conduct, moral sense, etc. Furthermore, to reinforce this stance, a number of religious groups and churches (which were also asked to express their opinion) emphatically condemned homosexual acts “the majority of them is asking for maintaining the criminal prosecution of these practices”. On the other hand, a number of NGOs, both Romanian (e.g. the Romanian Institute for Human Rights and the Helsinki Committee) and international (including the International Commission of Jurists and Amnesty International) provided the Court with their negative opinion on the legality of the provision in question and called for its abolition. Faced with all of this, the Constitutional Court decided purely on the grounds of the inconsistency of article 200 with article 8 of the ECHR, as interpreted in the line of cases on sexual orientation by the Strasbourg Court. On the grounds that Romanian constitutional provisions have to be interpreted in accordance with international treaties to which Romania is a party, and that Romania is a party to the ECHR, the Court found that it had a duty to remove the inconsistency between the anti-homosexual provision of the penal law and article 8 of the ECHR. At the same time, it was careful to emphasise that the scope of protection afforded by art. 8 is limited to homosexual acts among consenting adults in private “under the condition that they do not provoke public scandal”. It seems to be implied by the Court that the last proviso (i.e. concerning public scandal) may be activated only when the act is committed in public and not in private, even if others somehow find out about it and become shocked or outraged. The latter interpretation (public scandal produced by a private act) would, of course, render the whole argument meaningless. Unfortunately, such an interpretation is not explicitly rejected by the Court: the public scandal proviso is listed in addition to the other conditions that would render the act actionable,49 and this small ambiguity is perhaps a sign of the Court’s effort to appear more moderate than in fact it was in striking down the anti-homosexual provision. 3. SPECIAL CASE OF AFFIRMATIVE ACTION While most of the constitutions in the region limit themselves to the prohibition of discrimination (or a general proclamation of the principle of equality), some go further by providing for special protection for certain specified social groups in
204
CHAPTER 8
particular social settings: for example, protection of children and minors in the workplace,50 and affording to them assistance in professional training and special health care;51 protection of women in the workplace,52 and, in addition, special protection of pregnant women;53 special assistance to and protection for mothers;54 protection of and aid to the disabled,55 etc. These provisions, however, have not been worded so as to mandate positive discrimination, that is, a deliberate programme of preferential treatment accorded to a group based on its disadvantage in access to a given social good. In fact, two of the constitutions (which, otherwise, contain provisions on special protection for certain groups) contain explicit prohibitions on granting any privileges based on various listed grounds. The Lithuanian constitution, for example, states that no restrictions can be imposed upon nor privileges granted to a person on as the basis of sex, race, nationality, etc;56 likewise, the Bulgarian constitution precludes the conferral of any privileges upon such grounds.57 This suggests that any acts or statutes that envisage positive discrimination in favour of women or ethnic minorities could be struck down on these grounds. The Slovak Constitution expressly limits the provisions concerning the rights of members of ethnic and national minorities with the qualification that the protection of their rights must not lead to, inter alias, “discrimination against [the Republic’s] other inhabitants”.58 There are, however, some exceptions to this general silence of CEE constitutions on measures of positive discrimination. The Hungarian Constitution goes beyond merely proscribing any discrimination, and provides that the state shall implement equal rights for everybody “through measures that create fair opportunities for all”.59 This may be seen as a mandate to the state to take positive action that will actually lift the position of disadvantaged groups vis-à-vis others in society. The Hungarian Constitutional Court considered a challenge to affirmative action in the form of special tax benefits to families with numerous children – hence not a typical “reverse discrimination” issue (which occurs in its most obvious form when privileges are granted to a group that has traditionally been discriminated against); nevertheless, it did represent an interesting test for the notion of equality as understood by the Court.60 The Court rejected the challenge, stating that “the ban on discrimination does not mean that any discrimination, including even discrimination intended to achieve a greater social equality, is forbidden”.61 The Court specifically acknowledged that the Constitution allows for positive discrimination aimed at eliminating inequalities of opportunity; it also noted that the anti-discrimination clause of the Hungarian Constitution (art. 70 A) gives effect to a broader notion of equality, which should be understood as an “equal right to human dignity”: “The ban on discrimination means that all people must be treated as equal (as persons with equal dignity) by law”.62 As one expert on the Court has noted, this formulation has been directly influenced by Ronald Dworkin’s distinction between equality in the sense of a “right to equal treatment”, a sense that is inferior and subordinate to “the right to being treated as an equal”.63 Dworkin had made this distinction precisely in the context of his discussion of “reverse discrimination” and its compatibility with the principle of constitutional equality.64
EQUALITY AND MINORITY RIGHTS
205
4. MINORITY ISSUES IN CENTRAL AND EASTERN EUROPE: AN OVERVIEW The region of Central and Eastern Europe displays a wide and untidy mosaic of ethnic, national and religious minorities within and across state boundaries. After the fall of Communism, nationalism was (alongside religious fervour) often the only force capable of cementing people together and counteracting the prevailing anomie;65 the downside of this, however, was that all too often such nationalism was characterised by rampant and violent exclusion, fed by a widespread sense of hostility towards “the other”. Communism had often swept various ethnic and national animosities artificially “under the carpet”: these tensions were thus not overcome but merely deprived of any open expression under the official orthodoxy of national unity. The Balkan wars and the split of Czechoslovakia are the most visible examples of this but various actual and potential cases of discrimination against ethnic and national minorities exist everywhere in the region. The characteristics of this discrimination vary, depending upon the nature of the minority in question and its relationship to the majority. Probably the most dramatic is the situation of Roma people, against whom hostility, discrimination and prejudice are strongly ingrained throughout the region,66 and rendered even worse by their added disadvantage of no specific territorial concentration that could help to build the capacity for effective political mobilisation. As a result, despite putting forward candidates in parliamentary and municipal elections in a number of countries, they remain virtually unrepresented in the political systems of countries they inhabit – and this despite the substantial numbers (around 20 million) belonging to this minority in the region.67 Some states in the region are relatively homogenous, with small and territorially identifiable minorities (such as Poland with its German, Ukrainian and Belarussian minorities). Apart, however, from Poland, the Czech Republic, Albania and Hungary, the presence of ethnic and national minorities in the countries of the region is quite sizeable, and varies between 10 and 55 percent of the population in any given state.68 Some of these minorities actually form a majority of the population in certain regions (e.g. Hungarians in the southern parts of Slovakia), which leads to understandable nervousness on the part of the national majority towards any claims for territorial autonomy. As an example, consider an initiative by a group of local mayors in Slovakia to establish a self-governing province populated predominantly by Hungarians – a move quickly condemned by Slovakian politicians as a threat to state sovereignty.69 More often than not, however, the minorities in CEE countries are not sufficiently concentrated in any given area to render the idea of territorial autonomy plausible. The presence of sizeable Hungarian minorities in Slovakia, Romania and Serbia (2.5 million ethnic Hungarians live in neighbouring countries, compared to just over 10 million inhabitants in Hungary itself) pose special problems both for Hungary, which tries to maintain a sort of tutelage over Hungarians abroad,70 and the host countries, which fear an irredentist minority encouraged by its neighbouring kin state. A similar case is that of Albania, a relatively
206
CHAPTER 8
homogenous state, but with sizeable Albanian minorities in neighbouring Macedonia and Serbia (Kosovo). A delicate and fragile situation persists in the Baltic states, in which a large proportion of the population is of Russian ethnic origin (for example, immediately after regaining independence, thirty percent of Estonia’s population, and 34 per cent of Latvia’s, was Russian-speaking),71 as a result of the USSR’s deliberate policy of encouraging Russians and other Slavs to settle in the outlying republics. This ethnic Russian population is often viewed with politically based distrust, as belonging to the formerly oppressing nation, regardless of how long they have lived in these republics. They were also initially denied many rights, including those linked to citizenship, on the “restorationist” theory that Soviet-era migration (which happened to be mainly Russian) was the result of an illegal take-over of the Baltic republics by the USSR in 1940. As an Estonian scholar Vello Pettai notes, “Although the Estonians’ argumentation was eminently juridical in that it related purely to the consequences of an illegal foreign occupation, its practical consequences in terms of the political marginalisation of a large share of the minority population were severe”.72 The dramatic legacy of ethnic and religious persecution left by the old regime, which has persisted in certain forms, is exemplified by Bulgaria where, since the early 1950s, the Communist Party had led a struggle against “expressions of nationalism and religious fanaticism among the local Turks”. Under this label, the persecution and harassment of Turks and Pomaks (ethnic Bulgarians who converted to Islam in the past) took various forms, such as the deprivation of Turks of their land; forced emigration to Turkey; forced renaming of the Pomaks and Turks (as recently as the mid-1980s); detention of those resisting these campaigns, even peacefully, in prisons and camps; a ban on the use of the Turkish language in public and on the celebration of Muslim holidays and rituals; official anti-Turkish propaganda, etc.73 As a result, Bulgaria entered into its democratic era with ten percent of its population (and this does not include the Roma minority)74 nursing fresh memories of persecution, harassment and discrimination. The disintegration of the old Soviet Union produced a situation in Russia in which a host of ethnic conflicts were very early on transformed into ethno-territorial conflicts: although Russia is more mono-ethnic than many other European states (with over 80 percent of the population being ethnic Russian), there are also more than a hundred different indigenous ethnic groups and thirty-one ethnic territorial units; ethnic conflicts are never far from the surface in the clashes between these units with “the Centre”.75 However, the “ethnic region versus the Centre” conflict (of which the Chechen and Tatarstan crises are examples) is only one of a number of different types of ethnic-based tension in the Russian federation: one knowledgeable scholar has identified others, such as the conflicts between ethnic regions themselves (e.g., Ingushetia v. North Ossetia); conflicts within a region between titular ethnic groups controlling “their own” territories (e.g. Karachai-Cherkessia), and within an ethnic group between sub-ethnic groups, clans, etc.; conflicts within a region between ethnic groups, including non-titular ones, for control over resources (e.g., anti-Chinese feelings in the Far East); and tensions arising out of competing claims to resources made by nations separated by frontiers of territorial units (e.g.
EQUALITY AND MINORITY RIGHTS
207
Ossetians, and, of course, Russians).76 In any event, the territorial solution cannot be seen as an adequate response to ethnic tensions as there is not a sufficient degree of correspondence between an ethnic group and a given territory: there are many groups that are spread over a number of different units (for instance, only one-third of ethnic Tartars, the second largest ethnic group in Russia, lives in Tatarstan).77 In sum, the nature of the problem varies from country to country; none however has a record free of discrimination and persecution on racial or ethnic grounds. Throughout the region, states attempted to consolidate the power of the majority by marginalising and under-valuing the cultural and political claims of minorities; the demands for minority rights were, more often than not, “defensive responses to the threats posed by assertions of majority nation-building”.78 Hence, the constitutional method of protecting minorities was one of the most contentious problems in the process of constitution drafting in the region, and in the practice of the constitutional court of at least some of the CEE states. 5. CONSTITUTIONAL DESIGN OF MINORITY RIGHTS: GROUP OR INDIVIDUAL RIGHTS? The only constitution in the region that fails to mention minority rights is the Constitution of Bulgaria. All of the others list various catalogues, with special prominence given to language and educational rights, the right to preserve one’s cultural and religious identity, etc. Minority language is clearly the main protected interest among minority rights (and will be discussed, in more detail, below).79 All constitutions, with the exception of the Bulgarian one, contain provisions granting a right to preserve one’s language and cultural identity. For example, the Constitution of Latvia provides as follows: “Persons belonging to ethnic minorities have the right to preserve and develop their language and their ethnic and cultural identity”.80 The catalogues of minority rights are often more elaborate, as in this provision of Romanian constitution: “The state recognises and guarantees the right of persons belonging to national minorities, to the preservation, development, and expression of their ethnic, cultural, linguistic, and religious identity.” 81 There are both negative and, at times, positive formulations of minority rights. They are formulated negatively when, as in the Latvian wording just quoted, members of a group are protected against possible infringements on their interests in the preservation and development of their culture, language, etc. Often, however, certain minority rights, in particular the right to education in one’s own language, are framed as positive rights imposing certain active duties upon the state. For example, in Hungary the Constitution states that “The Republic of Hungary shall provide for the protection of national and ethnic minorities… [and] education in their native languages”,82 and the Macedonian Constitution declares that “The Republic guarantees the protection of the ethnic, cultural, linguistic and religious identity of the nationalities”.83 Positive state duties are sometimes restricted to particular obligations, particularly in the sphere of official communications and interaction of the citizens with governmental bodies. For example, in Estonia, there is a very specific regulation concerning the official use of language, which provides:
208
CHAPTER 8
“In localities where at least half of the permanent residents belong to an ethnic minority, everyone shall have the right to receive answers from state and local government authorities in the language of the ethnic minority”.84 Finally, several constitutions provide for the rights of minorities to participate in public affairs qua minorities. The Hungarian Constitution proclaims that “national and ethnic minorities will be assured collective participation in public affairs” and that “The laws of the Republic of Hungary shall ensure representation for the national and ethnic minorities living within [the] country.”85 The Constitution of Montenegro goes even further by envisaging a system of proportional representation not only in the “state authorities” but even in public services: “Members of the national and ethnic groups shall be guaranteed the right to proportional representation in the public services, state authorities and in local self-government.”86 The constitutions do not, on the whole, attempt a definition of the term “minority”, nor refer to a definition enshrined in any other international document (which is not surprising, given the lack of any such precise definitions in the major international agreements on this subject). A couple of constitutions do, however, make statements in this regard. The Constitution of Macedonia relates minority protection to “inhabitants belonging to a nationality”87 (in the context of the right to use a language other than Macedonian as an official language) or “[m]embers of nationalities”.88 The Russian Constitution, instead of providing protection for minority groups, or the individuals belonging to them, generalises the problem: it grants traditional minority protections to all citizens. Thus, article 26 states: Everyone shall have the right to determine and state his national identity. No one can be forced to determine and state his national identity. Everyone shall have the right to use his native language, freely choose the language of communication, education, training and creative work.
The Constitution of Slovenia distinguishes between different types of minority groups in its provisions on the protection of minorities. For example, it states, in article 61, that “Each person shall be entitled to freely identify with his national grouping or autochthonous ethnic community, to foster and give expression to his culture and to use his own language and script”. However, in addition to this, there are specific rights subsumed under the heading “Special Rights of the Italian and Hungarian Ethnic Communities in Slovenia”.89 Here, these groups are given additional rights such as “to establish organizations, to foster economic, social, scientific and research activities … to plan and develop their own curricula [for education]…. In those areas where the Italian and Hungarian ethnic communities live, their members shall be entitled to establish autonomous organizations in order to give effect to their rights…”.90 One may add that, while Italians and Hungarians may be seen as “indigenous” groups in Slovenia because they have inhabited that area for centuries, they are not the most numerous ethnic minorities: Croats, Serbs and Muslims constitute proportionately larger minorities in Slovenia than do the Italians and Hungarians.91 The only explanation for this apparent abnormality is that the issue of the relationship between ethnic Slovenians and ethnic Italians and Hungarians in Slovenia is politically less explosive than the relationship between the members of the ethnic groups that made up ex-Yugoslavia. Hence, it was safer to
EQUALITY AND MINORITY RIGHTS
209
accord a special, elaborate and advantageous minority status to Italians and Hungarians than to Serbs, Croats and Bosniaks (Bosnian Muslims). Different treatment is also accorded to the Roma people. Article 65 states that “The status and rights of Gypsy communities living in Slovenia shall be such as are determined by statute”. This would seem to suggest that they do not fall within the general provisions on minorities and are not considered to be a minority group. Most of the constitutions of the region phrase minority rights in the language of individual rights, as held by “persons belonging to national minorities…”.92 In some cases, however, the language of group rights is used. For example, the Hungarian Constitution states: “National and ethnic minorities shall have the right to form local and national bodies for self-government”.93 Slovenia also takes this approach, albeit in relation to the rights of Hungarian and Italian minorities only (the others are treated as individual rights).94 Thus, these states truly create constitutionally guaranteed group rights. Several constitutions use both the language of group and individual rights, depending on the nature of the right proclaimed. For example, the Polish constitution uses group-rights language with regard to the establishment of educational and cultural institutions for national minorities95 and individual-rights language when dealing with the freedom to maintain one’s customs, tradition and culture.96 What difference does it make? The main constitutional dilemma with regard to the protection of minorities is whether the best way of protecting members of (national, ethnic, religious etc) minorities is simply by strong protection of individual rights backed up by a robust non-discrimination principle, or whether there should be a special constitutional principle (or set of principles) that confers special rights upon minority members. The former (liberal-individualistic) approach dominates the thinking about the protection of minorities in the United States: the idea is that if every citizen, regardless of their (inter alia) national or ethnic group membership benefits from the same strong civil and political rights, then any special group-based protection is redundant, and potentially dangerous.97 This may be called a “liberal-neutralist” (or individualistic) approach. In the continental European setting, however, this approach has been seen as largely ineffective and insufficient. In Europe, there is much less faith in the beneficial effects of the extension of individualistic liberal principles to a situation in which anti-minority prejudices and hostility are deeply ingrained, and are also displayed by those who are entrusted with the enforcement of general rules. Further, the liberal-individual approach is considered well-suited to the particular situation of immigrant societies, where the dominant concern of new minorities is to enjoy the same rights as the older population and to integrate themselves into a larger society governed by neutral rules. In contrast, when the claims for protection come from groups that have been present in a given territory for a long time, or that find themselves sharing the same nation-state due to changing borders or to forced movements of population (hence, forced rather than voluntary migration) etc, the purely individualistic method appears much less capable of providing real and effective protection to minorities.98 This distinction corresponds to what some refer to as the difference between “minorities of force” (minority groups that desire equal treatment but are denied equality by dominant majorities) and “minorities of will” (groups that desire to
210
CHAPTER 8
maintain their separate identities and thus demand different treatment from the dominant group).99 The terminology may be somewhat misleading, because the latter groups are also often victims of force and oppression exercised by the majority, but the nature of their claim is different. The demand of minorities of will is for constitutional entrenchment of minority rights distinct from, and operational alongside, universal individual rights. If non-dominant groups wish to preserve their identities, threatened as they are by extinction, then a prescription of equal protection and strict non-discrimination is not sufficient. Special measures designed to protect minority groups may be called for. Rights such as the right to use one’s own language in dealings with the authorities, or of special representation in local or national bodies, do not lend themselves to individualistic wording; at a certain point the group that is entitled to such privileges has to be identified, and at this point the universal-individualistic approach ceases to be suitable.100 If these forms of protection are viewed as important then they have to be accompanied by certain limiting clauses: not every single foreign-language speaker (or, put more cautiously, non-majority-language speaker) can be given the opportunity to use her or his language in communications with the state, and not every minority (however small) can be granted recognition through special representation in the legislature. The right then ceases to be universal; the question of feasibility will have a bearing on the choices that are made to single out some groups, but not others, for legal privileges. Perhaps one of the main reasons why the individualist-liberal rejection of the notion of minority rights is more ingrained in the Anglo-American constitutional systems (in particular, in the United States, and to a lesser degree in countries such as Canada, Australia and New Zealand) than in Europe is that in the former, but not the latter, setting there is a problem that has traditionally given liberal theorists a headache: how to reconcile a universal commitment to individual human rights (including the right to autonomy) with respect for the traditions of minorities that often do not practice autonomy and are (by liberal standards) quite oppressive towards their members. This may be seen as the fundamental liberal dilemma when it comes to minority rights. On the one hand, a liberal is committed to extending some fundamental dignity-based rights to everyone: no-one, regardless of their group membership, should be denied freedom of choice about basic personal matters, fair political representation, free expression, non-discriminatory treatment, physical inviolability etc. On the other hand, those minorities, often indigenous ones, that do not respect fundamental equality between men and women, that practice corporeal punishment, and that do not respect the individual’s right to control his or her life to the degree deemed necessary by liberals, pose a threat to these fundamental values. Hence, the liberal theorist is concerned about the position of the most vulnerable members of those minorities – often women and children – who are threatened with deprivation of all those individual rights that non-minority citizens take for granted. Group rights aimed at the protection of the identity of the group as a whole give to that group a degree of immunity from interference by the wider community into its “internal affairs”, which may extend to the interests of the vulnerable insiders. As Michel Rosenfeld observed (deliberately pushing the dilemma to the extreme): “it appears impossible for any constitutional regime to
EQUALITY AND MINORITY RIGHTS
211
guarantee at once a minority group’s survival and the most fundamental rights of an individual dissident within that group”.101 This perception animates much of the liberal critique of group rights. As noted by another author, who has recently made by far the most eloquent and passionate defence of such liberal universalism: [I]t seems overwhelmingly plausible that some groups will operate in ways that are severely inimical to the interests of at any rate some of their members. To the extent that they do, cultural diversity cannot be an unqualified good. In fact, once we follow the path opened up by that thought, we shall soon arrive at the conclusion that diversity is desirable to the degree, and only to the degree, that each of the diverse groups functions in a way that is well adapted to advance the welfare and secure the rights of its members.102
The prima-facie hostility of the Anglo-American legal systems to minority rights can be seen as resulting largely (though not solely; one should not discard much more invidious factors related to racism) from this dilemma. However, in the continental European setting, and in particular in CEE, this dilemma is out of place; the problem just identified simply does not ring true. The pattern of relations between an ethnic majority and minority (or minorities) plainly does not fit the description of “liberal majority versus oppressive minority”;103 therefore, the fundamental philosophical reason for distrusting the very idea of minority rights does not apply (or applies to a much lesser degree) to the European situation. Obviously this does not negate the fact that a “multicultural” solution, with an explicit recognition of separate minority rights, is often seen as a threat to the culture of the majority, and to state sovereignty. The problem, then, is not whether a liberal-neutralist model or a diversity-accommodating model (that is, a pluralist model) should be adopted; this dilemma seems to have been answered overwhelmingly in favour of the latter. As a Serbian legal scholar, Tibor Varady, concludes: experience in [CEE] countries has shown that ethno cultural neutrality and groupneutral regulation cannot accommodate cultural pluralism, and cannot guarantee stability and peace between ethnic majorities and minorities. Traditional liberal attitudes lack empathy towards maintaining diversity, and cannot provide solutions in traditionally multicultural environments where equality presumes an equal right to maintain one’s distinct identity.104
And it is significant that virtually the same argument has been officially endorsed in Hungarian law, namely in the 1993 Act on the Rights of Ethnic and National Minorities, which proclaims in its introduction that “minority rights cannot be fully guaranteed within the bounds of individual civil rights; thus, they are also to be formulated as rights of particular groups in society”.105 The question is, how to reconcile the diversity-accommodating constitutional regime of protection of minority rights with the values underlying the rule of law, namely equality before the law and the prevention of arbitrary privileges and discrimination. At first glance, it might appear that minority rights are necessarily group rights. This, however, is not the case: one should distinguish between group rights sensu stricto, where the beneficiary and rightful claimant of a right is a group as a whole (and where, presumably, the right is exercised in accordance with the decisions of some authorised leaders and representatives of the group), and, on the other hand, rights that are conferred upon individuals by virtue of their belonging to an ethnic,
212
CHAPTER 8
national or other minority. Group-specific rights may be individual in the sense that a claim-holder is an individual, even although the basis of his or her claim is that he or she belongs to a group. As one Russian scholar has noted: “Any public support to institutions that address specific needs of persons belonging to minorities can be justified in terms of individual rights”.106 Whether the distinction is significant in practice is another matter; the point can be made that individuals are best able to exercise their minority-based rights when they act in concert with other members of the same minority: “we may insist that a certain right . . . is due to individuals but the enjoyment of that right is nonetheless unthinkable without others”.107 But this is a practical and contingent matter; and need not always be the case. As another commentator has observed, “for some aspects of minority/indigenous protection – such as standing to bring complaints before international bodies – it may be important to determine whether a right is collectively or individually held”.108 Some rights, by their very nature, better lend themselves to a collectivist articulation than to an individualist one, or vice versa. Generally speaking, one may venture the proposition that, if a right more closely resembles an exemption from a general duty than a claim to the provision of certain services, it is perfectly imaginable and practicable to claim that right individually and regardless of others, even if the basis for the claim is membership of a certain group. In that case, there is no need to collapse an individual group-based right into a group right in the strict sense. One might think, for example, of officially recognised conscientious objection to military service based on belonging to a religious group of which one is the only adherent in a given country. As a general proposition, therefore, these two understandings of rights are quite distinct from each other. In the case of a group right sensu stricto there must be an officially recognisable (and recognised) corporate body that speaks on behalf of the group. This raises the obvious problem of whether that body has the legitimate authority to represent all of its members. The process of decision-making as to the exercise of a right is then taken away from the individuals concerned – who may actually reject the group’s authority to represent them – and centralised in one group-representative body. In the case of individually-worded group rights (where the individuals are entitled to claim, for example, for access to education in their own language) the problem of representation disappears; however, as a practical matter, it is inconceivable that such a right can be satisfied unless there is a sufficient critical mass of people to claim it. 6. LINGUISTIC RIGHTS The legal regime governing the relationship between the official state language and minority languages is a good test of the degree of accommodation of minorities in CEE countries. Usually the greater the fear of, or intolerance towards, the minorities in a given state, the lower the willingness to open up a significant public space for permitted use of that minority’s language in official interactions. As one commentator notes, in some CEE countries (in particular the Baltic States) the assumption seems to be that “minorities should not be denied the right of enjoying
EQUALITY AND MINORITY RIGHTS
213
their culture and using their language, but it must only be within their ‘own’, isolated environments”.109 This, as Will Kymlicka observes, indicates a worrying reversal of assumptions in comparison with the liberal-pluralist ideal: rather than assuming that minority languages can be used in social life unless there is a compelling reason for uniform language regulation, the assumption now is that the state language should be used in all aspects of public life, except for narrowly defined designated areas and environments.110 Almost all CEE constitutions contain provisions stipulating the official language of the State;111 and all but one112 contain provisions stating that minorities are allowed to use their own language. More than half of these constitutions contain the bare permission for groups to use their languages;113 and the majority of these do not attribute these rights to minorities114 but rather state simply that all people have the right to use their native language, a statement obviously directed to minorities. However, eight constitutions go further in their wording, and proclaim the rights of minorities to foster, preserve or develop their own languages.115 Some constitutions allow for the languages of minorities to be in official use within the state. This is mandated either by allowing the minority language to be used officially in all aspects of public life within a certain locality (where a majority of the inhabitants of a certain area belong to that minority),116 or simply by allowing for members of a minority to interact with certain state bodies using their own language.117 In addition, a large majority of these constitutions grant the right to be educated in one’s own language,118 although it is usually weakened by a stipulation that the exercise of this right will be regulated by statutory acts.119 Two of the constitutions that grant the right to be educated in one’s own language assert that the official State language must be learnt concurrently with this;120 while the constitution of Bulgaria is quite unique in laying down a definite obligation on citizens to learn the state language.121 Finally, a number of constitutions contain provisions guaranteeing that persons accused of an offence, or finding themselves in court, have the right to receive information in a language that they understand122 – including the right to a translator in court should they not be fluent in the official State language.123 The tension between the establishment of an official (state) language and the right to use, in official contexts, minority languages when minorities are relatively sizeable and territorially identifiable, remains a constant theme in a number of countries in the region; in particular in the Baltic states (particularly with regard to their Russian populations) and in countries such as Slovakia or Romania (with their Hungarian minorities). In Slovakia, the tension is symbolised by two clauses of art. 6 of the Constitution: the first clause provides that “Slovak is the state language on the territory of the Slovak Republic”; the second, that “The use of other languages in dealings with the authorities will be regulated by law”. The right envisaged by the latter clause has been a constant bone of contention in the relationship between the authorities and the Hungarian minority in Slovakia. The statute on state language of 1990 allowed the use of minority languages where minorities constitute at least twenty percent of the population; governmental practice, however, has often been contrary to that rule. The Slovak Constitutional Court considered a challenge to a number of provisions of the 1995 Law “On the State Language”,124 one of which stated that individuals were obliged to make written petitions only in Slovak. The
214
CHAPTER 8
Constitutional Court held that this contravened the constitutional right to use a minority language in official communications, even though the details of this right were to be regulated by statute. Not surprisingly, Macedonia – a country plagued by strong ethnic conflict between the Macedonian Slav majority and the Albanian minority – has seen a number of challenges to its laws relating to minority linguistic rights. In one decision, the Constitutional Court rejected a challenge to a law that allowed public radio to broadcast in the languages of national minorities (as well as in Macedonian).125 The challenge was based on the official-language provision of the Constitution (Art. 7(1)). The Court found, however, that the state has a constitutional duty to protect the ethnic, cultural and linguistic identity of members of national minorities (Art. 48(2)), and that the ensuing minority rights are not dependent upon a national minority being a majority in a certain locality. Hence, the provision for multi-linguistic public radio did not amount to the creation of multiple official languages in the Republic, and thus did not contradict the constitutional establishment of Macedonian as the only official language. In two other decisions, however, the same Court took a less pro-minority position, and struck down certain provisions as inconsistent with the officiallanguage rule. The difference here was that both concerned the dealings of citizens with public authorities (rather than the regulation of public radio), and, more specifically, the judicial process. In the first of these decisions, the Constitutional Court struck down a provision of criminal law that obliged courts to deliver summonses and other written documents to members of non-Macedonian nationalities in their own language.126 In a Solomonic decision, the Court upheld (on the basis of the principle of fair trial) the right of those persons to use their own language in lodging petitions and in proceedings before the court; on the other hand, however, the Constitutional Court decided that the official-language constitutional provision applied to official documents of the courts unconditionally. The second decision related to a provision of the Law on Civil Procedure, which stipulated that, in local self-government units (in which persons belonging to a national minority are the majority or a substantial part of the total population), the notification of trial dates should be written in the language and alphabet of the national minority.127 The Constitutional Court found this provision of the law to be in violation of the constitutional provision that declares Macedonian, in the Cyrillic alphabet, to be the official language of the state (Art. 7(1)). It should be noted that the opposite decision was genuinely open to the Court; articles 7(2) and 7(3) of the Constitution stipulate that, where there is a majority or a significant number of a national minority in a local self-government unit, their language shall also be in official use (as determined by law), alongside the official Macedonian in the Cyrillic alphabet. However, despite these qualifications on the official-language rule, the Constitutional Court held that courts undertaking official activities cannot use languages that are not official – thus they can use only the Macedonian language in the Cyrillic alphabet – regardless of the proportion of a given national minority resident in any particular area. In other words, the Court found the multi-linguistic provisions of the Constitution to be inapplicable to official court proceedings.
EQUALITY AND MINORITY RIGHTS
215
Poland provides an unwholesome example of a rigid, homogenising constitutional attitude towards the official state language. The constitutional provision that declares that “Polish is the official language”128 leaves no room for the introduction of any minority languages into official fora, even in a restricted manner. While there is an additional sentence in this article, to the effect that that the official-language rule “shall not infringe upon national minority rights resulting from ratified international agreements”, at least one prominent critic of this constitutional provision has argued that it does not add anything to the first sentence, and does not open up the possibility of introducing official minority languages.129 It is therefore not an exception to the rigid rule: “national minorities have not acquired in this Constitution a right to depart from a general rule that Polish is the official language”.130 The above-quoted critic reviews all of the international treaties between Poland and its neighbouring states, and concludes that none contains a rule permitting a minority to have its language officially recognised in Poland. If the constitution-makers had wanted to allow for such a possibility, they would have said so explicitly in the official-language provision.131 The only occasion that the Polish Constitutional Tribunal has been asked to consider the meaning of the “official language” provisions was in its “interpretive decision” of 14 May 1997;132 hence, before the new Constitution entered into force. The subject-matter of the Tribunal’s interpretation was a 1945 decree on the official language (previous Polish Constitutions had not dealt with the issue at all); however, according to the authoritative commentators, this decision also applies to the new Constitution,133 and can thus be seen as a statement of the current official position of the Constitutional Tribunal on the issue of the “official language”. The Tribunal was asked by the President of the Supreme Audit Chamber to provide an interpretation of the official language provisions by saying to whom exactly they apply, and also to which types of official actions they apply. The reason for this request for an interpretation was that the Audit Chamber had ascertained, in the process of its auditing activities, that some “decisions [were being] taken on the basis of documents and reports in foreign languages and that no translations into Polish were made of the contracts and agreements entered into by Polish state institutions or the companies set up by these institutions”.134 The direct trigger for the decision was therefore unrelated to minority languages. Nonetheless, at the end of its lengthy decision (which confirmed that the requirement to use the official language applies to all state institutions, and to all of their official actions) the Tribunal dropped a hint that, as far as citizens are concerned, the official-language rules are applicable only “indirectly” (when they communicate with state bodies), and that constitutional rights and freedoms defined the limits of the duty of state bodies to communicate in the official language. As the Tribunal stated in the very last sentence of its decision: “A citizen, whenever he wants to exercise his fundamental freedoms and rights, cannot be forced to comply with the provisions establishing the official language”.135 Unfortunately, this pronouncement was left hanging in the air: no specific criteria regarding how to reconcile the official-language provisions with the rights of members of minority groups have been identified (to be fair, the Tribunal was not asked to do so in this particular interpretative decision). However, the limits on these rights seem to be very strong and rigid: as the above-quoted authoritative commen-
216
CHAPTER 8
tator136 notes, under the present Constitution the right to use a minority language in public “does not imply that state organs have a duty to issue official certificates (e.g., birth certificates) or conduct court proceedings in the language of ethnic minority”.137 As an example of an acceptable use of this right, he mentions the possibility of conducting a campaign for election to the parliament or to local selfgovernment bodies in a minority language. In other words, no duties upon state bodies are implied by the “fundamental freedoms and rights” to which the Constitutional Tribunal referred and which, allegedly, establish the limits of the officiallanguage provisions. The situation of Russians in the three Baltic republics – in particular in Latvia and Estonia – raises special issues, due to the mixed political and ethnic nature of the problem. On the one hand, just as nowhere in the old Soviet Union were Russians treated as “immigrants”, Russians in the Baltics have not considered themselves as such. The USSR was their natural, single space (with only a sham federal structure), and they knew that, if they settled anywhere within the borders of the USSR, they would be “at home”, without the need to learn local languages, to adapt to local culture, to adjust to a different legal system etc. In fact, Russian was the lingua franca within the entire USSR. Hence, the liberation of the Baltics from its oppressive neighbour left a large number of Russians in those countries, many of them having lived there for some generations by this time, having made use of Russian-language institutions, schools etc. At the same time, however, not only were Latvians and Estonians unhappy to maintain the status of Russian as an official (even if not the official) language in their countries, but they also resented the very presence of Russians, frequently viewing them as the remnants of the old, oppressive regime. Particularly in Estonia, in which immediately after independence the Russian-speaking population was estimated at half a million (one-third of the total population), there was a real fear on the part of ethnic Estonians of a move towards a bi-national, Estonian-Russian state, in terms of which a bi-lingual policy would be a symbolic first step. As a knowledgeable insider, Vello Pettai, has noted: the strength of Estonian nationalist feeling was such that this [bi-national] destiny for the state was rejected by the vast majority of Estonians, and their entire struggle for the restoration of their independence represented for them as much an effort to stem this binational future as a desire to regain their formal sovereignty.138
The same writer observes elsewhere that any steps towards accommodating the Russian-speaking minority into broader social institutions, and of encouraging their meaningful political participation, have so far been weak and ineffective: “Too many Estonian and Latvian memories of Soviet russification remained, while the Russian communities themselves were slow to really mobilize their strength and press for meaningful change”.139 Hence, right after independence, and particularly in Latvia and Estonia, Russians found themselves in a situation not unlike illegal immigrants: stripped of their citizenship and even denied permanent resident status.140 The Constitutional Review Chamber (CRC) in Estonia was twice asked to decide on the constitutionality of imposing Estonian language requirements on electoral candidates running in national and local elections. The 1997 amendments to the
EQUALITY AND MINORITY RIGHTS
217
Language Act provided for language requirements of electoral candidates (as well as the tightening of the Estonian language proficiency requirements for non-Estonian employees in the public and private sectors). As one commentator notes, the law had been “motivated by nationalist desires to make sure that no non-Estonian-speaking person could be elected to parliament or a local council”.141 The President – who challenged the law,142 which of course was very controversial for the Russianspeaking community – did not attack it head-on but rather on technicalities. First, he claimed that the law was unconstitutionally vague: by formulating language requirements in such a vague manner, it ran afoul of the constitutional principle that limitations on a given right cannot distort the nature of that right (Art. 11); thus, he argued that the vague formulation of the language requirements in employment tended to distort the right to non-discrimination. The second charge was that, by delegating the task of controlling electoral candidates’ knowledge of Estonian to the executive branch (namely, the Minister of Education), the law threatened the constitutional separation of powers; the executive would potentially be able to harass these candidates after they have been elected. As is clear, the challenge was based not on minority-rights grounds; after all, it would be hard for minority rights-based arguments to prevail constitutionally over an argument derived from a very strong constitutional rule concerning the official language (Art. 6 and Art. 52(1)), which leaves very little room for any public uses of minority languages. There was, however, at least some room for this: the Constitution provides that “in localities where the language of the majority of the population is other than Estonian, local government authorities may use the language of the majority of the permanent residents of that locality for internal communication to the extent and in accordance with procedures determined by law” (Art. 52(2)). While the last limiting clause effectively delegates to statute the authority to define the use of languages other than Estonian at the local level, the Language Act’s requirement for electoral candidates in local elections to fulfil certain linguistic criteria seems to effectively annihilate the possibility opened up by Art. 52(2), as far as local elected councils are concerned. The Chamber, however, did not follow this path of reasoning, choosing instead the narrow line of argument suggested in the presidential challenge.143 In fact, it preceded its argument regarding invalidation with a nationalist salvo, stating that one of the duties of the State is to preserve the Estonian nation and culture, as evidenced by the Constitution’s preamble and state language provisions. On the basis of these provisions, and additionally of the provision that everyone has the right to address the authorities in Estonian and to receive answers in Estonian (art. 51(2)), the Court inferred that the challenged requirement of proficiency in the Estonian language for the candidates to parliament and councils was not unreasonable. As a matter of fact, the Court did conduct an “activist” or an “expansive” interpretation, but in the direction of undermining any possible claims for minority language rights! On the basis of the Constitution’s preamble (admittedly, a non-typical basis for a constitutional court’s reasoning), which declares that the State will guarantee “the preservation of the Estonian nation and culture through the ages” (note there is no mention of the language!), and on the further basis of the principle that Estonia is a democratic republic (Art. 1), the Chamber concluded that language requirements for electoral candidates could be justified. The argument
218
CHAPTER 8
went as follows: for a democracy to function, those who exercise power must understand what is happening and must use a single communicative system for their mutual communications; hence, only one language (Estonian) may be used in local councils and in the parliament. However, as Pettai has noted, even if this argument properly applies to the actual proceedings in the councils and in the parliament once they are formed, “it did not address in any way the issue of restricting candidate rights during elections themselves”.144 It was, therefore, on narrow technical grounds that the Chamber eventually struck down the controversial provisions. It agreed with the challenger that the law was impermissibly vague insofar as the requirements for employment were concerned, and also that, by delegating the power to regulate the language requirements for election candidates to the government, it was contrary to the separation of powers: decisions connected with electoral rights should be made by the legislature and not the executive. The postscript to the decision is that the parliament properly saw the Constitutional Review Chamber’s decision (and an analogous decision handed down a few months later)145 as a “green light . . . to legislate language requirements for electoral candidates”,146 which it did in November 1998 by passing amendments to the electoral law, which the President soon promulgated despite protests from Russian community leaders. These language requirements were eventually repealed in November 2001 under direct pressure from the OSCE and the EU, not as a result of a constitutional challenge. 7. THE SPECIAL CASE OF MINORITY REPRESENTATION IN PUBLIC AUTHORITIES The most far-reaching proposals for political protection of minority rights contain the demand for special political representation at a national or local level, implemented through some “sui generis minority mandates”.147 There are, of course, a variety of milder methods of strengthening the political representation of minorities, such as by increasing “the role of the minorities ombudsman, local and national minority self-government, effective official lobbying mechanisms incorporated into parliamentary decision-making, and parliamentary candidates of minority parties and social organizations, who have gained their mandates under ‘ordinary’ election laws”.148 There are also some ways of facilitating the representation of minority parties in parliaments, without at the same time creating special quotas of seats set aside for these parties: some constitutions provide, for example, preferences in the form of a more lenient election “threshold”. This is the case in Poland, where electoral committees representing ethnic minorities do not have to pass the five- percent threshold to achieve parliamentary representation.149 Similarly, in Lithuania, the organisations representing ethnic minority parties were exempt (until the 1996 amendment to the 1992 election law) from the four percent threshold needed to elect candidates under the proportional rules (which apply to one-half of the MPs, the remaining half being elected through a majoritarian system).150
EQUALITY AND MINORITY RIGHTS
219
Often those seats set aside for minority parties in parliament are subject to a specified electoral result. For instance, the Romanian Constitution reserves one seat in the parliament for each ethnic minority organisation that fails to obtain a sufficient number of votes to get elected in the normal manner;151 the electoral law clarifies that this is subject to obtaining at least five percent of votes.152 In effect, Romania has emerged with the most extensive system of minority representation in the region; and, in successive elections, the number of ethnic parties has continually risen, due to the low entry barrier, sometimes even raising the question of the fairness of the system towards non-ethnic parties.153 In Slovenia, the Hungarian and Italian minorities can elect at least one candidate each to the National Assembly.154 In Croatia, where the Constitution remains silent on the matter, according to electoral law all nine specified ethnic communities are entitled to one seat each, plus any ethnic community that exceeds eight percent of the population (only the Serbian minority meets this requirement) is entitled to additional proportional representtation.155 In Hungary, apart from the constitutional right to be represented in national and local bodies,156 national and ethnic minorities have a constitutional right to form their own minority self-governments.157 The statute on ethnic and national minorities provides that, when a local government is elected with over half of the representtatives elected as minority candidates, they may declare themselves as a minority self-government. By the mid-1990s it had been reported that over 800 such minority self-government units existed in Hungary,158 although it must be added that serious doubts have been expressed as to the resources available to, and powers and effects of these bodies.159 When it comes, however, to formal parliamentary representation of ethnic minorities, despite an impressive number of assorted legislative proposals aimed at designing an acceptable system,160 no political consensus has emerged as yet to allow the adoption of a statute to regulate this issue – even though the Hungarian Constitutional Court declared that the absence of mechanisms to implement the constitutional requirement for parliamentary representation of minorities was unconstitutional.161 At the opposite end of the spectrum are the countries that either openly or tacitly oppose the possibility of any formalised ethnic representation in parliamentary bodies. This is the case of Albania162 and Bulgaria,163 which ban (in their statutes and the Constitution respectively) any parties based along ethnic lines (although the Constitutional Court in Bulgaria has softened this regime, in a decision to be discussed below); and also Russia, where “a number of tacit regulations create a legal climate in which, despite the multiethnic conditions, there is no ethnic party in the more powerful chamber, the Duma”.164 There are also a number of CEE countries that, while not prohibiting it, make no constitutional or statutory allowances for the political representation of minorities in the highest and most powerful representative organs: the Czech Republic, Slovakia, the Ukraine, Belarus, Macedonia and Moldova all belong to this category.165 One example of an intervention by a constitutional court actually preventing a system of ethnic representation is provided by the Slovakian Court.166 In that country, there was an attempt to introduce by legislation a rule of proportional representation of ethnic groups. It was struck down under the constitutional
220
CHAPTER 8
principles of equality and political competition. The case before the Constitutional Court related to the local self-government electoral law of 1998 which stipulated that, in towns and villages where national minorities or ethnic groups lived, the total number of deputies in local elections must be divided proportionately, resulting in a faithful reflection of the ratio between Slovaks and individual minorities: in effect, a quota system for particular minorities (and Slovaks) was created. The Court found that such a system was contrary to the constitutional rule of equal access to public offices (Art. 30(4)), the principle of equal dignity (Art. 12(1)), and to the constitutional provision that states that the regulation of political rights must facilitate political competition in a democratic society (Art. 31). In effect, the Court rejected any idea of “preferential quotas” in order to improve the status of a national minority or ethnic group, and opted instead for the individual-civic principle: all citizens are equal in exercising their political rights, regardless of group membership. The question of defining who belongs to an ethnic minority, in cases where the group is given certain special political rights, arose in Slovenia.167 The Constitution provides for some extensive special rights for Italians and Hungarians, including the rights of these two ethnic communities to be “directly represented at the local level” and also to be represented in the National Assembly (Art. 64(3)). This gave rise to the question of how to identify the members of those minorities, as they possess special voting rights: in effect, they may vote twice – for a member of their national community, but also for other delegates as well. The challenged provision stated that there would be special electoral rolls set up by the self-governing Italian and Hungarian communities and then confirmed by an organ of the jurisdiction. Another article of the law stated that members of these communities not living in the regions in which these minorities make up a significant part of the population could be inscribed on this roll upon their written request. The Constitutional Court was concerned that the criteria for determining who belonged to the minority community were not defined either in the Constitution or in any other law; thus, inscriptions in the electoral roll were being carried out without any statutory guidelines on who should be included. Since membership of these communities is a status to which the Constitution attaches special rights, the criteria of belonging to the minority must be statutorily determined. While everyone has a constitutional right to express freely his or her affiliation to any nation or national community, if the will of each individual were, in this particular context, to be decisive, the potential for abuse and undermining of the “real” will of the minority community and their constitutional rights (contained in Article 64 of the Constitution) is readily evident; others could be entered on their electoral roll and affect the outcome of elections. In consequence, the Court declared the legal provisions to be unconstitutional on the basis of their conflict with the rule of law, and demanded that the legislator correct this state of affairs. Probably the most important decision on racial and ethnic equality in CEE (although not argued in terms of non-discrimination) was the decision of the Bulgarian Constitutional Court of 22 April 1992, on the status of the Movement of Rights and Freedom (MRF).168 According to the petitioners in this case, a group of fifty-three Bulgarian Socialist Party deputies, ninety-nine percent of the membership
EQUALITY AND MINORITY RIGHTS
221
of this Turkish-based organisation belonged to the Turkish minority. It had twenty four deputies in the parliament and was a crucial member of the coalition dominated by the Union of Democratic Forces: without its support, the liberal government formed after the October 1991 elections could not survive. The petition to the Constitutional Court demanded that the MRF be outlawed (and, in consequence, its MPs excluded from the parliament) on the basis of article 11(4) of the Constitution, which prohibits political parties based on ethnic, racial or religious lines.169 Of course, the very idea of outlawing a political group on the basis of its ethnic composition and programme is anathema to the principle of liberal diversity.170 This was included in the Constitution at the insistence of the post-Communists, who won a majority in the first multi-party parliament and could thus shape the design of the hastily adopted Constitution in accordance with their own preferences, which also included an embrace of nationalism as their new creed. To question the validity of this constitutional pronouncement was naturally beyond the Constitutional Court’s reach. According to one convincing interpretation, the petition to outlaw the MRF was dictated less by ethnic and national animus and more by pragmatic partypolitics considerations: “the attacks against MRF were used – not very successfully – as a public relations device to boost the political fortunes of the former communists”.171 In any event, the “pro-MRF” judges rejected the argument of the petitioners, and engaged in a rather creative interpretation of the Constitution, stretching perhaps their power of interpretation to its limits, given the relatively clear text, but, admittedly, in a good cause. They interpreted the true meaning of Art. 11(4) as a ban on parties that actually exclude potential members on the basis of their national or ethnic origins. Since the MRF’s constitution did not contain any such exclusionary rules, the ban contained in Art. 11(4) simply did not apply to it. A party cannot be said to contravene Art. 11(4) on the basis that the majority of its members belong to a particular ethnic or religious group – the Constitutional Court argued – because, if this were the case, all parties that have the word “Christian” in their name would have to be judged unconstitutional. The “pro-MRF” judges also engaged in a discussion of the MRF’s programme and discerned that it was opposed to the ideas “of autonomy, national chauvinism, revanchism, Islamic fundamentalism and religious fanaticism”,172 although, at the same time, it demanded a broadening of the protection of rights of ethnic and religious communities. The Court also sketched a historical-justice analysis: it recalled that the rights of the Turkish minority had been blatantly violated in the past, as was acknowledged in the parliament’s declaration of 15 January 1990, and that the emergence of the MRF on the political scene must be seen as a natural reaction of the Turkish minority to these violations. It actually stated that the very formation of the MRF was an “immediate consequence of the acts perpetrated by the totalitarian regime against a part of the Bulgarian citizens”.173 In addition, the judges argued that the MRF was not a “party” but a “movement”, as evidenced by its registration in a regional court as a “political organisation” and, additionally, by the refusal by a regional court to re-register MRF as a “Party for Rights and Freedoms”. This was a courageous decision, and it provoked a passionate debate, all the more so as it was effectively decided by a minority of judges (five against six): the MRF’s parliamentary presence survived only because the Constitution provides that
222
CHAPTER 8
decisions of the Constitutional Court must be taken by “more than half of the votes of all Justices”, which means, in effect, that seven judges must be in favour of declaring a party (or a statute, for that matter) unconstitutional.174 As Venelin Ganev has argued, the decision of the Court “effectively obviated [the] nationalistic, restrictive intent” of Art. 11 (4) and “contributed towards the advancement of the process of ethnic reconciliation”.175 With the benefit of hindsight, it is possible to say that the decision has certainly not aggravated ethnic relations in Bulgaria but rather has, if anything, contributed to a further reduction of negative stereotypes towards the Turks in that country. As one Bulgarian expert in ethnic politics observes, “The fact that [the Turks and ethnic Bulgarians who have adopted Islam] are represented in public life by an independent political organization finally legitimized them in the eyes of Bulgarian society”.176 8. CONCLUSIONS In contrast to civil and political rights, discussed in the previous chapter, the jurisprudence of the constitutional courts of CEE on equality has not marked such a radical shift away from the values and political principles of the ancien régime: it was more a matter of the careful fine-tuning of standards of non-discrimination as written into the new constitutions. For reasons noted earlier in the chapter, the constitutional clauses of equality served as useful and attractive legal bases for constitutional challenges, and the courts have made a valuable, if careful, contribution towards removing some aspects of discrimination in their respective societies, such as against women, gays and lesbians, etc. When it comes to minority rights and ethnic relations in general, the contribution by the constitutional courts has been modest, and it would be much too generous to claim (as some Western observers have done) that these courts were crucial elements in shaping the “toleration regimes”.177 Apart from the shining exception of the Bulgarian MRF case, there have been virtually no significant decisions by constitutional courts in CEE that would provide support for this statement. In fact, there have been rather few decisions dealing with ethnic/national problems, even in the places where one would expect them. The case of Estonia is quite instructive in this regard: ultimately, it was only international pressure, and not the intervention by the Constitutional Review Chamber, that compelled the parliament to amend the law discriminating against the Russian-speaking minority. In other post communist countries, very few ethnicityrelated decisions have been made by constitutional courts, and, where they have, they would hardly support the thesis that these courts have played a central role in shaping the regimes of toleration.
CHAPTER 9 “DECOMMUNISATION”, “LUSTRATION”, AND CONSTITUTIONAL CONTINUITY
The post communist countries of Central and Eastern Europe (CEE), at the outset of the transition to democracy, had to face not only the difficult problems posed by the present and the future, but also those stemming from the past: how to handle the widespread violations of human rights, the travesty of legality, the pervasiveness of collaboration with the secret police, and the often blatant crimes conducted for political reasons that characterised the immediate past. The peaceful, negotiated form of transition that prevailed in the region meant that the various perpetrators of these unsavoury acts have remained very much part of their respective societies during the transition to democracy, and often belong to the political elite that was (co-)responsible for the move away from the very political system in which they had been active participants. How is it possible to reconcile legal tolerance towards those people with the requirements of political justice and with the exigencies of a democratic society? Should politics be “cleansed” of those discredited by their direct involvement with the Communist regime, with the risk of creating a category of second-class citizenship and the resulting danger of disloyalty towards the democratic state, or rather adopt the “let bygones be bygones” attitude and implement, for the sake of national reconciliation, a policy of forgiveness, even if not of forgetting? “Amnesty but not amnesia” was indeed one of the slogans of some of the liberal democrats immediately after the transition. This, however, was easier said than done: can one move to a new, democratic system of government without settling the accounts of the immediate past? The worldwide precedents suggest that – with very few exceptions, such as in Spain where the transition from the Franco era, which occurred in 1976-1978 and was presided over by King Juan Carlos, involved an amnesty for all political crimes committed in the past (both by government forces and the opposition), the sealing of police files, and a general abstention from any purge of those implicated in the Franco regime1 – most post-authoritarian systems established various state-sanctioned mechanisms of retribution, apology, shaming, purges, reaching “truth and reconciliation”, etc. This is, of course, only one aspect of so-called transitional justice, i.e. of the legal mechanisms adopted in order to help come to terms with the legacy of an immediate authoritarian past by a successor democratic regime. (Some authors use the concept “retroactive justice” interchangeably with “transitional justice”;2 I prefer to use the adjective “transitional” throughout because the word “retroactive”
224
CHAPTER 9
suggests that the main rationale of these forms of justice is backward looking, which is, in itself, a substantive argument that is often question-begging, as will become clear below). Other possible forms of coping with persisting manifestations of the authoritarian legacy include trials of the perpetrators of politically-motivated criminal acts in the prior period, restitution of unlawfully (or unfairly) taken property, truth commissions, access to the files of former secret police, official apologies, etc. Each of these ways of handling the past by transitional polities raises a host of difficult and controversial moral and legal problems; as such each deserves separate treatment. I have chosen the two that I consider to be the most controversial and the most widely discussed questions of transitional justice in post-communist societies: the issue of “lustration” (or purges) of public officials for their past involvement in the Communist regime, in particular collaboration with secret police, and the question of reopening of the statutes of limitation for criminal offences committed for political reasons during the Communist era. The moral and political significance of both of these issues is beyond any doubt: “lustration” is seen by its proponents as the main device of “decommunisation”, that is, of cleansing the public sphere of the newly democratised society of those who have shown their utter disregard for the values of democracy and liberty while suspension (or reinstatement) of statutes of limitation is seen as the removal of selfserving immunity that the perpetrators of crimes had conferred upon themselves. By its opponents, lustration is viewed as a tool for divisiveness and as an expression of collective responsibility for the past (one author even suggested that “the process of de-communization in Eastern Europe has sometimes been carried out with something like Communist disregard for individual rights”),3 and tampering with statutes of limitations as a violation of the principles of legality of the very sort that characterised the old regime. Both of these forms of transitional justice also raise the most dramatic ingredient of the transition, namely what to do about the persons implicated – sometimes, in a criminal way, and, almost always, in a morally repugnant way – in the oppressive regime. It is this that has proved the most difficult question the in post-communist societies of CEE, and not the issues of property (as in restitution) or of providing redress to the victims (apologies, truth and reconciliation, access to secret files, etc). Most importantly, for my purposes, both of these forms of transitional justice have elicited diverse but important responses in the constitutional discourse in post-communist countries, not the least in the judgments of the constitutional courts. 1. MAIN DILEMMAS RAISED BY DECOMMUNISATION AND LUSTRATION LAWS A whole host of moral, political and practical concerns rendered the issue of “lustration” and decommunisation one of the most vexed and divisive questions of post communist transition. The very terms came to denote different things in different places; perhaps the most widespread use of the notion of “lustration” applied to those instruments that were meant to reveal the real (and largely hidden) past involvement of some people in the Communist regime and its practices, and,
“DECOMMUNISATION” AND “LUSTRATION”
225
depending upon the circumstances, that specified further consequences such as the loss of a right to hold certain public positions. However, regardless of the consequences, the starting point of lustration was the exposition of the real past of certain individuals, mainly for their prior collaboration with the secret police; it has been used as an instrument policy, the centrepiece of which is to reveal the actual but hidden truth about the exact involvement of some people in the old regime. “Decommunisation”, in contrast, was supposed to consist in the removal of exmembers of the Communist Party (above a certain rank) from prominent positions in the new democratic system. Thus, it encompassed not simply the exposure of individual guilt of past collaborators with the secret police, but also the more collective removal of ex-Communists as a group from positions of influence upon political decisions in the democratic state. Of course, the borderline between “lustration” and “decommunisation” thus understood is not precise, as group characteristics (such as belonging to a particular sphere of “nomenclatura”) are seen – often accurately – as a sufficient sign of deeply discrediting involvement in the invidious practices of the past. Nor does past- or future-orientation mark a practicable distinction between “lustration” and “decommunisation”. To be sure, lustration has been often defended on the practical grounds based on the political requirements of democracy rather than with securing justice for previous wrongs. But precisely the same considerations – alongside more moral, justice-related arguments – have been raised in the calls for “decommunisation”, viewed as a guarantee that a democratic system of government will not be eroded by those more used to authoritarian methods of exercising political power. For this reason, my use of both words here, lustration and decommunisation, follows the conventional usage in CEE countries. “Lustration” applies to the screening of persons seeking to occupy (or actually occupying) certain public positions for evidence of involvement with the communist regime (mainly with the secret security apparatus), while “decommunisation” refers to the exclusion of certain defined categories of ex-Communist officials from the right to run for, and occupy, certain public positions in the new system. However, in public debates on the moral and legal rationales for and against the policies covered by these concepts, the two have been often lumped together, as in this statement by Adam Michnik, one of the leading critics of lustration and decommunisation in CEE: The principle of de-communizing, the idea behind it, is that a certain number of Communist functionaries of the Communist regime, or of the Communist party, would be stripped of their constitutional rights en bloc, only for that reason – that they held certain positions in the Communist party. The lustration idea is that by using the materials of the political secret police, the past of various personalities active in public life would be examined. In other words, the de-communizing idea is a direct continuation of the Bolshevik concept of citizens’ rights, summed up by the English writer George Orwell in his novel ‘Animal Farm’, that all animals are equal but some are more equal than others.4
This quote from Michnik clearly shows the central moral and, in consequence, legal dilemma that arises when a post-authoritarian state attempts to come to terms with its immediate past. Can a break with the past be made in a way that does not
226
CHAPTER 9
itself compromise the legal and constitutional attributes of a democratic state: the rule of law, equality before the law, the individualisation of guilt, the nonretroactivity of legal sanctions, etc? If not, and if the past is thus going to be a matter for study by historians only, will it not keep haunting the newly democratised society? Significantly, many thoroughly democratically-oriented, non-Communist participants in the debate (such as Michnik himself) have expressed strong reservations concerning the policies of lustration and decommunisation. Vojtech Cepl, an eminent Czech legal scholar – and later, a judge of Constitutional Court – admitted right at the outset of the post-Communist era to his own “deep ambivalence about lustration”. On the one hand, he recognised the need to know about evil committed in the past, and the need (inter alia) to preclude the possibility of “entrusting our future to people who can be continually blackmailed”; on the other, he expressed concern both about “careless and indiscriminate lustration”, and also about the main thrust of the lustration being directed against low-level executioners of orders rather than against those who actually issued those orders.5 Similarly, JiĜina Šiklová has identified the principle of collective guilt as the principal objectionable characteristic of the lustration law.6 (This, incidentally, is admitted even by some defenders of the policies of lustration who actually argue that collective determination and a presumption of guilt reduce, rather than strengthen, the deleterious character of the lustration laws, because the “deliberately overbroad, prophylactic bans on officeholding” carry less stigma than individualised determinations of unsuitability for office would!)7 In addition, much of the debate around lustration has focused on the unreliability of the documents that, inevitably, were to become the main proof of past collaboration; namely, the files of the former secret police apparatus. Much of the documentation is incomplete, having been destroyed during the days of transition (in Czechoslovakia around ninety percent of secret police registers were destroyed after the Velvet Revolution: the unreformed secret service still controlled the archives until June 1990, and similar facts have been reported in many other CEE countries).8 More importantly, those documents that remained often did not make it clear whether a person was an actual informer/collaborator or merely a candidate for such a role; the police agents often inflated the numbers of collaborators co-opted in order to improve their status and financial rewards. As a result, what was found in the police archives was both over-inclusive (many names of phoney agents) and underinclusive (missing files of many genuine collaborators). Another theme in the debates around lustration concerned its consequences for a newly democratised society: is it wise to re-open old wounds and create divisions in a society that needs a high degree of coherence to handle the challenges it is facing? Those who answer this question in the negative have relied heavily on the divisiveness of the effects of lustration in their argumentation. Lustration was embarked upon in order to clear up the secrets of the past and to make for greater transparency and openness, but its immediate consequence must have been the exclusion of some people from the first category of citizenship; as Ruti Teitel puts it, “perhaps paradoxically, greater access [to the old state files] would still mean political exclusion”.9 The exclusion occurred not only through formal measures of screening for important posts, but also at a psychological, interpersonal level.
“DECOMMUNISATION” AND “LUSTRATION”
227
Šiklová gives a telling account of the results of “unofficial” lustration conducted by some Czech newspapers that carried the list of real or alleged collaborators of secret police: “Many people, who found themselves listed, literally collapsed when they saw their names. Others searched for their own name with trepidation and then, with perverse pleasure, for the names of friends and family members”.10 And what about the intrinsic morality of lustration: is the moral guilt of ex-informers (some of whom were blackmailed or threatened into what they did) sufficient to inflict upon them now such a tremendous social penalty in the form of – at the very least – moral opprobrium? And conversely, is the sense of outrage on the part of the “lustrators” always based on respectable motives? Wrongful motives may taint otherwise plausible policies, and if that is the case, then lustration may originate from sources that are dubious in nature, such as the sense of guilt and shame for not having done much to counter the dictatorship. Significantly, some of the bravest dissidents under Communism turned out to be strong opponents of “lustration” (such as Vaclav Havel or Adam Michnik), and, conversely, some of those complicit in some forms of oppression, or at least passive towards it, emerged as virulent proponents of lustration. As Jon Elster suggests: “I suspect that some spontaneous executions that took place in the aftermath of liberation in France and Belgium, as well as some demands for lustration and retribution in Eastern Europe, owed more to a bad conscience than to rightful anger”.11 Reluctance to endorse lustration may be therefore seen, partly, as a prudential strategy of not allowing public policy to be dictated by illegitimate motives.12 The central tension, it seems to me, in lustration laws and the public discourse around them can be seen as a clash between the retributive, past-oriented rationale that can best account for the shape and reach of the lustration laws and, on the other hand, the consequentialist, future-oriented arguments that are used as official, public arguments for lustration. The sources of the retributive kind of rationale are well encapsulated in the following observation by Vojtech Cepl: “After decades in which everybody was afraid of everybody else and even close friendships were marred by suspicion, the people of Czechoslovakia are ready for some sort of retribution…”.13 This applied, with lesser or greater force (depending on the harshness and repressiveness of the old regime, and on the role that members of the Communist elite played in the transition itself) to all of the countries emerging from the collapse of Communism, and the basic urge for historical justice, according to which those who were the perpetrators and supporters of the fallen regime should pay, in one way or another, for their past sins, was understandably strong. However, the historical-justice argument rarely figured in the more formal, legal discourse surrounding the lustration laws, in particular insofar as this discourse was controlled by constitutional courts: it seemed that to base such laws on the need for retribution would come too close to revolutionary vengeance, and would contradict too much the dominant rule-of-law ideology.14 Hence, the dominant politico-legal argument concerned the need to protect the new democratic regime from the dangers posed by staffing its key institutions with people whose democratic credentials are doubtful, to say the least. Not only can they be bound by various forms of old loyalties and open to blackmail but, more significantly, their past is evidence that they are not
228
CHAPTER 9
suitable to function in a democratic state. The following is a good statement of this kind of rationale: The democratic method . . . requires a guarantee of the certainty that its office holders will, under all circumstances, heed the democratic rights of citizens. … Anyone who consciously participated in suppressing the rights of citizens is a potential danger to a democratic society, and thus does not meet the prerequisites for important positions in state administration…15
The tension to which I referred earlier is that the lustration laws seem badly suited to the function of protecting the democratic state against non-democratically minded personnel; there is no reason to believe that a large number of the old functionaries of the authoritarian system who behaved as they did for a large variety of motives, including fear, opportunism and a general lack of better prospects, will necessarily undermine the democratic rules of the game in a totally new apparatus, operating in accordance with different modes and procedures. A variant of this rationale relies not so much on the argument that those implicated in the past in collaboration have no qualifications to participate in the actions of a democratic state, but rather that they are liable to various forms of blackmail on account of those past affiliations. This point has frequently been made by proponents of lustration laws; and yet the speculative nature of the argument renders the proposed method of handling this danger distinctly over-inclusive. How many ex-Communists would indeed be blackmailed by former security agents? It could even be argued that blackmail is more likely to occur when a lustration law is in place – because the official sanctions for false declarations make the blackmail more effective than when the only sanction for past involvement is social reprobation. In turn, the rationale that seems to best account for the shape of lustration laws, namely the retributive one, was considered inappropriate to figure prominently in public justifications for such laws. It was only rarely that official legal discourse explicitly admitted that the aim of lustration may be “a genuine public disclosure of the nature of the previous regime, to guarantee a measure of redress, and simultaneously to symbolize the irreversibility of the changes, through revealing the activities of the secret services”.16 The problem with this perspective on lustration is that the law is a very imperfect instrument for performing these tasks, and, in the process of achieving public disclosure and a symbolic break with the past, the law risks creating all sorts of negative side effects related to the collective guilt approach. (Not to mention the obvious argument that if the point of “lustration” is purely symbolic then it should be operated with symbolic means – such as emphatic declarations about the past – rather than by more punitive methods involving the exclusion of certain persons from certain positions). The symbolic function of “lustration” as a ritual breaking with the past seems inconsistent with the nature of a liberal-democratic state, which abstains, as far as possible, from engaging in actions of a didactic or symbolic nature. The ideological neutrality of such a political system demands that the functions of performing rituals of symbolic transition are better left to non-state actors: media, non-governmental organisations, professional associations, writers, journalists, historians and other opinion-leaders, rather than to state officials. In the context of post-Communist lustration, the danger of involving officialdom in symbolic rituals aimed at “breaking with the past” lies in the fact that
“DECOMMUNISATION” AND “LUSTRATION”
229
the past has been the subject of contestation, dispute and divergent interpretations: there is no single, consensual view as to the nature and effects of Communism. One can, of course, deplore such a state of affairs; however, to disregard this ideological disagreement, and to call for public acts the aim of which is (among other things) to establish an emphatic, official interpretation of the immediate past, risks placing those who disagree beyond the margins of the legal and political framework of the newly democratised polity. This point was stated, in an exaggerated and clearly self-interested way, by those post-Communist deputies to the Czech Parliament who objected that the collective assessment of the evils of Communism, conveyed in the parliamentary act of 1993 entitled “On the Lawlessness of the Communist Regime and Resistance to It” (discussed in more detail below in Part 3 of this chapter), contained “a doctrinaire evaluation of an historical period of the former Czechoslovakia, introduced in the form of a statute [that] excludes other opinions and conclusions resulting from scholarly knowledge of historical facts”.17 This is an interesting act as it represents an uncommon attempt to elevate the evaluation of a period of communism to the rank of positive legislation. The statute contains also some more practical rules – concerning, for example, the suspension of statutes of limitations for crimes committed under Communism and not prosecuted earlier for political reasons, to which I will return below. But at this point it is useful to look at the “declaratory” aspect of the statute, which, both in its Preamble and in the first part, asserts that the Communist Party of Czechoslovakia, its leadership and its members are responsible for the manner of rule between 1948 and 1989, proclaims the joint responsibility of those who supported the communist regime for crimes and other arbitrary acts committed during this period, and declares the Communist regime to be criminal, illegitimate and abominable.18 In the explanatory report on the statute the Parliament indeed made it clear that one of the rationales for legislating in this way was that “there has not been a more comprehensive definition and characterisation of the injustice and crime of the dictatorship, by which the whole society became so deeply and systematically marked”.19 Is an act of parliament the right mechanism to provide a “definition and characterisation” of a particular historical period? Framing the question in such general terms seems to yield an immediate negative response; the matter is, however, made more complicated by the fact that the law concerns the directly preceding period, and that the “definition and characterisation” it provides may be seen as instrumental to, and supportive of, the real aim of the statute, which was to suspend the limitation period for the prosecution of certain crimes. What is worrying about including such a characterisation of Communist rule in the actual articles of a parliamentary statute is that it sets out a fixed ideological orthodoxy as to how the past should be viewed. This was precisely the argument of those who challenged the statute before the Constitutional Court, and is encapsulated in the sentence about a “doctrinaire evaluation of a historical period”, and “exclusion of other opinions”, already quoted above.20 The suggestion of “exclusion” of other opinions is extravagant, particularly when linked to the preposterous claim that such a declaration is contrary to the freedom of research in the conduct of scholarly and journalistic activities.21 Nevertheless, the fact is that any such authoritative statement involving an assessment of the immediate past will create an official orthodoxy on
230
CHAPTER 9
matters about which there is considerable disagreement, even if that orthodoxy does not violate any specific constitutional rights (no-one is compelled to adopt, or punished for not adopting, this official line in their writings, lectures, etc). The Court responded to this claim by saying that the declaration of illegitimacy of the regime, and the joint responsibility of all supporters of it, was merely an acceptable expression of the Parliament’s moral-political view on the matter. It creates no criminal sanctions in itself and does not contradict the right to research or express alternative opinions (as the challengers claimed), and thus breaches no constitutional rights. The Parliament – according to the Court – had a right to express “its moral and political viewpoint, by means which it considers suitable and reasonable within the confines of general legal principles – and possibly in the form of a statute”.22 However, distinguishing between the declaratory and the binding aspects of a statute, and asserting that controversial moral judgements may be contained in a legal act as long as they are non-binding, is not quite satisfactory, because the legitimacy of the legislator to speak for the entire society is put to a severe test. Of course, it is a commonplace that, when there is disagreement over a particular rule or practice, the legislator, upon deciding that it is a suitable matter for legal regulation, of necessity sides with the majority, and will thus in the end leave the minority unsatisfied. When a statute does this through imperative rules, it is inevitable; when, however, the Parliament does so through declaratory statements, the divisive and alienating effect is something that can be avoided. In a thoughtful analysis of the Czech statute, JiĜí PĜibáĖ has defended the parliamentary declaration, and the Court’s support thereof, by distinguishing between the Parliament acting “as a representative body of the sovereign political will of the people”, in which case it “functions to strengthen the fictionalized general will”, and, on the other hand, when “it expresses moral or political judgments”, in which case it “must naturally exclude competing political or moral judgments that are incompatible with Parliament’s expressed judgment”.23 The former function (“fictionalized general will”) is performed when the Parliament issues imperative rules; the latter (“moral or political judgments”) when it issues declaratory, non-binding statements. However, this distinction is questionable because it converts an empirical, political observation about moral disagreement in society into a normative theory legitimating legislative declarations of norms about which people do disagree. When the legislator has to regulate a particular matter in an imperative way, the appeal to majority rule (subject to some precautions mandated by the existence of minority rights) is inevitable, and the “fictionalized general will” is in such cases much more “fictionalized” than “general”. Partiality is simply inescapable: when some people demand the suspension of statutes of limitations for politically motivated crimes and others reject it, there is no way for the Parliament to stay neutral as between these demands for a particular imperative solution. When, however, what is at stake is the assessment of a particular fact or period, with no specific binding consequences attached thereto, and there is considerable controversy in a society as to this assessment, the endorsement of one viewpoint in the controversy is politically costly (because it alienates a section of the population) and morally dubious (because it goes against the liberal principle of neutrality of the state). One may object, naturally, that no such neutrality is warranted when the assessment of an authori-
“DECOMMUNISATION” AND “LUSTRATION”
231
tarian, oppressive regime is at issue. Perhaps. And yet the very fact of an actual disagreement in the society as to the details of such an assessment suggests that evaluations of a declaratory nature create a danger of leaving those who disagree beyond the pale of the new political community; even if only in a symbolic way; they are placed in a position of second-class citizenry. To say that they have brought it upon themselves by siding with the wrong viewpoint is not a satisfactory response, because it relies upon an idea of moral homogeneity that is anathema to a liberal approach to the state. Hence, such evaluations are better left outside the sphere of authoritative actions of the state, and relegated to the operations of civil society. In the end, it is a question of whether a democratic society wants to entrust an authoritative body – even if democratically elected and accountable – with the task of declaring an official truth about the past. The name of an institution set up in Poland, the Institute of National Memory,24 exemplifies this danger; the Orwellian title betrays an ambition to “nationalise memory”, to establish an institutional truth about the past, and to exclude those who dissent from that truth. This, it should be noted, is not an argument against any societal attempts to establish the truth about the past. After the traumatic overthrow of a repressive regime, which had been producing deceptive, propagandist accounts of its own, this reaction aimed at constructing the truth about the past, through various official or quasi-official bodies such as Truth Commissions, is a natural and healthy antidote. It is also an important means of providing satisfaction to those who have been directly victimised by the previous regimes. There is nothing necessarily illiberal in a posttransition society attempting to regain the truth through collective institutions. However, the difference between these truth commissions and the Czech declaration is that the former – such as the Commission on Historical Clarification in Guatemala or Commission for Truth and Reconciliation in South Africa – typically work their way from detailed, single-case accounts to generalised characterisations of the regime, rather like the way in which an individual historian constructs a synthetic account on the basis of cases, facts and events. They are also meant to establish a sort of consensus, or at least a non-adversary forum for reflection on what happened; in such commissions, [p]ublic knowledge about the past is produced through elaborate processes of representation by perpetrators, victims, and the broader society, grounding the historical inquiry with a basis for social consensus. It is a truth that is publicly arrived at and legitimated in nonadversarial processes that link up historical judgment with potential consensus.25
Equally importantly, “[t]he truth reports are not generalized accounts but detailed documentary records. . . . The greater the detail, the stronger the counterweight to prior state silence. The more precise the documentation, the less is left to interpretation and even to denial”.26 But the Czech statutory declaration is nothing but interpretation, and – at best – a “generalised account”. This is a generalised characterisation – accurate, in my view – of the nature of the Communist period, and of the interpretation given to that era in recent Czechoslovak history. As such, it is not the basis for a consensus-seeking establishment of the facts about the past, but rather a dissensus-provoking state orthodoxy about the ideological lenses through which the past should be viewed.
232
CHAPTER 9
Before presenting an overview of the systems of lustration and decommunisation as adopted in CEE, and the responses of constitutional courts thereto, two remarks are necessary that will help to situate these developments within a broader context. First, the adoption and implementation of these programs has not taken place in an international political vacuum. In particular, the by-and-large lukewarm attitude of many of the constitutional courts may, to some extent, also be explained by external influences that, on balance, inclined the legal systems of CEE against rather than in favour of lustration. In contrast to the post-World War II situation, when the Western allied powers pressured Germany to conduct a wide-ranging denazification of its state apparatus, after the fall of Communism the stance of Western democratic governments and commentators was quite the opposite or – at best – indifferent. Indeed, some Western observers came to view the very idea of “corrective justice” as antithetical to constitutionalism itself. This argument is nicely encapsulated in the following assertion by Bruce Ackerman: “An emphasis on corrective justice will divide the citizenry into two groups – evildoers and innocent victims. … Constitutional creation unites; corrective justice divides”.27 One commentator, of Eastern European background, complained recently that “seeing lustration with ‘Western eyes’, i.e. democratic and liberal but lacking an understanding of the totalitarian experience, became extremely popular among centre and left-wing politicians, who persuaded their electorates that ‘doing nothing’ was the best and most prudent strategy of breaking with the legacy of the past”.28 However, this explanation, with its implied accusation of ignorance of local conditions and ideological bias,29 is only partially true. More importantly perhaps, Western observers were keen to see the new democracies of Central and Eastern Europe become “normal” as soon as possible, and this meant, crucially, the adoption of the characteristic standards of consolidated democracies, including robust guarantees of the rule of law, due process, and formal equality. A degree of universality is written into such a notion of “normalcy”, and it is true that the “main motivation for the ‘westerners’ who became interested in post-totalitarian retroactive justice was support for liberal values, like democracy and human rights and a wish to spread justice around”.30 Quasi- and nonjudicial forms of investigating into people’s past outside the usual mechanisms of criminal process, as well as the creation of whole categories of people deprived of certain political rights, are anathema to such a political programme, and the universalistic liberal ideal was used as a yardstick to judge the preparedness of the new democracies to join first the Council of Europe, and then the European Union. Western European political elites developed a stake in the integration of the exCommunist East into the overarching political structures of the continent, not least for the sake of stability and peace in Europe. The criteria of “normal” democracy, untainted by any extraordinary measures related to its immediate non-democratic past, were instrumental in achieving these aims. The resolution adopted in 1996 by the Parliamentary Assembly of the Council of Europe, which outlined how lustration procedures should be designed in order to be compatible with the principles of the rule of law, is very interesting in this regard. 31 While it does not condemn the very idea of lustration outright, and indeed acknowledges (without disapproval) that “some states have found it necessary to introduce administrative measures, such as lustration or decommunisation laws”,32 it does set
“DECOMMUNISATION” AND “LUSTRATION”
233
out a number of criteria that such laws must meet in order to be compatible with “a democratic state under the rule of law”, the most important being that lustration laws must be based on the principle of individual, rather than collective, guilt, which “must be proven in each individual case”.33 This would undermine much of the lustration as contemplated (and practised) in CEE because the general rule usually is that the very fact of having occupied certain positions in the apparatus of the Communist regime is sufficient to disqualify the individual in question from certain public offices. This may seem a pedantic distinction, because the performance of some roles (for example, in the security apparatus, or as a collaborator with those services) almost inevitably meant participation in the consolidation of the authoritarian system. But this goes to the very heart of the debate about the principle of lustration: putting obvious crimes to one side (as it is not the purpose of lustration to deal with these), can a newly democratic state protect itself by removing a whole category of people from performing certain roles? By rejecting the collective approach, the resolution of the Parliamentary Assembly seems to be going further than it claims to go, and can be read as delegitimising the very principle of lustration. Further, the resolution (and even more emphatically, the working document accompanying this resolution) makes it clear that lustration may only be used for purposes related to the protection of the democratisation process, and not for “punishment, retribution or revenge”.34 This serves to emphasise even further the individual as opposed to the collective approach, as there is an inherent tension (as suggested above) between the utilitarian purpose of lustration and the collective identification of the persons to be removed from public offices. Such a collective approach to lustration serves well the purpose of emphatic retribution for the sins of the past, but it is exactly this that the resolution (and most of the official rhetoric concerning lustration in CEE) specifically prohibits. The second remark concerns the example of the former German Democratic Republic (GDR), and its effect on other CEE countries. The agreement on the reunification of Germany of 31 August 1990 provided an opportunity to dismiss from the civil service the former senior party functionaries of the GDR who had held public offices, as well as the leaders of trade unions and those who had cooperated with the Stasi (the East German secret police). This was followed by a law on 20 December 1991 concerning the files of the former GDR security police, according to which a wide range of officials were checked for possible links with those security services. As a result, thousands of public officials, judges, university professors and schoolteachers, among others, were discharged. Germany, however, was for one readily evident reason not a suitable model for the rest of the post-Communist of CEE states to follow, having been absorbed into the former West Germany. Part of the reason why the “decommunisation” in the former GDR was so efficient and carried out with such zeal was that West Germany had the accusations of incomplete denazification still fresh in its collective memory, and that shaped and hardened its resolve to cleanse the state more thoroughly this time. The Germans had the unique “opportunity” to apply the tools of external screening for past sins, rather like the Allied powers in post-War Germany or the Americans in post-War Japan: put simply, German decommunisation and lustration was made easy by the disjunction between those who were implicated and those who did the screening. However, in
234
CHAPTER 9
other CEE states this was not the case: many of those in charge of (or only in favour of) decommunisation had been implicated in the past; the division between the tainted and the clean was blurred, consistently with the fact that the totalitarian systems left hardly anyone completely immune from responsibility for the systemic evil that characterised them. And, of course, the more diffuse the responsibility for, and complicity in, the repugnant system, the more difficult it is to come to terms with the past. This is not to say that responsibility for past wrongs was equal for all citizens of post-Communist states, nor to deny the moral righteousness of the dissidents, but rather to show that the “luxury” of having a class of uncontaminated citizens exercising moral scrutiny of those who were putatively implicated in the authoritarian past never existed in CEE. This fact undoubtedly importantly affected the erratic and incomplete moves towards lustration and decommunisation – in contrast to the German experience, where the circumstances existed for an exercise of classic “victor’s justice”.35 In contrast, in all other post-Communist societies, “transitional justice” was of much purer in character – in the sense that “the society is in a real sense judging itself”, which in turn meant, crucially, that “[i]n transitional justice, many of the judges themselves have been implicated in the regime they are judging”.36 Nevertheless, the example of the ex-GDR did exert some pressure upon lustration proposals elsewhere, and provided a model for the strongest proponents of lustration and decommunisation in other CEE states.37 2. LUSTRATION AND DECOMMUNISATION IN CENTRAL AND EASTERN EUROPE “Lustration”, or vetting, may be considered either harsh or lenient as a function of three particular characteristics: (1) the range of positions in the old Communist regime (or types of involvement with the regime) that trigger lustration-related consequences, (2) the range of current positions the holders of which are subject to lustration, and (3) the consequences that follow from the finding that a person belonging to a range (2) has been implicated in the circumstances defined by range (1). As will become clear, there has been a considerable spectrum of approaches to “lustration” and “decommunisation” in CEE countries in terms of all three characteristics.38 While no single factor explains these differences, three variables seem to have been most important in determining the severity and seriousness of the means applied to “settle the accounts”. First, the degree of repressiveness of the previous regime: the harsher and the more pervasive it was, the more robust the decommunisation. Second, the way in which the collapse of the old regime occurred: in negotiated transitions, ex-officials of the Communist regime played an important role as interlocutors and collaborators with former dissidents, and it was both politically impossible and socially awkward to propose measures aimed at disabling them from performing official political roles in the new system. In addition, an act of forgiveness could be seen as a form of reward for good behaviour in the transition period. As one commentator observed: “When the regime and the opposition negotiate the conditions for holding free elections, the regime demands treatment with a velvet glove”.39 Third, the political strength and influence of parties
“DECOMMUNISATION” AND “LUSTRATION”
235
connected to the old regime in the first years after the transition has been significant: when participating in governing coalitions (or even in opposition, but with a high degree of influence), they could – and did – use their clout to prevent any strong and effective lustration measures. But note that, ironically, the relative strength of postCommunist parties well after the transition period may have the effect of keeping the issue of lustration salient, because it is an effective tool of political struggle for their opponents. That was precisely what happened in Poland where, as one observer notes, “[i]n a political climate in which former communists were able to resurrect themselves successfully, the issue of ‘dealing with the past’ became increasingly salient and was instrumentalised by the Polish right as part of the political power struggle”.40 All three of these factors are, naturally, to some degree inter-related. For instance, more repressive regimes, as in the case of the former Czechoslovakia, ended in a less negotiated way than the more liberal ones, such as Hungary or Poland; hence the connection between the first and the second factors. Furthermore, negotiated transitions usually allowed ex-communists to retain a number of key positions and influence over legislation, hence the connection between the second and the third factors. None of these factors can account fully for the differences in approaches to lustration, for example, those between the Czech and Slovak systems, which had, after all, shared the same legacy of repression and collapse of the communist regime, and yet which – due to different domestic political constellations after the split of Czechoslovakia – adopted diametrically opposed practices. A fourth, secondary factor (secondary because it has been largely affected by the three just mentioned) has to do with the timing of the consideration of the issue of lustration: when, due to political circumstances, it has been effectively placed on the political agenda relatively late (as in Poland, where the first successful lustration law was enacted only in 1997), the measures adopted were weaker partly because of the growing unreliability of the files on which lustration determinations could be made, and partly due to the fact that public emotions had become, with time, less pronounced. 2.1
The Radical Model: The Czech Republic
It is in the Czech Republic that the strongest position on decommunisation and lustration has been adopted. Not only do the laws (confirmed, as we shall see below, by the Constitutional Court) cover a relatively wide group of persons, but they also ban such persons from a fairly wide range of posts. Therefore, in terms of all three criteria suggested above, the Czech approach to lustration can be seen as the harshest. This, undoubtedly, can be explained mainly by the combination of two factors: the particularly severe and pervasive nature of the police state of Czechoslovakia in the years after the Prague Spring until late 1989, and the weakness of Communist or post-Communist political forces in the first months and years after the transition. Secret police infiltration in Czechoslovakia was extremely broad: police “tried to infiltrate every school, office and institution, particularly those employing
236
CHAPTER 9
intellectuals or involving any contact with foreigners”.41 Those suspected of having “anti-state” views (or, in academia, of espousing non-Marxist theories) were subjected to a very wide range of discriminatory practices and harassment, ranging from arbitrary arrest and the loss of employment to sanctions such as of university places to the children of the accused. The very form of transition (a matter related, of course, to the nature of the pre-transition regime) is of relevance here: in contrast to the negotiated transitions in Poland or Hungary, where the spirit of reconciliation and collaboration of all parties prevailed (at least in public rhetoric), in Czechoslovakia the changes occurred in a more revolutionary (though non-violent) way. This enabled the adoption of a “new beginning” approach, further strengthened by the dissolution of the federation, which reduced any sense of legal continuity, so powerful in Poland or Hungary. These differences explain why the approach in the Czech Republic to “dealing with the past”, not just in the area of lustration but also with regard to property restitution or suspension of statutes of limitations for politically motivated crimes, was markedly different than elsewhere in the region. The Czechoslovak lustration law of 4 October 1991 decided that those who had collaborated with the security services of the Communist regime, or had held certain positions within the regime (a category that has a very wide reach and covers even such people as secretaries of the Party at district level), could not now hold high positions in governmental bodies and organisations for five years.42 Even relatively low-status former functionaries of the Party, from the township level up, were barred from holding certain positions, regardless of whether they had collaborated with the state security agencies or if they had been involved in criminal activities.43 The offices from which those persons were barred included positions in state institutions and also key positions in the military, the judiciary, universities, state-run media and many other state-run enterprises (including joint-stock companies where the state was the main shareholder). Each employee or prospective employee within the range of categories covered by lustration was required to ask for a certificate of “negative lustration” that would be submitted to the relevant employer, or in the case of elected officials, to the parliament. (In the years that followed, there were 310 000 requests for lustration vetting, and of these, 15 000 were “positive”, i.e. the persons in question were pronounced collaborators; most of these, however, did not result in the loss of employment status because those thus vetted did not hold high-level offices). Although the law created a right for those found to be former collaborators to challenge the finding before a special commission and then to appeal commission’s finding to a court, this right of appeal was denied to those found to be higher-level collaborators, and to the former top officials of the Communist Party. It is clear that the legislators have not changed their view on this matter in subsequent years, and that the Czech Republic has not taken a different view to that of Czechoslovakia. (In contrast, even though in Slovakia the law was never formally repealed, contrary to the express wish of the first Slovak government of Vladimir Meþiar, it became dormant and without any de facto force. The Constitutional Court of Slovakia has flatly refused to strike down the lustration law, on the basis that the matter had already been considered by the Constitutional Court of Czechoslovakia). Originally, the 1991 law allowed for this lustration process to continue until the end of 1996, but, in September 1995, this was amended to extend the period to the end of
“DECOMMUNISATION” AND “LUSTRATION”
237
2000. Interestingly, this extension was done by a parliamentary vote to overturn the veto by President Vaclav Havel, who argued that retention of the law would suggest that the country was still in a revolutionary phase, and that a normal “rule of lawbased” system had not been yet created.44 In January 1999, the Parliament rejected a bill to annul the law and, in 2000, the law was amended so as to remove the cut-off period (thus leaving it open indefinitely).45 Much earlier, at the beginning of the independent existence of Czech Republic, some additional lustration rules were added (in the so-called “small lustration law”)46 to the 1991 act (which became known as the “large lustration law”). Moreover, a law was passed in 2001 that expanded the application of this lustration law to public higher education institutions.47 The initial 1991 law was challenged before the Constitutional Court of the Czech and Slovak Federal Republic; in fact, it was easily the most important case considered by that Court before the “divorce” of the two constituent entities of the Republic, and the subsequent substitution of the Court with two successor constitutional courts for the new states.48 The Court found the lustration law to be generally constitutional: it justified its decision on the grounds of security and democracy, noting the extraordinary nature of transition periods, in that during them the institutions of government are much more susceptible to a relapse towards the totalitarian system. It did, however, find one provision to be unconstitutional – one that included those who were merely candidates for clandestine collaboration in the list of those subject to the law. The Court found this to be contrary to Art. 4(3) of the Charter on Fundamental Rights and Human Freedoms, which provides that all statutory limitations on freedoms must apply in the same way to everyone. In its reasoning, the Constitutional Court came down heavily on the side of material rather than formal justice, arguing that, until 1989, all important positions in the administration were filled by those conforming to cadre orders that required them to feel a strong sense of responsibility towards the Party. In addition, the Court felt that the regime’s most important tool for the maintenance of power, the State Security organisation, would try to influence or reverse the present democratic developments in the country (the Court cited some State Security directives of 1989, which stated that they should attempt to infiltrate the new power structures and obtain positions of influence). This, in conjunction with the fact that about ninety percent of the Security Service’s files were destroyed or hidden, “created a real and potentially very perilous source of destabilization and danger, which could easily threaten the developing constitutional order”.49 The Court went on to say that a democratic state not only has the right to eliminate an unjustified privilege once enjoyed by a certain group of citizens, but it also has “the duty to assert and protect the principles upon which it is founded, thus, it may not be inactive in respect to a situation in which the top positions at all levels of state administration, economic management, and so on, were filled in accordance with the now unacceptable criteria of a totalitarian system”.50 It is necessary for employees of the state to be loyal to democratic principles. Thus, the basic justifications for lustration are stated as being “[the state’s] own safety, the safety of its citizens and, most of all, further democratic developments…”.51 The analogy drawn was to the security clearance checks that are operative in various consolidated
238
CHAPTER 9
democracies in appointment practices; the vetting of candidates for their past practices is a rational and necessary way of avoiding unnecessary risk in staffing practices. The Court did not, however, endeavour to explain what risk is created by appointing an individual with a discreditable past to a position that does not involve any particular opportunity to trample upon the civil or political rights of others, or to take security-related decisions, such as in universities or in the media. In other words, the extensive range of positions to which lustration applied somewhat undermined the security rationale provided for the law. In contrast to the sweeping range of the positions covered by the Czechoslovak (as it was then) lustration law, security clearances are tolerated in Western democracies only with respect to a narrow range of top positions; to apply politically-sensitive criteria regarding a person’s past as part of the qualifications for employment is, in most cases, anathema to a liberal democratic state.52 Therefore, the only way in which the Court could justify the law was by an appeal to special transitional circumstances (which in fact it did); then, however, the analogy to the screening practices elsewhere in democratic states rings rather hollow. As to the intimation of collective guilt implied by the law, the Court justified the banning of whole groups of persons by saying that such practices can be found in other countries, giving the example of the Federal Republic of Germany after the absorption of the former GDR. It compared the situation under this law to that subsisting under the Communist regime, and stated that this law was to affect a small number of people and be in force only for a short period of time (until 1996, as originally intended). The Constitutional Court then proceeded to examine possible conflicts of the lustration law with the Charter of Fundamental Rights and Human Freedoms, and with international agreements such as the International Convention on Civil and Political Rights. It held that the law in question was not discriminatory, as it merely provided extra preconditions for posts, preconditions that were deemed to be crucial to the fulfilment of the relevant functions. As to the right to work and to freely choose employment, the Court again counteracted this with the argument that the state has the right to protect the democratic constitutional system and the its own security; the “right to enthrone democratic leadership and to apply such legal measures as are apt to avert the risk of subversion or of a possible relapse into totalitarianism…”.53 Finally, responding to arguments about legal certainty, it stated: “Respect for continuity with the old value system would not be a guarantee of legal certainty but, on the contrary, by calling into question the values of the new system, legal certainty would be threatened in society and eventually the citizens’ faith in the credibility of the democratic system would be shaken”.54 It is at this point that the Court’s appeal to substantive norms of justice – to “natural human rights”, as the Court put it55 – that prevails over purely procedural notions of legality becomes most visible. The discourse becomes highly moralistic, and the appeal is to the substance of democratic norms as contrasted with a notion of the state that would be “amorphous with regard to values” and that would be concerned with “the observance of any sort of values and any sort of rights” provided they were “adopted in the procedurally proper manner”.56 Throughout, however, the rationale provided is consequentiality and pragmatic: the barring of ex-Communist officials and collabo-
“DECOMMUNISATION” AND “LUSTRATION”
239
rators is presented as a necessity dictated by the requirements of a democratic society rather than as an act of historic justice, or “some sort of retribution” as mentioned in the article by Cepl quoted earlier. Indeed, the Court carefully distinguished between the alleged ex-post facto unlawfulness of the holding of certain offices (which the law, in Court’s opinion, did not establish) and, on the other hand, the setting of certain conditions for holding some prominent positions in the future; this distinction served to rebut the charge of retroactivity. Later commentaries often emphasised the fact that, in the case just discussed, the Constitutional Court was referring to the transition period from totalitarianism to democracy, and found that special protections were needed at this time. For this reason, it is interesting to note that a second decision has been handed down on this law, this time by the Czech Constitutional Court, that is, the successor court to the Constitutional Court of the former federation.57 It related ostensibly to the 2000 amendments, which removed the time restrictions for the validity of lustration provisions: the challenge was based on the argument that the main rationale for the earlier law, namely to counter “the risk of subversion or a possible return of totalitarianism”58, was no longer valid after the consolidation of democracy, and that there was no reason to extend what had been transitional, extraordinary measures par excellence. However, the Czech Constitutional Court held that amendments cannot be challenged by themselves, and thus that the whole law had to be reevaluated. Overall, the Court declared itself in agreement with the earlier decision on this matter, and much of its decision is simply a summary, with approval, of the 1992 judgment. It admitted that part of the earlier decision was based on the extraordinary nature of the period of transition, and the intention to have the law in force only for a short time. However, it held that these were not central, crucial justifications, and that the Court had also based its decision on the fact that, in all democratic systems, the state should be able to require of its officials loyalty to democratic principles, and lay down certain prerequisites in order to protect the security of its citizens and further democratic development. In particular, the Court refused to consider whether the social and political conditions as regards the maturity of democracy were satisfied (and hence, the completion of the transition period in which special lustration measures were justified), deeming it to be a “social and political question, not a constitutional law question”.59 The Court thus held that the need for loyalty does not exist during a transition period only. In addition, an important component of such loyalty is that not only should the public services actually be loyal, but that they should also appear as such to the general public; any lack of this is a danger to democracy. The Court referred, with approval, to the 1993 Law on the Lawlessness of the Communist Regime and Resistance to It (see below), which stated that all “functionaries, organisers and instigators” of the old regime have joint, collective responsibility for its actions.60 It went on to say that “[t]hus, it is evident that an individual’s close connection to the pre-November regime and its repressive components is a circumstance capable of having an adverse effect on the trustworthiness of a public position which that individual holds in a democratic state”.61 Therefore, bearing in mind the legitimate aim of the law and the fact that it applies to a restricted number of jobs, the Court found the law to be constitutional. The Court did note that in other countries of
240
CHAPTER 9
Central and Eastern Europe such laws are limited in time; nevertheless, at the time of that particular decision, “lustration laws [in other European states] are still valid and are in effect”.62 However, the Court stated that the legislator should create a new law governing entry into the civil service, which would recognise the decreasing relevance of positions held in the Communist era, and would supersede the lustration law. 2.2
The Intermediate Model: Albania and the Baltic States
The Czech position represents one extreme pole in the range of approaches to lustration in CEE, when considered from the point of view of the range of positions screened and the consequences for the “positively lustrated” (in the sense of confirmation of a person’s past involvement with the totalitarian regime). The Albanian lustration laws came close to this position: in autumn 1995, two decommunisation laws were passed that would deny ex-communists the right to run for the parliament and also work in the judicial system, state banks and other financial institutions, the army and the mass media.63 In the case of candidates for elections, the law established a special commission to determine whether a prospective candidate had served in one of a variety of capacities under the prior regime or was registered in the secret police files as a collaborator. The Constitutional Court upheld much of this law: it only struck down the provisions concerning the lustration of journalists working for independent newspapers. Earlier, a law of January 1993 had stated that both former members and employees of the Communist Party, and the collaborators with security services, could not practice as (private) lawyers for five years. That law was however overturned by the Constitutional Court in May 1993 on the basis of a variety of constitutional reasons, including the constitutional principle that lawyers constitute a free profession (Art. 16), the constitutional right to work, the principle of separation of powers (the law granted to a licensing commission powers that are normally reserved for the court) and the principle of presumption of innocence.64 All three Baltic States have also introduced bans on former operatives of the KGB and those who collaborated with them from holding positions in the parliament or government. The case of the Baltic States was somewhat special in comparison with other CEE countries, due to the fact that the secret service operatives there were more directly related to the annexation of their territories by the USSR and both during and immediately after the restoration of independence, they posed direct threats to the sovereignty of these states. For example, the Lithuanian government adopted a series of resolutions in 1990 and 1991 warning against the activities of operatives of the USSR Committee of State Security, calling upon them to register with the Lithuanian Ministry of the Interior, and demanding that they break all links with the secret services of the USSR. But it was only in 1998 that Lithuania introduced the lustration law that banned former KGB employees from being employed in almost all state institutions as well as in a number of types of private institutions (banks and credit unions, security services, communication enterprises, as private lawyers, notaries, etc) for ten years.65 At the same time, the law contained
“DECOMMUNISATION” AND “LUSTRATION”
241
a provision allowing the suspension of this lustration measure in respect of some exKGB employees under certain conditions: in particular, if they revealed all of the information pertaining their former links with and activities for the KGB. For such cases, the law set out a procedure whereby the Centre for Research into the Genocide of the People and Resistance of Lithuania and the State Security Department would jointly adopt a measure recommending the suspension of lustration measures against a particular person. This recommendation would then be considered, and a decision taken, by a three-person commission appointed by the President of the Republic; its decisions would be subject to confirmation by the President himself. This law was challenged by a group of MPs before the Constitutional Court, which upheld the lustration in general but struck down the exemption provisions on the basis that it empowered the President to make decisions regarding restrictions of constitutional rights – something that, under the Constitution, only the Parliament was empowered to do.66 When it came, however, to the main thrust of the statute, the Court found no defects in the lustration procedure. In particular, it rejected one of the arguments of the petitioners, who had claimed that lustration allowed for criminal sanctions to be meted out by non-judicial bodies, and also that it denied the presumption of innocence that is normally adopted in criminal proceedings. It also rejected the argument that exclusion from certain positions violates the constitutional right to free choice of occupation. The Court reasoned that the right to equal opportunity to work in a state office is not absolute, and that “the requirement for loyalty and credibility in connection with service in a state office is common and understandable”.67 The Court then extended the same argument to positions in private enterprises – such as banks, credit unions, security services – and also occupations such as private lawyers or notaries, making the somewhat sweeping statement that “In private enterprise there are also such areas and posts which are of crucial importance to the society and the State, therefore the State may set special requirements to those who want to work in such jobs”.68 As one can see, the Court expressed an extremely high level of deference to the legislature, thus largely giving it carte blanche to determine any position in society, public or private, as being of sufficiently high importance to warrant the exclusion of ex-secret service employees. At this point, one may easily begin to suspect that a rule ostensibly defended in a consequentiality way – as a rational protection of important governmental purposes – is in fact driven by a desire for retribution (how big a danger to the State does an ex-KGB operative pose when working as a notary?). Nonetheless, it must be emphasised that the Baltic-style lustration did not go as far as the Czech one, as it applied only to ex-operatives of the security services of an occupying empire, and not to ex-Party or state officials. 2.3
The Lenient Model: Poland, Hungary and Bulgaria
The opposite pole, at the lenient end of the range of lustration policies, is occupied by those systems, such as Poland, Hungary and Bulgaria, in which the ostensible object of lustration is to verify and reveal past of public officials (or of
242
CHAPTER 9
candidates for public positions) without, however, banning the those “positively lustrated” (i.e., certified as connected in the past to the security services) from the offices in question. It is significant that in all three of these countries the Communist parties (or the reformist factions within them) took an active and often positive role in the transition (as in Poland and Hungary), or occupied a strong position of power soon after the transition.69 Both in Poland and Hungary, the terms of transition were negotiated in round-table discussions between Communist Party officials and the opposition, and the loyalty of Communist Party activists (however renamed and reformed) to the negotiated rules was a central factor in the peaceful and eventually successful move towards democracy. The weak character of the retribution meted out by the lustration measures in the years after the transition (as well as the fact that they came relatively late) can therefore be seen as functionally linked to the role of the ex-Communists in the transition itself, and also to their influence upon legislation in the early post-transition period. Also, in all three countries the role of the Constitutional Courts in shaping the lustration laws, often in ways contrary to the wishes of the parliamentary majority of the time, has been quite significant. In Hungary, the legislature passed a fairly mild lustration law in 1994, which provided for the creation of three-judge panels that would examine whether certain present public office holders (the President, ministers, high officials, members of parliament, judges, certain journalists and persons with high positions in state universities and public companies) had collaborated with the state security services or been members of the fascist Arrow Cross Party. Even if a person were found to be “guilty”, the information would only be made public if they refused to resign from their post. They could still keep their job even if such information was publicised. This was therefore a compromise solution between discharge from the post in question and no lustration at all; it also contained the possibility of no public disclosure, but on condition of resignation. The Constitutional Court found the 1994 law to be partly unconstitutional and offered some advice on the necessary changes to the content of the law, and a time limit (July 1996) for its enactment.70 The important thing about this decision is that the Court declared that one of the two main functions of lustration (alongside its moral and symbolic function of exposing the past), namely to protect the process of democratic transition from being undermined by those who had compromised themselves in the past, was no longer relevant: the transition had actually occurred and its protection could not be now used as a rationale for lustration.71 In this respect, the Hungarian Court’s reasoning departs radically from that of its Czech counterpart, which only much later chose to rely on the same argument to maintain the validity of its lustration law. (This had something to do with the lateness of the Hungarian law, which came fives years after the transition began, while the Czech Court relied heavily on its Czechoslovak predecessor’s decision, which had reviewed the law very soon after the Velvet Revolution). In any event, the Hungarian Court applied the standards of a normal democratic society rather than the extraordinary circumstances of a transition period. Under such standards, the Court explained that there is a necessary balancing act between, on one hand, the right of informational self-determination (protection of personal data) in Art 59 of the Constitution and, on the other, the right to acquire and disseminate information
“DECOMMUNISATION” AND “LUSTRATION”
243
of public interest, protected under Art. 61 of the Constitution. It held that public persons do have a smaller sphere of privacy than private persons, and thus it would be just to come down in favour of the principle of freedom of expression/acquisition of information. Thus, such “pure” lustration is constitutionally permissible for persons holding public office. However, it saw a problem in this law, in that those persons who had been spied on during the old regime did not have their “rights of informational self-determination” properly secured. Also, those subject to background checks were chosen too arbitrarily. For instance, the list included all journalists in the public (but none in the private) media. This, the court held, was without justification, arguing that the law should either apply to all journalists who have a role in shaping public opinion (including those working for private media), or to none of them. In addition, the Court felt that persons with high positions in state universities or businesses do not participate in public affairs, and thus it was not reasonable to include them in the lustration law. Finally, the Court found an unconstitutional omission in the Parliament's failure to properly secure the rights of privacy and informational self-determination of all citizens by providing for the right of people to check their own files. After the Court’s decision had been handed down, the Parliament (now with a new majority, this time composed of the Socialist party, which had grown out of the reform wing of the old Communist Party) tried to comply with these suggestions by enacting, in 1996, a new – and greatly weakened – lustration law.72 This reduced the number of people subject to the law (to persons who must take an oath before the parliament or the President, and those elected by the parliament). It also narrowed the concept of collaboration with the state security apparatus to include only those who belonged to the state security organisation, actually submitted reports on others, informed on others, or were paid by the organisation. The mere presence of someone’s signature upon a declaration of willingness to spy was no longer to be sufficient. The Hungarian position is therefore moderate: the established fact of earlier collaboration with the security apparatus is not, per se, a sufficient ground for the dismissal or screening out of a person; and, if a person resigns from their post, the fact of their collaboration need not even be publicised. (As it happens, there have not been, thus far, any cases of individuals resigning due to past activities).73 And it is only collaboration with the security apparatus of the totalitarian regime, not any other involvement in the politics of the Communist era, which forms the subjectmatter of the background check. The current Polish position is quite similar in that it applies only to collaboration with the secret police, and “positive lustration” does not necessarily result in the loss of one's position, or ineligibility for such a position. However, the model adopted is somewhat different from the Hungarian one in that it requires a number of positionholders and candidates to declare publicly whether they had been collaborators, and it is only in the case of a false declaration, established as such by the courts (the socalled “lustration lie”) that a person loses the right to hold public office, or to run in elections, for a period of ten years. That penalty is, strictly speaking, for lying about one’s past, and not for the fact of one’s prior collaboration with the secret police in itself. The range of positions to which it applies is relatively broad: it includes all
244
CHAPTER 9
elected state officials from the President downwards (including parliamentary candidates), all ministers, state functionaries above the rank of deputy provincial governor, judges, prosecutors, barristers (on the basis of the 1998 amendment to the law), and leading figures in the public media. As a result of the law of 11 April 1997, around 22,000 public officials were obliged to disclose all forms of voluntary, secret collaboration with the security service between 1944 and 1990. This system was preceded by a number of failed attempts to adopt a more radical, Czech-style lustration process in Poland. In particular, in May 1992, the Parliament passed a resolution stating that the Minister of Internal Affairs must provide information as to whether any person applying for, or holding, a position in the parliament, as a government official (with an extremely wide range of positions subject to this requirement, from local council officials upwards), or as a judge, prosecutor or barrister (advocate), was connected to the secret police under the past regime.74 This led to a major political crisis (as the Minister of Interior undertook the task of providing the parliament with the names of those whose files were in the archives of the Ministry with great zeal), and to the eventual dismissal of the Government by the parliament at the initiative of the then President WaáĊsa (whose name, incidentally, was on the list of “top agents” produced by the Minister). The resolution in question was later found unconstitutional by the Constitutional Tribunal, in its decision of 19 June of the same year. The Tribunal found that the resolution violated individual dignity, without providing any means of protection to those thus “lustrated”: it led “to the violation of the good name of the persons to whom the information [demanded by the Parliament from the Minister of Internal Affairs] applies and creates a sui generis penalty of infamy”.75 According to the Tribunal, this was inconsistent with art. 1 of the so-called "Little Constitution" (in force at the time), which contained a formula for the creation of a democratic Rechtsstaat. The following year, in May 1993, a law was passed that demanded that all candidates for elections to the parliament swear that they had never worked for the security services. However, the Constitutional Tribunal, when asked by the Chief Justice of the Supreme Court to provide a “universally binding interpretation” on the matter, held that a false declaration by a candidate would not represent sufficient grounds for setting aside the election – a decision that removed the teeth from the provision in question.76 A law of narrower scope (because it only applied to judges), but with harsher penalties, was passed in December 1997. It amended the 1985 Act on the Courts of General Jurisdiction, so as to allow the removal of judges who had collaborated with the previous regime, or prevent them from receiving pay rises: the law applied to judges who, in the years 1944-1989, had committed acts contrary to the principle of judicial independence by issuing unjust sentences, or judges who had otherwise impaired a defendants’ right to a fair trial. The central article of the law – article 6, which provided that statutes of limitation would not apply until 31 December 2000 in the case of disciplinary proceedings against judges – was found defective by the Constitutional Tribunal on procedural grounds.77 According to the Constitution, the National Council of the Judiciary (NCJ) has the right to express opinions on all matters related to bills dealing with the judiciary, including, naturally, this one.
“DECOMMUNISATION” AND “LUSTRATION”
245
However, art. 6 was introduced as an amendment to the bill in question and went beyond the scope of the initial legislative proposal. Because it occurred at a late stage of the legislative process, the NCJ was not consulted. This was a controversial decision, as indicated by two strongly worded dissenting opinions by Justices Rymarz and Zdyb, who denied that there was a procedural defect, claiming that there was no firm obligation on the Parliament to consult with the NCJ.78 In particular, one of the dissenters, Justice Zdyb, warned against elevating procedures to such a high level that they will be able to justify any evil and injustice. It is important to note that the Court emphasised the narrow, procedural basis for its decision, and made it clear that professing an opinion on the substance of the law would be inconsistent with the principle of judicial restraint.79 And, given that the procedural grounds were quite controversial, the Tribunal effectively avoided facing the issue of lustration openly. It was, however, unable to avoid doing so a few months later, when it had to conduct a review of the main lustration law of 11 April 1997 (described above), this time in an ex-post review process initiated by a group of deputies belonging to the post-Communist Democratic League Alliance.80 The law, as explained above, required high-level public officials to make public declarations concerning any collaboration with the secret police under Communism, and penalised those who made false declarations with a ban from public office for ten years. Those challenging the law claimed that it required public officials to incriminate themselves, and also that it enabled the resumption of proceedings against a person who has already been cleared of charges. (There were many other grounds for the challenge, including the rather bizarre one that the lustration law violated the constitutional duty to defend the Motherland!) On the main points, however, the Tribunal held the law to be constitutional: it found that the statute did not require individuals to incriminate themselves because those who admitted their prior collaboration were not eo ipso barred from office; only those who made the false statements were. (This argument was rejected by one of the dissenting judges, Zdzisáaw Czeszejko-Sochacki who suggested that the requirement not only violated the right against self-incrimination but also that of privacy and “informational autonomy”).81 On the other hand, the Tribunal found inadmissible the provision allowing for the resumption of proceedings should new facts about the lustrated person’s past become known: the Tribunal decided that it introduced “a state of permanent uncertainty on the part of a lustrated person … and therefore unconstitutionally violated his/her liberty”.82 At the very least, the Tribunal suggested, there a time limit must be set for the possibility of resumption. One important aspect of this decision was the binding interpretation of the notion of “collaboration” with the security forces of the Communist state: the challengers alleged that the definition of “deliberate and secret collaboration” was unduly vague. The Tribunal provided a definition of collaboration by making it clear that a simple commitment to collaborate, even if evidenced by a person’s signature, is not sufficient: what matters is whether this was followed by specific actions that constituted collaboration, such as “operational gathering of information” and conveying this to agents of the secret police.83
246
CHAPTER 9
This was not the end of the Constitutional Tribunal’s encounters with the Polish lustration law. In June 2000 it issued a decision in the course of a “concrete review”: a judge of the “lustration court” (one of the branches of the appellate court in Warsaw) asked the Tribunal two questions related to the case before him.84 The first question concerned the November 1998 amendment to the law of April 1997, on the basis of which the lustration procedure was still to be carried out in respect of a public official who had resigned from office (before the amendment came into force) in order to avoid lustration. The Tribunal decided that continuing (or resuming) the lustration process in respect of those who had already resigned from office (but only before the amendment entered into force, that is, in the period in which the resignation resulted in the abandonment of the lustration proceedings) was unconstitutional because it violated the principle of the citizens’ trust in the state and in the law, which could be derived from the constitutional principle of the democratic Rechtsstaat. The violation was due to the fact that even those who resigned from their offices on the basis of rules enacted in April 1997 in order to avoid the lustration procedure could be called before the lustration court on the basis of an amendment made in November 1998 t: this “suggests that the lawmaker does not observe the principle of loyal behaviour towards citizens”.85 The second question was weightier: it concerned the “presumption of innocence” in the lustration proceedings. This went to the very heart of the lustration process as established in 1997: as a hybrid process combining characteristics of criminal and administrative procedures, it contained elements that would not be acceptable in fully fledged criminal proceedings. One such element was the rule that, if the lustration court found that there was insufficient evidence of a “lustration lie”, it was supposed to discontinue the proceedings. This, according to the critics of the procedure (and reflected in the question put to the Tribunal) suggested that the presumption of innocence was not respected in the lustration proceedings. Under such a presumption, all doubts should be interpreted to the benefit of the “defendant”, and so insufficient evidence should result in a decision by the lustration court that the declaration of the individual in question was truthful. However, proponents of the existing system argued that the lustration proceedings were not the same as criminal ones, and therefore the rules of criminal procedure in general only apply to it insofar as certain points were not dealt with specifically by the law on lustration. Put simply, the question boiled down to whether the lustration procedure was a form of criminal procedure and was therefore governed by the general rules of fair trial. In June 2000, the Tribunal used a strategy of avoidance with regard to this fundamental question: it simply found that to answer it was not necessary in order to consider the specific case before the lustration judge who had initiated the concrete constitutional review. As the lustration proceedings had to be dropped anyway, in consequence of the answer to the first question, the second was considered moot. This was a golden – and missed – opportunity for the Constitutional Tribunal to fine-tune the shape of the lustration process, by reducing the negative impact of an inconclusive finding by a lustration court on the reputation of the individual in question. One of the worrying aspects of the scheme of lustration adopted in Poland was the way it combined a relatively low level of legal punishment (no criminal sanctions; no sanctions at all if a person makes a truthful declaration) with a high
“DECOMMUNISATION” AND “LUSTRATION”
247
level of social opprobrium for those called upon to defend the truthfulness of their declarations. The “discontinuation” of proceedings for lack of sufficient evidence is hurtful to those to whom it applies: it does not vindicate their position, and permanent doubt remains over their reputation. There have been some legislative attempts to alter this scheme (by compelling the court to “acquit” a defendant in lustration proceedings when the evidence is insufficient) but they have thus far been unsuccessful.86 Most recently, the Constitutional Tribunal considered the issue of lustration in June 2002, when it found an amendment to the 1997 law unconstitutional on purely procedural grounds.87 The amendment limited the scope of application of the lustration law by removing from its ambit former collaborators with military intelligence and counter-intelligence. Further, the amendment exempted from the list of objectionable collaborations those actions that “did not endanger personal and civil rights and freedoms” – an arguably highly imprecise clause that, according to the critics, would render lustration totally ineffective. The amendment was pushed through the legislative process by the post-Communist Democratic Left Alliance (DLA) in November 2002, when it was a major partner in the governing coalition. However, the central part of the amendment (the exclusion just described) was added by the higher chamber (the Senate) to the proposal already approved by the lower chamber (the Sejm); it was only after the Senate proposal returned to the lower chamber that the DLA managed to persuade one of the formerly undecided parties (Samoobrona) to support the bill and thus mobilise a sufficient parliamentary majority. The Tribunal found the process defective because, under the rules for legislative procedure, an amendment suggested by the upper house cannot extend the scope of the Sejm’s original proposals; the Senate had therefore overstepped its powers. In the opinion of the Tribunal, by changing the definition of objectionable collaboration, “the Senate has in fact exercised the function of legislative initiative, which it is not constitutionally empowered to do”.88 Once again, the Tribunal used a procedural gambit to avoid facing the substantive issue. This would be a laudable approach were it not for the fact that the decision as to whether the Senate had in fact overstepped the limits of the Sejm’s bill was in itself controversial, and highly debatable – a point raised by some of the dissenting judges.89 They argued that the Senate is only prevented from throwing out a Sejm bill altogether and proposing its own, brand new bill in its place. The distinction between an “amendment” and a “new proposal” is, of course, largely a matter of judgement; and so basing a decision on such an important matter upon such a thin procedural ground seemed like a mere pretext for overturning a significant parliamentary statute, with the real bases for the decision remaining unstated. This point came up clearly in the dissenting opinions of those judges who criticised the amendment and would have invalidated it on substantive rather than procedural grounds. For example, President of the Tribunal Marek Safjan argued in his dissenting opinion that the amendment was enacted in a procedurally correct manner but was substantively faulty: for instance, the exemption of those ex-collaborators who had not endangered other people’s rights and freedoms would inject a highly subjective and vague criterion into the definition of collaboration, and this would lead to inequality of treatment of the accused in the lustration process.90
248
CHAPTER 9
A similarly moderate approach to “lustration” has prevailed in Bulgaria, despite contradictory moves at times towards, and at times against, a stronger system of “purifying” certain public domains of ex-Communists. In 1992, soon after the electoral defeat, in October 1991, of the post-Communist Bulgarian Socialist Party (BSP) and the rise to power of the liberal Union of Democratic Forces (UDF), the Bulgarian legislature passed a series of three laws applying, respectively, to the banking sector, to social-security entitlements, and to academia. The first “decommunisation” law would have established that officials of Communist Party organisations appointed within the last 15 years were barred from managerial positions in banks for 5 years. However, the Constitutional Court struck this down on the basis that it violated the right to work; the dissenting judges (the case was decided by a 7-4 majority) claiming that the law merely established some additional, legitimate professional qualifications for the jobs in question.91 The second law, passed in 1992, held that time spent working in Communist Party organisations was not to be included in pension calculations. This law was held by the Constitutional Court to contradict the constitutional guarantee of welfare rights (Art. 51 (1)).92 Finally, the third statute, the so-called “Panev law” (after its author Georgi Panev), disqualified former members of the Communist Party who had held certain posts in the old system and former instructors of Marxism-Leninism from holding leading positions in scientific and educational institutions. The law caused great controversy: in the opinion of one eminent lawyer and scholar of the older generation, and thus implicated in the highly ideological scholarship of the Communist era, “it had very negative consequences [because] many distinguished scholars withdrew [from academic activities]”.93 In fact, the law did not prevent those people from maintaining their teaching posts and from lecturing; they just could not occupy leading positions in their institutions. This law was found to be not unconstitutional94 by the Constitutional Court: the Court held that there was no violation of the right to work, as the lustration laws simply set out requirements of professional standards.95 It has been suggested that this change in approach (compared to the decisions invalidating the statutes on decommunisation in banking and pensions) was not due to any legal reasoning, but rather to the Court being influenced by the change in the political situation between the dates of these two cases. As Rumyana Kolarova notes, “While in the first case [concerning banking] the presumption of the justices was to soften the retributive anti-communist policy of the parliamentary majority, in the second case they had to decide against the backdrop of a newly formed and unstable parliamentary majority”.96 In any event, the “Panev Law” was in repealed by parliament in March 1995 (after a change in the composition of the legislature resulting from the election at the end of 1994, in which BSP won an absolute majority of seats).97 As from 1997, with a new UDF parliamentary majority (won in the elections of April 1997) the Bulgarian legislature started to pass wider-reaching (but, at first, quite mild) lustration laws. In July 1997, it passed the Law on Disclosure of Secret Police Files.98 This set up a supervisory body to examine the past of all members of parliament and government, the President, Constitutional Court judges and other judges and prosecutors. If they were found to be listed as agents in the card index of the former State Security apparatus, or had personal files as agents in the archives of
“DECOMMUNISATION” AND “LUSTRATION”
249
the Ministry of Interior, they could either resign, or the information would be made public. However, they could still keep their job if they were found to have collaborated. When the law came before the Constitutional Court, it was found to be partially constitutional.99 Thus, members of parliament and senior government officials could be checked in such a way, but not the President and the VicePresident of the Republic, or Constitutional Court judges. The Court grounded this decision on the principle of separation of powers, and not on the basis of particular rights provisions. The only serious attempt to establish a strong, Czech-style lustration system was made by the legislature in 1998, when it passed a far-reaching lustration law (the Administration Act) that would have prevented high-ranking Communist and secret service officials or collaborators from obtaining high government and civil service positions for 5 years. Those already in office were to provide a declaration, within 30 days, denying that they had held such a position or had acted as collaborators. Any admission of guilt was to lead to removal from office.100 However, in a decision of 21 January 1999, the Bulgarian Constitutional Court rejected this attempt by the legislature to expand lustration along Czech lines.101 The reasoning they gave for this was that it violated the rule of law, was discriminatory, and was contrary to international human rights principles. It was also held to violate the equality principle and the right to work (Arts. 6 and 48 of the Constitution).102 Thus, the current legal situation is that established by the 1997 law, with the only modifications thereto being those that have been introduced by the Constitutional Court itself: members of parliament and senior government officials (but not the President or Constitutional Court judges) will have any former collaboration made public if they do not admit to it. They can, however, keep their jobs. 3. RETROACTIVE EXTENSIONS OF STATUTES OF LIMITATION While lustration raises the question of the appropriate administrative and quasijudicial procedures for coping with the involvement of individuals in the repressive state apparatus of the past, it is really not a part of criminal law; those acts that represent the central concerns of lustration proceedings – collaboration with the secret services, involvement in the propaganda mechanisms of the oppressive regime and its ideology, performing leading official political and administrative roles, etc. – were not criminalised by law at the time of their commission. Lustration, however, does not exhaust the “unfinished business” of coping with Communism’s legacy. An even more difficult issue is how to deal with those acts related to the political oppression of the ancien régime that, formally speaking, were criminal at the time they were committed. If such cases can be re-opened and tried now, they face an evidentiary problem after the elapse of such a long period of time. However, as the example of Poland indicates, the trials of torturers in Stalinist prisons can meet the popular demand for retributive justice, while at the same time the symbolic, lenient punishments imposed make sure that the revenge exacted upon these now very old men is not too offensive to our notions of humanitarian treatment of the elderly. A more difficult issue – both legally and morally – arises when the criminal acts of the
250
CHAPTER 9
past fall under the statutes of limitation (established, of course, under the old legal system) that have already expired, meaning that the only way to open these cases is to revoke those periods of limitation themselves. Can a new polity attempt to prosecute time-barred crimes committed under predecessor regimes? This, let us note, is a matter different from a retrospective determination that certain acts, while formally legal under the Communist regime, should be now deemed illegal ex post facto. The element of retroactivity is very minor in the case of reopening certain cases barred under statutes of limitations: it does frustrate the expectation of an offender that once the specific period of limitation has run the course, he or she has nothing to fear from the law, and indeed does so ex post. However, the truly objectionable aspect of retroactivity, which consists in the violation of the principle nulla poena sine lege, is absent here. This is a point to which I will return below. The two main contrasting approaches to the question of statutes of limitation in CEE are exemplified by the Czech and Hungarian law on the issue. Consistently with their already noted differences in approach to lustration, these two legal systems also display distinctly different attitudes to the dilemma of dispensing material justice versus observing strict legality. In the Czech Republic, in July 1993, the Law on the Lawlessness of the Communist Regime and Resistance to It stated that the Communist regime was illegitimate and that all those who supported it (“[t]hose implementing the communist regime from their positions of officials, organizers and supporters at the level of political and ideological activity”) are jointly responsible for the crimes committed by it. This law, apart from declaring the communist regime “criminal and illegitimate”, also lifted the statute of limitations for criminal offences committed between 1948 and 1989 and not prosecuted for political reasons. It also allowed for the reconsideration of criminal convictions handed down during the Communist period. The Czech Constitutional Court upheld the 1993 law.103 (I have already discussed an aspect of this decision above, in Part I of this chapter). The Court criticised the idea of constitutional neutrality with regard to values, which it associated with legal positivism and with a “legalistic conception of political legitimacy”; in contrast, it found that the Czech constitution is not neutral and merely procedural, but rather “incorporated into its text also certain governing ideas, expressing the fundamental, inviolable values of a democratic society”.104 There is no contradiction – it was held – in the fact that old laws are now re-interpreted in the light of new democratic values. At this point, the Court engaged in a discussion of the principles of the rule of law (or, in its language, the state based on law), and refused to accept a formal definition of legality, instead merging legality with legitimacy, the latter being contingent upon the democratic character of the state (“A political regime is legitimate if, on the whole, it is accepted by the majority of citizens”).105 Hence, the fact that some pre-1990 laws have continued in force cannot be said to endow the old regime with legitimacy: “even while there is continuity of ‘old laws’ there is a discontinuity in values from the ‘old regime’”.106 At this point, and crucially from the point of view of the statutes of limitations, the Court introduced evidence to demonstrate that the Communist regime did not even comply with its own laws: this proved that the constitutional principles (under the
“DECOMMUNISATION” AND “LUSTRATION”
251
Communist-era Constitution) of universal observance of the law “became fictional and hollow whenever the [Communist] party recognized such to be advantageous for its political interests”.107 This is correct and convincing, but this line of argument sits uneasily with the overall theme of the decision, which held that the regime was, on the whole, unlawful (because illegitimate because it exercised power without majority approval). For if the illegality of the system as a whole is asserted, then to demonstrate the non-compliance of the illegitimate system with its own laws is unnecessary; in fact, it weakens the force of the argument (aimed, as it is, at the question of statutes of limitations) because it suggests that, had the regime complied with its own laws (their illegitimacy notwithstanding), the suspension of statutes of limitations would not be justified. In a way, it is an example of a situation in which having two reasons makes the argument weaker than if there had been only one.108 On this basis, the suspension of the limitation period was held to be constitutional because, in the opinion of the Constitutional Court, during the period of Communist rule the law had been simply a tool of the regime and thus limitation could not be said to have actually been running. In fact, the Court’s reasoning is quite complex at this point: it asserts that it does not suspend the operation of the limitation period but rather states that, “for criminal acts, which on political grounds were not prosecuted by the regime then in power, [the statute] declares the period of time during which the limitation was not able to run, even though it should have run”.109 The basis for this argument lies in the provision of the Criminal Code that states that all periods of time during which it was not possible to bring an offender before a court “due to legal impediments” do not count as part of the period of limitation. The statute simply declared that the Communist era was precisely such a period, as far as “criminal acts that on political grounds were not prosecuted by the regime then in power”110 are concerned. In other words, the Court is suggesting that it views the statute not as introducing an impediment to the running of the limitation period, but rather as deeming the institution of limitation to be fictional during the period of lawlessness. This is an ingenious argument, but is it convincing? It may be said that it merely recharacterises the suspension of the statute of limitations in order to render it more palatable; after all, the strategy adopted by the Court had exactly the same effect as an explicit ex-post suspension of the statutes of limitation would have. As for the problem of retroactivity, the Court held that this law did not infringe constitutional guarantees against the retroactive application of laws, because a statute of limitations is not a law but merely a procedural requirement; as such, it is not governed by the Charter of Fundamental Rights. What is banned by the principle of non-retroactivity is the ex post designation of certain acts as criminal, or imposition of a higher punishment than the one that was in force at the time when a crime was committed; this, however, is clearly not the case with regards to statutes of limitations.111 The Hungarian approach to statutes of limitations has been markedly different to the Czech one. For one thing, the Hungarian legislature created a whole series of laws on this issue for the simple reason that the Constitutional Court repeatedly declared each new one to be unconstitutional. At first, the legislature seemed to
252
CHAPTER 9
ignore the Constitutional Court’s reasoning in the creation of new laws, but, in 1993, it started to take account of the jurisprudence developed in these cases and tried to structure its laws accordingly. In 1991, the Parliament passed a law (the so-called Zetenyi-Takacs Act, named after its drafters) stating that, between 1944 and 1990, limitation periods would not be considered to have run for the crimes of treason and murder, where such crimes were not prosecuted for political reasons.112 The law was not particularly wide, as it only covered acts that were crimes at the time they were committed, and even then only those cases in which the decision not to go to trial had been taken for reasons (and not where there was a trial but with an acquittal on political grounds). The President refused to sign the law and sent it to the Constitutional Court, which then declared it to be unconstitutional.113 In this judgment, referred to as the Zetenyi decision (after the abbreviated name of the law that was struck down), it based its decision partly on the unacceptable vagueness in the law, arguing that “political reasons” is not a term capable of one clear definition to cover a period of time spanning almost fifty years. More fundamentally, it based its decision on its conception of the state based on the rule of law. It held that, in a constitutional state, “[n]ot only must the legal provisions and the operation of state organs comply strictly with the Constitution but the Constitution’s values and its conceptual culture must permeate the whole of society”.114 Given that the transition to democracy occurred on the basis of legality, no distinction can be made by the Court between laws enacted before and after the new Constitution; every law must therefore conform to the Constitution and every law must be reviewed in the same way. On this basis, the Court applied the principle of legal certainty – as a fundamental requirement of the rule of law – to the act suspending the statute of limitations, and found the law deficient in this regard; therefore, the law was held to be a form of retroactive legislation, thus offending against the rule of law. The Court declined the opportunity to appeal to “the unique historical circumstances of the transition”,115 and refused to suspend constitutional requirements on the basis of the exceptional nature of the circumstances that warranted – in the eyes of the legislature – this law. As a result of this invalidation, the Parliament adopted, in February 1993, an “authoritative resolution” (an instrument used to interpret other laws) stating that the period 1944 to 1989 should not be included in the time used to calculate limitation periods. In addition, at the same time, the Criminal Procedure Act of 1973 was amended to state that public prosecutors must prosecute in certain situations, even if the statutory limitation period has run its course.116 The authoritative resolution of the Parliament and the amendment to the Criminal Procedure Act were both declared unconstitutional by the Constitutional Court.117 It did this, in addition to procedural grounds, for the same reasons as in the decision described above. In the same month the legislature also created a law entitled “Procedures Concerning Certain Crimes Committed During the 1956 Revolution”, which stated that war crimes and crimes against humanity, being crimes under international law, do not have a limitation period (they are outside the scope of domestically-defined crimes). Thus, on the basis of this law, past crimes could be prosecuted. This law interpreted the events of 1956 as war crimes and crimes against humanity, and made specific reference to the Geneva Convention of 1949 (concerning treatment of
“DECOMMUNISATION” AND “LUSTRATION”
253
civilians during war) and the New York Convention of 1968, which dealt with war crimes and crimes against humanity, and which provided for periods of limitation. President Goncz again sent this law to the Constitutional Court, which this time struck down some parts of the law but upheld others.118 It held that, generally speaking, statutes of limitations cannot be extended. However, in two situations this could occur: (1) Where there was no statute of limitations in existence at the time the crime was committed, or (2) where it related to a crime against humanity or a war crime, covered by international law, which had no statute of limitations. The Court justified this view on the basis that Art. 7 of the Constitution declared that Hungary must respect the rules of international law. In addition, crimes against humanity and war crimes are seen as being so serious that they can be treated apart from other crimes. Given that the New York Convention, to which Hungary is a party, does not prescribe any limitation periods in respect of such crimes, it is reasonable that the domestic situation in Hungary could mirror this approach. In other words, prosecutions could be brought under international, rather than domestic, law. However, the Constitutional Court did invalidate one article of the law under review, which referred exclusively to crimes defined by domestic law. The Court also instructed Parliament to change the law to conform with its decision and, for example, define the crimes it covered in the same way that international law defines crimes against humanity and war crimes. The new law entered into force in October 1993. These two approaches, the Czech and the Hungarian, illustrate well the question of the degree to which the principle of legality should control considerations of substantive justice. The Hungarian approach (as exemplified by the Zetenyi decision of 5 March 1992) gives strong priority to the formal principle of legality, resulting in the rejection of retrospective changes to the statutes of limitation, contrary to the decision of the Parliament. It is rather hard to see what values underlying the principle of legality support such a conclusion. As noted above, the principle of nonretroactivity in criminal law protects the interests of a defendant in not facing consequences that he or she could not have anticipated at the time of planning or committing a given act. This concern, however, is absent from alterations to statutes of limitations: we are not dealing here with acts that did not constitute crimes at the time of their commission, but rather with crimes that went unpunished due to a combination of official tolerance of such acts and the state’s decision as to the period of limitation. The length of this period, it may be assumed, did not weigh substantially on the mind of the offender when he or she undertook to commit the crime in question; rather, it was the anticipation of impunity that was decisive. So there is no substantial ex post facto frustration of the offender’s expectations and no unfair change of the rules by the state. The only aspect of retroactivity that may seem morally offensive here is “a vested interest in repose, on which an offender is entitled to rely”119 once the limitation period has expired. But is such an interest justifiable and legitimate? According to Stephen Schulhofer, “a wrongdoer cannot reasonably claim a right to rely on stated limitation periods in planning his criminal activities, so extensions enacted before the limitation period expires can fairly be applied to offences committed beforehand”.120 But if the first part of that sentence seems compelling (as it does), the second part does not necessarily follow: why
254
CHAPTER 9
would extensions enacted after the limitation period expires be in any different relationship to the wrongdoer’s right (or a lack thereof) to rely on the limitation periods as known at the time of the crime? After all, if the offender indeed factored knowledge about the relevant statutes of limitations into his or her decision whether to commit a crime (a highly unlikely proposition), then it does not really matter whether the change in the period in question comes before or after the expiry of the period as known at the time of the crime. From the point of view of those concerns normally central to arguments against retroactive criminal justice, there does not seem to be any particularly relevant difference between the reinstatement of expired statutes of limitations and the extension of non-expired ones, and both seem to present rather mild, indirect and relatively unobjectionable forms of retroactivity. To be sure, there are also other arguments concerning statutes of limitation than those related to reliance by the offender on knowledge thereof (the evidentiary problems exacerbated with the passage of time, the criminal’s changed personality that may allow us to see him or her as, morally speaking, a different person, etc.). And yet, even these arguments do not undermine the rightness of retrospective changes in statutes of limitations; rather, they vindicate the very idea of statutes of limitations, and in particular the proposition that they should not be excessively long. These are also quite contingent arguments, and courts normally have ample opportunity to take into consideration in their judgments factors such as insufficient evidence or the changed moral character of the offender. One argument that definitely is effective against ex post facto renewals of an expired statute of limitations is that based on “a vested interest in repose”, but one wonders how weighty the legitimacy of this interest is when compared to the countervailing arguments in favour of the reopening of such periods, when the failure to prosecute the offences in the first place was due to deliberate state policy of tolerating politically-motivated violations of individual rights. This is an important consideration in the case of post-Communist legal systems: they are confronted with unpunished offences of the past being time-barred not for the usual reasons that some crimes go unpunished elsewhere (such as the weakness and inefficiency of the police, prosecutorial and judicial institutions), but for the specific reason that those crimes were inspired, mandated and tolerated by the state, as a result of which limitation periods were allowed to expire. The argument that these periods should be unchangeable advanced by the Hungarian Constitutional Court does not ring quite true: the Court explained that “[t]he statute of limitations in the criminal law guarantees lawful accountability for criminal liability by imposing a temporal restriction on the exercise of the State’s punitive powers. Failure to apprehend [the criminal] or the dereliction of duties by the authorities which exercise the punitive powers of the State is a risk borne by the State”.121 However, the “guarantee” that the Court refers to has no relevance here because the “failure to apprehend” and the “dereliction of duties” were part of the purposeful policy of the Communist state, and therefore a respect for statutes of limitations cannot, in this case, perform the role of an incentive for the State to act lawfully in the exercise of its penal functions. The Court stretches the artificiality of its argument to its limits when it talks about “a risk borne by the State” that has failed to prosecute the criminals and allowed the statute of limitations to pass, as if the
“DECOMMUNISATION” AND “LUSTRATION”
255
“risk” in question were a matter of the negligent behaviour of the state, which must then pay the price of its negligence by letting some criminals get off scot-free. Here the non-identity of the “state” before and after the transition is most crucially relevant, and the fiction of continuity at its most absurd. For, in terms of the Communist state, it was not a matter of a “risk” at all but rather of deliberate and lawless protection of offenders, while on the part of the successor state the “price” in the form of non-prosecution is unrelated to its negligent criminal policy. There is, admittedly, something impressive and respectable about the Hungarian Constitutional Court trying hard to confine its treatment of the past to the strict limits of legality, and to make sure that it does not condone parliamentary acts that may be seen to violate the rule of law. The Court goes a very long way in its Zetenyi decision to emphasise that precisely what distinguishes the current notion of constitutionalism from the shameful past is a strict adherence to various precepts of legality, including the observance of legal certainty, even if in some circumstances the ex-Communists – those very people who violated the rule of law in the past – were to benefit from its application today. It is not clear, however, that in this particular case judicial interference with the parliamentary statute was justified, or whether the balance between legality and substantive justice was struck wisely. One can understand, and applaud, the general strategy of behaving as if the exceptional character of the transitional period did not matter here, and as if the usual standards of a mature constitutional democracy were to be scrupulously applied.122 However, the “as if” mode betrays the weakness of this strategy in this particular case because the matter was exceptional, and the specific reason for the expiry of the statutes of limitation for some particular crimes committed in the past must have been an important part of the parliament’s moral and political decision to reopen those periods. This specific reason: a lawless and reprehensible refusal by the old regime to punish those who committed some of the most severe crimes as defined under the law valid at the time, seems to effectively vitiate the general moral reprobation for various forms of retroactivity in criminal law. Put simply, it would seem perverse if the crimes committed in the past were to go unpunished solely because those who committed them were part of the system that protected them, and made sure that, as long as the system lasted, their crimes would remain unpunished. The Czech approach seems, therefore, more compelling: when statutes of limitation were part of a deliberate practice of unlawfulness, it is not a requirement of legality that those limitations be respected now. Note that it does not apply to the law across the board: it is not the case that whatever was done in the name of law under the Communist period should be now disregarded. In fact, the arguments for a generalised principle of continuity are very strong. The transition to democracy in CEE occurred in the context of a peaceful and negotiated settlement within (by-andlarge) old legal rules; there was no revolutionary overturn that would have warranted a radical legal rupture. Continuity rationales are, however, stronger in some cases than others, and in the case of statutes of limitations they seem to be singularly weak. In this regard, the Czech Constitutional Court’s characterisation of changes in periods of limitations as a procedural rather than a substantive issue is quite proper; as outlined above, the particular questions related to the statute of limitations do not raise the fundamental issue of unfairness to the offender. As Eric Posner and Adrian
256
CHAPTER 9
Vermeule have observed, such changes in the statutes of limitations “only apply to those who have, after all, violated underlying substantive law in effect at the time of the violation; usually that substantive law has moral content, as opposed to being a strictly regulatory offence”.123 The defect that the Hungarian Court found in the suspension of statutes of limitations is highly questionable both because there was no clear prohibition in the Hungarian Constitution of retrospective laws (the whole constitutional textual basis of the Court’s reasoning lies in the concept of the rule of law, from which the Court inferred the principle of legal certainty, from which it inferred the ban on retroactivity) and because the objectionably retrospective nature of the reinstatement of statutes of limitations is far from clear upon consideration of the principal rationales underlying the general abhorrence of retroactivity in criminal law (as argued above). This is why the intervention of the Court in the parliamentary action aimed at bringing the perpetrators of some of the crimes of the past to justice can be seen as an arrogation of the power, by the Court, to dictate the terms of the transition, under the guise of self-righteous legalism and a commitment to the rule of law. It is for this reason, perhaps, that the decision was so broadly applauded by Western observers and commentators:124 they could identify with the Court speaking the idiom of liberal constitutionalism and the “civilised” rule of law, as opposed to the apparently revengeful and populist parliament. But there is nothing canonical about this particular interpretation of the rule of law, and by denying the Parliament the authority to define the parameters of transition – the proportions of continuity and discontinuity with old legal system – the Court decided that its own highly arbitrary interpretation of the rule of law should prevail over politically defined understandings of the right mix of legalism and substantive justice. Ruti Teitel expressed this argument well: “the statute-of-limitations decision [by the Hungarian Court] represents a controversial power grab by the court. It is a brilliant power grab in that it appears to represent a victory for the rule of law”.125 Very early after the fall of Communism in Hungary (remember, this judgment was handed down in March 1992) the Constitutional Court successfully, but controversially, challenged the Parliament’s power to define the terms on which the polity handled the darkest legacy of the immediate past: by repeating the mantra of the rule of law (without a textual anchor in the constitution, and under a highly arbitrary interpretation of the concept) it established itself as the ultimate interpreter of the legal parameters of the transition. In the words of Teitel again: “The [1992] case [on statutes of limitations] stands for the proposition that the authority to assess the legality of the prior regime does not lie with Parliament, but instead with the Constitutional Court”.126 This, rather than the details of the argument, is what fundamentally distinguishes the Hungarian approach from that adopted by the Czech Constitutional Court, which, in a decision handed down one year after its Hungarian counterpart (and discussed above), expressed proper deference to the political judgement as to the (il)legality of the former regime and the degree of (dis)continuity between the Communist system and its successor. Michel Rosenfeld, who applauds the Hungarian decision, has articulated the central issue here in terms of the following paradox:
“DECOMMUNISATION” AND “LUSTRATION”
257
past injustices and the flouting of the rule of law may be redressed through application of a retroactivity law, but since the determination of what constitutes a ‘political reason’ itself depends on political criteria, use of a retroactivity law seems bound to undermine adherence to the rule of law.127
True, the Hungarian law under scrutiny in that decision provided that the suspension of the statute of limitations would apply to those crimes that had not been prosecuted under the Communist regime “for political reasons”. But I am not sure why, in itself, this proviso should render the law inconsistent with the rule of law; the rule of law would be endangered if the political criteria were to apply to deciding whether the act constituted a crime, or what punishment should be meted out. However, the extension of the period of limitations for crimes not prosecuted in the past uses the “political criteria” only in a way that tracks the motives behind the failure to prosecute these crimes in the first place. It is therefore not the sort of infusion of political criteria into legal considerations that is particularly offensive to the principle of legality; rather, it merely identifies the crimes that should have been punished, but (for wrongful reasons) were not. This is not to say that the Hungarian law in question (the Zetenyi act) was well drafted, and that it complied in its entirety with the rules of legality. Note that the law applied only to two types of crimes that were not prosecuted despite the fact that they should have been, prosecuted: treason and homicide. Ruti Teitel was correct to observe that the re-imposition of liability for treason (in contrast to liability for murder) actually might offend against the principle of nulla poena sine lege, as the notion of treason under the Communist era was significantly different from that currently prevalent in Hungary; therefore, those who would be punished now for treason as interpreted in democratic Hungary would in fact be liable for an act that was not unlawful at the time of its commission. One can, no doubt, disagree over how far this argument goes: whether “treason” (as defined under Communism) remains exactly the same “concept” then and now, while only the “conception” of what particular behaviour actually constitutes treason has changed. There would be precedents for such an approach: for example, the criminal trials of French citizens in post-Second World War France were based on such pre-war criminal prohibitions as those applying to treason or to acting in conjunction with an enemy power. However, there is something problematic about it from the point of view of the nulla poena maxim: what matters is whether a person committing a particular act had (or could have), at that time, knowledge of its illegality. Those Hungarians who had collaborated with the USSR in a treacherous (from today’s perspective) way most probably had no idea that what they were doing was in fact an act of treason: it was an act officially and authoritatively condoned by the state. Those, however, who carried out murders of political prisoners, could not have thought that their acts were “legal”; rather, they knew that they enjoyed impunity for their crimes. The state never authoritatively and officially announced that the torturing of a political enemy to death is a legal and right act; rather, they denied that such practices took place at all. This difference, which may be a difference of degree only (as it relies on the degree to which particular behaviour is officially condoned as legally correct), is the one between legality and “illegality with impunity”, and it informs the difference between those instances of the suspension of statutes of limitations that violate the
258
CHAPTER 9
principle of punishing only for those actions that were known to be illegal at the time of their commission, and those that do not. 4. CONCLUSIONS: TRANSITIONAL JUSTICE AND CONSTITUTIONAL CONTINUITY As is clear from the above discussion, the “unfinished business” of having to deal with the immediate Communist past elicited somewhat varied responses in CEE. In terms of the two main issues discussed here, lustration/decommunisation and the suspension of statutes of limitations in respect of certain crimes, the moral and political dilemmas raised by these matters have yielded different models and approaches. On the issue of lustration and decommunisation, the main dilemma identified here was the clash between the socially dominant concept of retributive justice (demanding that those involved in the running of the Communist system, and in particular those who had discredited themselves by collaborating with the dreaded secret police, be removed from prominent public positions in the new democratic polity, for some time at least) and the official, legal-political rationale given to the lustration and decommunisation laws, which was future-oriented, consequentiality and prudential in its nature. While the latter types of argument – concerning the need to protect the democratic system against those who lack the necessary qualifications to manage it, or are liable to dangerous blackmail – constituted the legitimate pattern of “public reason” provided for such laws, the shape of the laws was largely overinclusive from the point of view of this rationale; thus the existence and influence of the past-oriented, retributive-justice rationale was revealed. In the end, lustration and decommunisation were undertaken on a reasonably wide scale only in the GDR (where its success was made possible by the special case of absorption of the whole defunct state into the new unified Germany, with its Western part playing the role of collective judge over its Eastern, Communismtainted brethren) and in the Czech Republic. The success of lustration there, compared to all other CEE states where, as was described above, the screening of public figures has been undertaken in only a very weak manner, and has focused mainly on securing truthful declarations rather than providing serious consequences for past misbehaviour, deserves some attention. There have been, no doubt, serious political factors contributing to such an outcome: a combination of the relative harshness of the pre-transition Communist regime (much harsher than, say, in Hungary or in Poland), the only grudging complicity of the old Communist elite in the process of transition itself (distinguishing it from a largely bona-fide, collaborative approach leading to round tables in Poland and Hungary), and the relative weakness of the Communists (or reformed Communists) in the immediate post-transition political constellation when the legal shape of lustration and decommunisation measures was formed. But these factors cannot explain everything: the first two were almost identically present in Slovakia (for obvious reasons), Romania and Bulgaria. An important stimulus for the strong measures adopted in the Czech Republic was provided by the attitude of the Constitutional Court, which was extremely deferential towards the legislature’s harsh lustration and decommu-
“DECOMMUNISATION” AND “LUSTRATION”
259
nisation approaches, and limited itself basically to fine-tuning the lustration system (for instance, by excluding the disqualification of former candidates for collaboration, largely on the basis that their files were unreliable). The composition of the Czechoslovak Constitutional Court, and then its successor court in the Czech Republic (both elected by the anti-Communist parliamentary majorities), was a guarantee to the legislature that they would not interfere actively with the recently enacted systems of lustration and decommunisation. The line of argument adopted by the Court in defending the laws – based on the political if not legal narrative of discontinuity, and of the “lawlessness” of the Communist regime – allowed it to undertake very minimalist scrutiny of measures aimed at purging the present of the remnants of the past, and to tolerate a discrepancy between the consequentiality, future-oriented rationale of the lustration/decommunisation laws and their sweeping reach, which indicated the retributive motives at work in the legislation. In contrast, the lustration practices adopted in Poland, Hungary and other states of the region were much more timid and limited, in the ways described above. They can be seen as principally symbolic measures emphasising the value of transparency more than of purge (although the scale of the actual purge that occurred in the Czech Republic must not be exaggerated either).128 In Hungary and Poland, in slightly different ways, the point was to elicit statements from public figures about their past rather than to remove those who had been implicated in discreditable activities from their current positions or to disqualify them from applying or running for such positions. This was partly for self-serving reasons (the influence of ex-Communists upon the law making process in the transition period) and partly in order to emphasise the non-revolutionary nature of the transition, which necessitated a degree of continuity, including at the personnel level. The discourse of continuity made it very difficult to draw sharp distinctions between different candidates for public offices based on the previous activities of those involved. Similarly, with regard to criminal liability for political crimes committed by the regime in the past, a fundamental line can be drawn between those constitutional systems that have adopted a dominant narrative of a sharp break with the past – and hence, no need to be bound by the legal commitments reflected in the statutes of limitations used as self-serving protections by the ancien régime – and those systems that asserted continuity and insisted upon the strict observance of the rule of law and respect for statutes of limitations. The latter were allowed to be suspended as a result of international human rights commitments (as in Hungary) or constitutional prohibitions on statutes of limitations for war crimes and crimes against humanity (as in Poland), but not as a result of the legislative restoration of liability for crimes with respect to which the statutes of limitations had already expired (contrary to the situation in the Czech Republic). The distinction lies in the understanding of the degree of constitutional continuity with the past, and also in the institutional division of roles played by the political institutions and the constitutional courts in defining the right mix of continuity and discontinuity. An arrogation, by the Constitutional Court, of the supreme power to define this crucial aspect of transition, became in Hungary the starting point for a broader grab for power in which the Court freely translated major political issues into constitutional ones, and provided its own
260
CHAPTER 9
philosophy of the true meaning of the “invisible constitution” in order to control the way in which the polity dealt with the unfinished business of the immediate past. The very concept of legal (dis)continuity is a troubling and ambiguous one, and to present the matter as a simple dichotomy tends to blur rather than clarify the real dilemmas raised by the legacy of the authoritarian system.129 This is not only because there is always a choice of the degree of continuity (rather than a simple option for continuity or discontinuity), but also, and more fundamentally, because even a general commitment to legal continuity does not dictate any particular choice of which specific legal commitments undertaken by the previous regime should be observed meticulously by the successor. Consider this observation from a book by Carlos Nino, focusing on post-authoritarian transitions in Latin America: when the new democratic regime is legally continuous with the old authoritarian one and the human rights violations to be tried were legally protected at the time of their commission or afterwards (say, by an amnesty law), the principles against ex post facto reversal of that legal protection create formidable obstacles to retroactive justice.130
This short sentence hides a number of dilemmas, none of which are conclusively resolved by the initial assumption of legal continuity of the new democratic regime with the old authoritarian one. Firstly, to what extent were the violations of human rights, by the authoritarian regime, indeed protected at the time of their commission? As we have seen earlier, many of the crimes committed (especially in the Stalinist period) by perpetrators within the Communist system were nominally prohibited but de facto tolerated, encouraged and indeed mandated by the system. As a result, the act of punishing a torturer in a Stalinist prison now is not a case of retroactive justice, but rather takes seriously the nominal criminal code in force at the time, and applies it à la lettre, though subject to a delay caused by the failure of the Communist legal system to operate according to its textual commitments. In addition, this characterisation abstracts from other sources of pre-existing law that can now be acknowledged as having been in operation then: international human rights law, constitutional law with its abstract (though unimplemented at the time) commitments to human rights, perhaps even (most controversially) universal natural law. While under the old doctrine, none of these sources affected the actual valid law at the time, under today’s understanding these higher orders of law vitiate much of the lower-level law that could be invoked in defence of the perpetrators of the crimes under the ancien régime, thus reducing the charge of retroactive legislation. They do not eliminate retroactivity altogether (because today’s interpretation is brought to bear on past actions), but they reduce it significantly (because these higher legal orders are thought to have been in force when the crimes in question were committed). One particular legal form of such partial reduction of retroactivity is by means of interpretive statutes, by which legislatures may declare that the positive laws existing at the time of the commission of the alleged crimes, properly understood, never actually authorised the relevant acts. Such an instrument to minimise the retroactivity of justice was adopted, for example, in Belgium after the 2nd World War, where the provisions against treason (valid throughout the Nazi occupation) were said to include not only military activities but also less direct forms of collaboration;131 this is precisely what the Hungarian legislature attempted
“DECOMMUNISATION” AND “LUSTRATION”
261
to achieve – unsuccessfully, due to the Constitutional Court’s intervention discussed above – when it issued, in February 1993, its “authoritative resolution” stating that the period 1944 to 1989 should not be included in the time used to calculate limitation periods. Secondly, Nino discusses a situation in which an amnesty applies in respect of earlier crimes; a subsequent reversal of the amnesty would constitute an ex post facto (and thus, questionable) removal of legal protection. To begin with, an amnesty is itself a retroactive act, so there is an element of retroactivity in any event; an act of reversal of the amnesty would erect one case of retroactivity against another. To be sure, lenient retroactivity (amnesty) is less problematic than harsh retroactivity (reversal of amnesty), but the point is that the option of not relying upon retroactivity at all is simply not available to us in such cases. Further, it makes a crucial difference whether we face an amnesty conferred upon the perpetrators by the ancien régime itself, a “self-amnesty”, so to speak (as was the case of the Argentine military in 1983),132 or an amnesty established during the process of transition itself, as a philosophical formula of dealing with the past by the polity as a whole (as in the Spanish post-Franco and the South African post-apartheid transitions). The moral force of the latter commitment is naturally much higher than the former one; the former is better seen as an attempt at unilateral self-protection by those who fear imminent retribution, and there is no particular moral or philosophical reason why the community as a whole should respect such self-serving actions, even if, formally speaking, it was cloaked in the form of a valid law. There may, admittedly, be some good pragmatic reasons for the non-prosecution of perpetrators of past crimes; these reasons, however, do not derive their force from the concept of legal continuity of the present regime with the old one. Now, the cases of statutes of limitation that protect crimes committed by the Communists during the Hungarian revolution of 1956 can be analogised to an amnesty conferred upon itself by an authoritarian regime: by non-prosecution of these crimes, and by thus allowing them to become time-barred, the old regime successfully brought about a state of affairs practically identical to what it could have achieved by conferring upon itself and its members a blanket amnesty. Consistently with what has just been suggested, there is no special, conclusive obligation deriving from the principle of legal continuity to meticulously observe those privileges, and no obvious reason why to prosecute despite them would be an outrage to the principle of non-retroactivity of justice. One way of glossing over the ambiguity and complexity of the concept of (dis)continuity has been to assert that the principle of legality requires a strong sense of continuity precisely in order to emphasise that the democratic, legality-based regime is not replicating the faults of the previous system, which paid lip service to the notion of legality, and the resulting principle of legal certainty, but which in fact did not observe them at all. This line of reasoning emphasises the virtue of the current system by insisting on the contrast between its character and the vices of the former one, even if it commits the new one to respect the commitments of the old. In the words of one theorist (who does not necessarily endorse this approach): “The justification for adhering to prior law in the transitional moment is that under prior repressive rule, adjudication failed to adhere to settled law. On the positivist view,
262
CHAPTER 9
transformative adjudication that seeks to ‘undo’ the effect of notions of legality supporting tyrannical rule would imply adherence to prior settled law”.133 No doubt, the self-righteous rhetoric of behaving better than one's predecessors (“we believe in the rule of law so we will observe the settled entitlements even if they were conferred by our authoritarian predecessors upon themselves”) has played an important role in the argument of post-Communist constitutional courts (and in particular in the Hungarian Court’s Zetenyi decision, discussed above),134 but the asserted necessity of adherence to prior settled law is highly problematic. It is a non sequitur to say that if a new legal system wants to observe the rules of legality, it must adhere to prior settled law no matter what its content. The range of options is much broader than either full observance of all the entitlement-conferring rules of the predecessor system or a revolutionary rupture with the legal past. The very nature of constitutional change in CEE reveals its partly noncontinuous character: these countries have not followed fully the routine path of constitutional amendment as written into the old Communist constitutions, but rather were initiated by, and developed in, extra-constitutional processes. These took various different forms, such as the Round Table negotiations in Poland and Hungary, the restitution of independence from the USSR (formally expressed in Declarations of Sovereignty – an extra-constitutional instrument par excellence), the restoration of pre-Communist constitutions in Baltic states, the Slovenian war of liberation and its subsequent separation from Yugoslavia, the violent overthrow of the regime in Romania, etc. Consequently, at a macro-level, the idea of full constitutional continuity was eroded from the outset. At a micro-level, the notion of “vested rights” has not been fully adhered to either: the fundamental transformation of the socio-economic system meant that mechanical adherence to previously vested rights would be neither feasible nor fair. Thus in Poland, the Constitutional Tribunal derived (in a fashion characteristic of the other constitutional courts of CEE) the notion of “vested rights” from the idea of a “democratic state based on law”. In claiming that the manner in which these rights are to be respected is dictated by the related principles of (1) trust of the citizen towards the state, and (2) nonretroactivity, the Constitutional Tribunal established at the same time that they were not absolute but subordinate to the principle of social justice. Accordingly, the Tribunal drew a distinction between those vested rights that deserve protection and those that were acquired unfairly or that can be extinguished if their protection would lead to unjust privileges for certain groups, in the radically changed economic situation.135 This indicates that the legislatures and constitutional courts have been, from the outset, dispensing a complex mix of continuity and discontinuity, and therefore that the principles of legality and certainty have never conclusively dictated complete adherence to the specific laws settled under the prior regime. It was always a matter of choice, even if the availability of that choice was often denied by the self-righteous rhetoric of scrupulous legality contrasted to the misbehaviour of the previous regime.
CHAPTER 10 RESTRICTIONS OF RIGHTS
Undoubtedly one of the best ways to deepen our understanding of the nature and scope of constitutional rights is through an examination of the statutory (and other) restrictions on them, to the extent permitted by the relevant constitutions. Rights often collide with other rights (both with other categories of rights, and with the similar rights of other people), and with other important social goals which are often also constitutionally proclaimed. How these rights can properly be restricted is a matter both of the constitutional design of the permissible grounds for such restrictions, and also of the interpretation by constitutional courts of these grounds and of the additional requirements, which together constitute a pattern of scrutiny of the limitations of rights. Both the constitutional design and the courts’ interpretation thereof will be discussed in this chapter. In addition, I will look at a matter that is partly beyond the main theme of this book, namely, the constitutional treatment of citizens’ duties. As rights and duties are often proclaimed in one breath in constitutional texts, with the latter often limiting the former, it is appropriate that some attention is also paid to this topic in this chapter. 1. CONSTITUTIONAL DESIGN OF LIMITS ON RIGHTS As with so many other aspects of constitutional design, all of the post-communist constitutions in CEE have opted for the continental European model in this regard, in which constitutional rights provisions are accompanied by instructions as to the constitutionally permitted grounds of statutory limitations of rights. This contrasts with the United States Bill of Rights model of absolute prohibitions, which leaves the task of establishing limits on rights to congressional acts as reviewed by the federal courts. In continental European constitutions, and in international instruments such as the European Convention on Human Rights and the EU Charter of Fundamental Rights, rights provisions are coupled with clauses setting out the grounds and criteria on which the rights may be legitimately restricted. This, naturally, does not fully dispense with the role of the official interpreters (including constitutional courts) in assessing whether a statutory provision conforms to constitutionally recognised standards (as the instructions regarding permissible limitations are never self-evident and are always subject to interpretation); their role, however, is correspondingly narrower. Nevertheless, within the European model as adopted in CEE, there are different formulations for structuring the relationship between rights provisions and
264
CHAPTER 10
permissible restrictions thereto; and which of these is chosen is far from inconsequential. Imagine a field of possible solutions as a spectrum, with, at one extreme, a constitution that constrains the legislator’s choice of limiting measures to a very high degree, leaving only a narrowly specified set of cases when rightsrestrictions are acceptable, and, at the other, a constitution that leaves broad discretionary powers to the lawmakers to restrict rights whenever they deem fit, with only very general and easily malleable criteria for determining the constitutionality of such restrictions. It then becomes possible to construct a typology of postcommunist constitutional bills of rights according to where they are placed along this spectrum. Broadly speaking, they fall into three categories. Constitutional charters of rights in the first category contain no general clause authorising the legislators to restrict constitutional rights under certain conditions, but rather mention in certain rights articles that those particular rights can be restricted on specific, named grounds. The four constitutions that belong to this category1 seem therefore to imply that the rights that are not accompanied by such clauses explicitly allowing for restrictions cannot be legislatively restricted at all. In turn, those articles that do mention rights capable of legislative restriction often vary the list of permissible reasons for restrictions, in a similar way to the approach adopted in the German constitution.2 For example, the Georgian constitution states, in article 22(1), that there is a right to move freely. Then, in subsection (3) of that article, it goes on to say: “Restriction of these rights is permissible only in accordance with the law, in order to guarantee state and public security as necessary for the existence of a democratic society, public health, prevention of crime and fulfilment of justice”. Then, however, a provision on “the right to receive and disseminate information”3 comes with different grounds for restrictions attached to it: “state security, territorial integrity or public order, prevention of crime, protection of rights and dignity of others, prevention of disclosure of the information recognised as confidential, independence and impartiality of the court”.4 It should be added that, although the general rule in these constitutions is to state reasons for which restrictions are permissible, some articles provide that a restriction can be made by law only, but give no substantive grounds justifying such restrictions. This arguably allows rights to be limited for any reason at all, and in effect relegates these specific rights to the category of unlimited legislative discretion, thus depriving them of strong constitutional entrenchment.5 The second category contains five other constitutions that, while containing a general clause for limiting rights, state that this clause applies only to the rights provisions that expressly allow for statutory restrictions.6 This, again, results in a situation in which individual articles offer specific justifications for and limitations on the permissible restrictions on the right in question. From the point of view of the scope of legislative discretion, the practical effect of these two categories of constitutional constructions of legislative restrictions is identical. The only difference is in the constitutional flexibility of tailoring the specific list of grounds for legitimate restriction to a particular right. A sixth constitution should be added to this group, namely, that of the Russian Federation. Although its chapter on rights and liberties contains a general restricting clause7 (without the proviso present in the five above-mentioned constitutions
RESTRICTIONS OF RIGHTS
265
restricting its application to certain rights), there is nevertheless another constitutional provision8 that explicitly prohibits imposing any restrictions upon a number of exhaustively listed rights.9 In effect, these rights acquire a status identical to those rights that do not allow for statutory restrictions in the five other constitutions that fall into this category. This prohibition on statutory (and other) restrictions of certain rights has been actually used by the Constitutional Court of Russia, in an important 1996 decision regarding the rights of criminal defendants.10 The Court was considering a provision of the State Secrets Act, which stated that lawyers could be barred from participating as defence counsel in criminal proceedings connected with state secrets, on the grounds that they are not authorised to have access to state secrets. The Court found that this provision unlawfully restricted the right contained in article 48 of the Constitution (a right to qualified legal counsel) because it would, in effect, limit the defendant’s choice of legal counsel to a limited number of lawyers (namely, those who were authorised to have access to state secrets). In accordance with Article 56(3) of the Constitution, the Court declared that the right to legal counsel may not be restricted in any circumstances, and the Federal Assembly was thus ordered to make the necessary changes to the legislation. The technique of constitutional drafting adopted makes a real difference between these first two categories, on the one hand, and the third category of constitutions (which is clearly the dominant model in CEE, comprising all of the constitutions in the region barring the ten already mentioned) on the other. Explicit permission for statutory restrictions of rights is contained in a general clause, which also lists the grounds upon which any right provisions in the constitution may be limited. These clauses are typically expressed in the following form: Restriction of personal rights and liberties shall be permitted only in the instances specified in law, in the interest of national security, public order, the protection of the morals and health of the population, as well as rights and liberties of other persons.11
The lists of grounds for restrictions vary somewhat from country to country: some lists are narrower,12 while others refer, in addition, to international human rights standards for restrictions of rights.13 Certain constitutions belonging to this category also explicitly require that limitations must be proportionate to the aim for which they are imposed – a point to which we will return, in some detail, below.14 In addition, even when specific constitutions in this group explicitly state that the list of permissible grounds of restrictions is exhaustive, some of the constitutional courts have established this principle in their case law.15 The fact that all of the constitutions belonging to this third category have general restricting clauses does not prevent them from also listing, in some specific rights provisions, further grounds upon which these particular rights can be limited. Such extra limiting clauses should, however, be considered as an addition to, rather than an exception from, the general clause. Some of these particular articles simply repeat one criterion for restriction from the general clause, restricting the limitation of that right to the criteria that are relevant to the particular situations in which a restriction on the right in question may plausibly occur.16 What, however, renders the third category of constitutions distinct from the first two is that, in addition to particular
266
CHAPTER 10
grounds for restrictions applying to particular rights, there is also a general clause that operates in a blanket fashion with regard to all constitutional rights and liberties. 2. CONSTITUTIONAL REVIEW OF STATUTORY LIMITS ON RIGHTS: PROPORTIONALITY SCRUTINY Five criteria for acceptable legislative limitations on rights cut across the typology of constitutions suggested above, and deserve special mention because they have played an important role in the judicial review of statutes under rights provision. In general, permissible restrictions must be: proportional to the goal they seek to achieve; in conformity with the essence of the right in question; established by statute; non-discriminatory in character; and be imposed for reasons other than the lawful exercise of another right. Of these, by far the most important and the most complicated is the proportionality requirement; hence, I will discuss it in some detail in this section. The requirement that a restriction must remain in proper proportion to constitutionally-mandated goals is perhaps the most powerful tool that the constitutions of the region granted to constitutional courts, as the scope for complex and discretionary judgements concerning the relative weight of competing social and individual interests is the greatest here. These judgements, in turn, involve considerations of policy even if, on the face of it, it is only the “constitutionality” of legislative measures that is being considered. Such “constitutionality” entails an evaluation of the relationship of the means to the ends, of their effectiveness, and even speculations about the availability of other, hypothetical, measures that may better attain certain goals. Naturally, such decisions cannot be made in isolation from complex (and often controversial) policy judgements at to the anticipated efficiency, costs and benefits of proposed measures. This is what, normally, we expect from the legislatures: their role, in adopting new laws, is to decide on the proper balance of the respective values of competing social and individual interests. A “balancing jurisprudence” is therefore an example of a quasi-legislative function performed by constitutional courts. It is in this area that the question of the democratic legitimacy of these courts arises with particular force. At the outset, it is important to recognise a strict link between the constitutional standards of “proportionality” and “necessity” in this field. A number of constitutions in CEE explicitly include a “necessity” requirement for statutory restrictions on rights, for example, that such restrictions are legitimate only if necessary17 or proportionate18 (or both)19 to achieve constitutionally defined goals. However, even if proportionality is not explicitly required in constitutional texts, such a requirement has often been established by the doctrine of the relevant constitutional courts, as is the case, for example, in Hungary, Slovenia20 or pre-1997 Poland. This “necessity understood as proportionality” was explicitly stated, for example, in the Slovenian Constitutional Court decision on bugging devices and privacy, in which the Court held:
RESTRICTIONS OF RIGHTS
267
By the ‘necessity’ of the infringement, the principle of proportionality is explicitly built into the Constitution . . . . This demands of the legislator that in determining the conditions for an encroachment [on constitutional rights] it enables a judgement of whether the encroachment is necessary, such that the desired aim cannot be achieved by less extreme means.21
The courts in CEE have clearly followed the path of the proportionality doctrine as developed by their Western counterparts, and, in particular, by the European Court of Human Rights (ECtHR). In the jurisprudence of the ECtHR, the requirement of “necessity” contained in Articles 8-11 of the Convention (namely, that restrictions on these rights must be “necessary in a democratic society”)22 has actually acquired a meaning synonymous with “proportionality”, or, to be more precise, the test of “proportionality” has been found to be an important one in establishing that the “necessity” requirement has been met. The ECtHR has established, in a number of decisions, an authoritative interpretation of the Convention’s formula “necessary in a democratic society”: that the interference with a right must correspond to a “pressing social need” and be “proportionate to the legitimate aim pursued”.23 As one commentator has noted, “from ‘necessity’ to proportionality is but a small step”,24 and this step has repeatedly been made; indeed, the notion of “pressing social need” has been authoritatively established as a test for “necessity”. Under this interpretation, “necessity” qua proportionality is a rather flexible notion that allows for a relatively broad range of measures to be found “necessary” – even if they are not “necessary” in the sense of being “indispensable”, or being sine qua non. It is significant that, at times, the ECtHR jurisprudence has held that “necessity” is analogous to the requirement that the reasons for a restriction be “relevant and sufficient”.25 The relationship between these various notions: proportionality, necessity, relevance, etc., is a complex one, and some attention needs to be paid to the conceptual issues involved before we can see more precisely what is at stake in the proportionality doctrine in CEE constitutionalism. To begin with the last point, necessity is understood as “relevance and sufficiency”. On the face of it, there may be a measure that is relevant (i.e., related to the achievement of an aim) and sufficient to achieve an asserted aim (that is, once the measure is applied, the aim will be achieved without any other action required) and yet not “necessary”, because that same aim can be achieved by using some other means. For instance, if we wish to make sure that children will not be assaulted in the streets in the evening we might impose a curfew upon minors at certain times: the measure will be “relevant” (there is a connection between the means and the aim) and “sufficient” (once you apply the means and enforce it strictly, it will be enough to achieve the goal) and yet hardly “necessary” in the sense of being sine qua non. In turn, if we find that a measure is necessary (in terms of a “but for” test) then it is ipso facto proportionate: if a legislator is required to pursue a particular goal, and there is a measure that is “necessary” to achieve that goal, then this measure cannot be found disproportionate. To find any “necessary” measure disproportionate would amount to disabling the legislator from pursuing a goal that it is obliged to pursue. Initially, therefore, we may conclude that any measure that is necessary is ipso facto
268
CHAPTER 10
proportionate, but not every measure that is unnecessary is ipso facto disproportionate (because not every measure that is proportionate is necessary). The notion of “relevance”, taken on its own, is an extremely rudimentary threshold test, which, in itself, hardly does any work at all. As Robert Alexy puts it, the test of relevance (which he translates into English as “suitability”) as used by the German Constitutional Court’s three-tiered principle of proportionality,26 can be conceptualised as a maxim that “excludes the adoption of means obstructing the realisation of at least one principle without promoting any principle or goal for which they were adopted”.27 However, in practice, any legislative action will have some costs, and therefore can be seen as “obstructing” (at least minimally) some “principles” (for example, of economy of effort and costs); therefore, if the measure in question does not promote any principle or goal for which it was adopted at all, then this can be seen as so irrational as to defeat the legislative restriction even without any further investigation into how it “obstructs” the realisation of any other principles. Here is how Alexy explains the test of “suitability”: “If a means M, adopted in order to promote the principle P1, is not suitable for this purpose, but obstructs the realisation of P2, then there are no costs either to P1 or P2 if M is omitted, but there are costs to P2 if M is adopted”.28 My point is that the italicised words pertain to an investigation that is, in fact, redundant; if M is not suitable to the purpose for which it was adopted in the first place, it can be easily struck down without any further ado: its negative effect upon P2 is something that can be left out of the picture altogether. I therefore disagree with Alexy’s claim that, under his example, “P1 and P2, when taken together, prohibit the use of M”.29 To include P2 as part of the overall investigation of suitability/relevance would lead us to the troubling conclusion that a lawmaker may legitimately take irrational actions (that is, actions that do not help achieve, even minimally, the purported aim) if only they do not have adverse effects upon other “principles”. That puts an unnecessary onus on those required to show that the costs implicated by an irrational measure can be expressed as an “obstruction of a realisation of a principle”; the mere irrationality should be enough to defeat the unsuitable measure. This is particularly true of the actions of public authorities, and, in particular, law- and policy-makers: they are minimally legitimate if they are performed for the right reasons, and the rightness of reasons includes, as a necessary condition, that the measure in question can help achieve the goal for which it was adopted. The test of “necessity” is more difficult to meet than the one of relevance, and therefore its use by those who scrutinise the law reveals a lower deference, or a higher degree of distrust, towards their respective legislators. Such a test involves a counterfactual: an inquiry into whether there are any other measures, less intrusive than the proposed one, that would also lead to the constitutionally mandated goal. Theoretically, postulating the test of “necessity” as a criterion for the constitutionality of restrictions on rights must result in an invalidation of any rightsrestricting measures as long as we can plausibly think of some other measures that do not restrict the right (or that restrict it to a lesser extent, or that restrict a less important right), and that also lead to the attainment of the goal pursued by the measure under challenge. And since it is almost always possible to think of some alternative means of achieving the same aim, the requirement of necessity may turn
RESTRICTIONS OF RIGHTS
269
out to be, as one American scholar has famously suggested in a similar context, “‘strict’ in theory and fatal in fact”.30 This, however, would be a pedantic approach to the “necessity” requirement, and an ultimately useless one. It does not take into account the “efficiency” of the attainment of the goal, understood as the degree of attainment of that goal reduced by the degree of negative side effects (defined in terms other than the failure to achieve fully the goal). Suppose that a measure under challenge, M-1, which restricts a constitutional right R-1, is adopted by the legislator in order to achieve goal G. Under a pedantic reading of the necessity test, M-1 is unconstitutional if we can plausibly think of another measure, M-2, that is also capable of achieving G but that does not restrict (or restricts to a lesser extent) R-1. But what if M-2 achieves G to a lesser extent than M-1 does, or/and produces higher negative side effects (in terms other than the restriction of R-1) that are also constitutionally relevant? These two aspects of the “efficiency” of attaining a goal, expressed together in the last sentence above, should be treated separately because they implicate the test of necessity in somewhat different ways. Consider two scenarios, corresponding to the two aspects of “efficiency” just mentioned. Scenario 1: M-2 achieves G less efficiently in the sense that it will take more time, or more resources, or both, to achieve this goal compared to the situation in which we employ M-1. All else being equal, M-2 represents less of an advancement towards goal G than M-1 does. This could perhaps be called a matter of “suitability”: the measure M-1 is not strictly “necessary” but is the most suitable to achieve G. But this should not result, per se, in the choice of M-1; this would render the “necessity” requirement redundant. Rather, we have to balance the degree of restrictions of R-1 (considered as losses, from the constitutional point of view) with the gains in terms of higher efficiency of achieving the aim G. If we conclude that the gains (in terms of a more efficient attainment of G) resulting from M-1 outweigh the losses (in terms of restricting R1), and on that basis uphold M-1, then in fact we have not met the test of “necessity”, but only that of suitability; that is, we have upheld the measure on the basis that it is “the most suitable” to achieve G. In other words, this requires the weighing and balancing between a right on the one hand and constitutionally meaningful goals on the other. Under a Dworkinian theory of rights, in such cases the rights should presumptively prevail, that is, goals may be decisive only if they are of particularly high urgency. Now consider Scenario 2: the relative inefficiency of M-2 (compared to M-1) consists in its negative side effects (“externalities”) other than the under-attainment of G. If these side effects can be characterised in terms of their negative impact upon other people’s constitutional rights (which, of course, is only one among a set of different categories of negative side-effects), then the balancing occurs within the realm of constitutional rights themselves. To choose M-1 over M-2 on the basis that the choice of M-2 would bring about a reduction of other rights of other people is then based on a balancing of the degree of restriction of R-1 caused by M-1 with the negative impact on R-2, R-3 etc. produced by M-2. This calls for a judgement of the relative importance of R-1 compared to R-2, R-3 etc., and also for a comparison of the degrees of restrictions of these rights. Suppose that we conclude that the restriction of R-1 resulting from M-1 produces lesser harm than the externalities
270
CHAPTER 10
produced by M-2. We may then say that M-1 is “necessary”, and this notion of “necessity” of restriction should be more palatable than in Scenario 1 because it does not involve the comparison of rights and (non-rights-related) goals. At the same time, the argument under Scenario 1 is more deferential towards a legislator; it renders the legislator's task of defending the statute easier. This is due to the fact that it is harder to identify a specific “right” (R-2, R-3 etc) that can figure in the justification for a restriction of R-1 than it is to identify a constitutionally meaningful “goal” (but non-right) to play this role. Especially when the goals are defined in an extremely broad and vague way, such as “public interest”, it is easy for the legislator to claim constitutional grounds for the restriction of a right. Either way, the “necessity” test calls for a balancing in both scenarios: either balancing of rights and constitutionally meaningful goals, or balancing among rights. It is the second scenario that is more palatable to those who “take rights seriously”, because some rights will be affected regardless of the choice made. This, in turn, is why those decisions of constitutional courts under the “necessity” test that follow the pattern of the second scenario are less problematic for a defender of a robust conception of constitutional rights. The first scenario also uses the necessity test legitimately, as the goals that figure in the rationale for rights restriction also have a constitutional value. There is no clear textual reason why they should be less appreciated than the constitutional rights in question. Such a reason can be only provided by a doctrine along the lines of the Dworkinian theory of rights as trumps over collective (even if constitutional) goals. For anyone accepting this theory, the idea that a social goal (irreducible to a clearly individuated right) may override a constitutional right is deeply problematic; therefore, the threshold of argument justifying such an override (and thus legitimating the restriction) should be placed appropriately higher than in the case of the second scenario. With this theoretical compass in hand, we can now look at some examples of proportionality-based decisions of constitutional courts in the region. Consider first an example of a decision in which a restriction on a right was defended on the basis of its proportional connection to the protection of other people’s rights: The Polish Constitutional Tribunal has frequently considered the question of the “proportionality” of restrictions to the constitutional goals proclaimed as justifying those restrictions, and it has explicitly derived the proportionality principle from the constitutional requirement of “necessity in a democratic society”.31 In 1999 it considered the application of this requirement to certain limits upon the constitutional right to ownership.32 Under scrutiny was a provision of the building law that provided for demolition by the state authorities of those buildings that were constructed without the proper planning permission (art. 48 of the statute). The proceedings were triggered by a “legal question” addressed to the Constitutional Tribunal by the Supreme Administrative Court (SAC), who expressed doubts as to the constitutionality of the provision under the proportionality requirement. The SAC was of the opinion that the same goal – to stamp out the “anarchy”33 in the field of land development – could be achieved “by applying means . . . that are less restrictive and at the same time more effective”.34 One of the arguments raised by the SAC was that this was a disproportionately harsh response to a failure by the developer to secure building permission because it affected equally both those
RESTRICTIONS OF RIGHTS
271
developers who had violated the “substantive” requirements of the construction law (and hence would have had no chance of obtaining a permit in the first place) and those who had merely committed minor procedural faults by failing to secure all the necessary documentation. The Constitutional Tribunal rejected this argument and asserted that even purely procedural failures create a threat to the rights and liberties of the third parties, given that one important purpose of procedural requirements it to guarantee the procedures by which third parties may assert that their rights and interests are affected by a proposed development. In conclusion, the Tribunal found that the regulation of art. 48 was “closely related” to the protection of constitutional values determined by art. 31 (3) of the Constitution,35 such as securing “public order”, protection of the environment, and also avoiding threats to other people’s rights and liberties. In fact, the Tribunal drew a clear distinction between these two grounds for interfering with rights: it held that protecting constitutionally mandated public interests and protecting specific rights of other people are “qualitatively different” types of situations.36 It has not as yet drawn the conclusions suggested above in this section of this chapter (that the legislator should be treated less deferentially in the latter situation than in the former), but it seems that only a small step is needed to reach such a conclusion. In this decision, the Tribunal characterised the proportionality principle (“found” by the Tribunal in the necessity requirement) as a “prohibition on an excessive legislative interference in the sphere of individual rights and liberties”.37 The Tribunal announced that this requirement calls for scrutiny of (a) the existence of a “real need” to interfere with a given right or liberty, and (b) the efficiency of those means and their necessity in the sense of protecting certain (constitutional) values in a way, or to a degree, that cannot be achieved by any other means.38 This immediately suggests the validity of our earlier analysis that “necessity” cannot be understood in a pedantic way, but also needs to take into account the efficiency dimension. In the formula adopted by the Constitutional Tribunal, the textual, constitutional requirement of necessity entails the criteria of indispensability (no other means can achieve the goal to a similar degree), usefulness (there is a real need to interfere), and proportionality sensu stricto understood as a rational and adequate proportion of interference with the right in question to the goals that figure as justification for the restriction. In this, it largely tracks the German Constitutional Court’s proportionality analysis.39 For an example of a restriction of rights justified in terms of its relationship to a social goal (in this case, defined as protection of children from alcohol) rather than other people’s rights, consider the following case from Estonia. Article 11 of the Estonian Constitution (on restrictions of rights) does not mention proportionality, but does contain a necessity requirement which has been transformed by the Constitutional Review Chamber into a proportionality condition. In its decision on licenses for selling alcohol,40 the Chamber considered subsection 1(2) of section 19 of the Alcohol Act, which provided that the licence must be revoked if its holder seriously violates the procedure for the handling of alcohol. The Chamber considered whether this limit upon the right to engage in commercial activities (art. 31) was consistent with the general clause of Art. 11 which provides that “Restrictions [on constitutional rights] may be implemented only insofar as they are
272
CHAPTER 10
necessary in a democratic society, and their imposition may not distort the nature of rights and liberties”. Thus, the Chamber concluded, restrictions must be proportionate to their desired aim.41 The aim of the statutory restriction was characterised by the Court as “protecting society against untrustworthy salespersons” and, in particular, making sure that alcohol is not sold to children.42 The Court then noted that the Alcohol Act gave no choice to the issuer of activity licences as to the penalty for a serious breach of alcohol handling. It left no room for evaluation, in each concrete case, as to whether a revocation of the license, constituting as it does a restriction of rights and freedoms, was necessary in a democratic society: "The law does not allow [the issuer of licences] to take into consideration the circumstances of breaching the procedure for the handling of alcohol, for example the age of the buyer, the quantity and strength of alcohol sold" etc.43 This, the Chamber concluded, was not legitimate: the legislator should give the executive the possibility of taking the circumstances of each case into consideration so that the infringement of people’s rights can be justified. Thus, the Chamber held that subsection 1(2) of section 19 of the Alcohol Act conflicted with Articles 31 (the right to engage in enterprise) and 11 of the Constitution, and thus was null and void. This decision nicely illustrates the interchangeability of the “necessity” and “proportionality” requirements. This equivalence is also illustrated by a Hungarian decision on preventive custody pending trial in which a restriction on one’s personal liberty rights, supposedly justified by the public interest in public safety, was challenged; the Court rejected the measure in question as disproportionate.44 This case related to Article 92(1)(c) of the Criminal Procedure Code, which allowed custody pending trial to be ordered if the person was accused of committing a crime that can be punished with imprisonment, and there were reasonable grounds for suspicion that they would commit another crime if released. The Constitutional Court noted that custody pending trial is not in itself unconstitutional, since it is essential in some cases and, as a preventive measure, it does not violate the presumption of innocence (guaranteed in Article 57(1) of the Constitution). To place a person in custody pending trial also does not violate Article 55 of the Constitution (that anyone suspected of a crime be either promptly released or be brought before a judge who shall promptly decide whether the person shall be released or detained). However, it also held that allowing detention in order to prevent someone committing another crime is disproportionate to the aim involved in Article 92(1)(c) of the Criminal Procedure Code (namely, that of the public interest). Therefore, the provision was found to be unconstitutional. Note that, in this example, what was at work was a notion of “proportionality” that makes sense only if viewed through the lens of necessity, rather than the other way round. If the aim of a measure is defined as broadly as “public interest”, and the prevention of commission of another crime certainly qualifies as “public interest”, then the judgment that preventive custody is “disproportionate” to the aim is very questionable, especially given that such custody was not mandatory, and it may have been reasonable for the enforcers of the law to believe that the most effective measure (in terms of the proportionality of means to ends) to take in order to prevent the re-commission of crime was to place the accused in custody. But such a measure certainly does not meet the “necessity” test in the strict sense of the word: one can
RESTRICTIONS OF RIGHTS
273
think of some other measures (arguably, less efficient and less economical) that may achieve the same aim; for example, surveillance, monitoring, a duty to report regularly to the authorities, etc. The connection between necessity and proportionality suggested above can be tested by a case in which a Constitutional Court upheld (rather than invalidated) a given restriction on a right, on the grounds that it met the condition of proportionality. As an example, consider another decision of the Hungarian Constitutional Court.45 It concerned a provision of the Criminal Procedure Code which stated that a document containing testimony could be used if the person who had testified refused to do so again in court. Such a restriction of the defendants’ rights, including the right to remain silent (the Constitutional Court said), can be justified if it is necessary and proportionate to the State’s duty to punish criminal offenders. The right to remain silent is not absolute: rather, it is a safeguard aimed at ensuring fair trials, which can in certain circumstances be removed if other safeguards of fair trail are put in its place. The Court found that there were such other safeguards envisaged by the law on criminal procedure: for example, the investigator is obliged to tell the accused at the pre-trial stage that he can refuse to make a statement. If he decides to make a statement nevertheless, it is reasonable that he cannot later decide that it cannot be used at trial. The Constitutional Court emphasised that this provision can only be used in order to clarify the facts or in the interests of another accused or the victim. In addition, the judge should obtain evidence from other sources, even if the accused made a full confession. These elements protect the right to fair trial, despite the restriction of the accused person’s right against self-incrimination. In this case, the Court found a particular restriction of a right proportionate to the purpose of the restriction (to facilitate the punishment of criminals) and used the “necessary and proportionate” formula, but obviously it would have been very hard pressed to show that such a restriction was “necessary” to achieve that aim. Reading the accused person’s earlier testimony may be useful (hence, proportionate), but it is an exaggeration to say that it is “necessary” in order to achieve the aim of effective crime control. Another device for striking down the law under the proportionality requirement is to note the defects of the purpose chosen by the legislator to justify a restriction. It is clear that proportionality, as a relational concept, involves scrutiny of both the relationship between the restriction and its purpose, and the purpose itself. If the purpose is faulty (due to vagueness, or because, even if precise, it has no constitutional status), then further scrutiny of the proportionality of the challenged measure is pointless: the law should be invalidated at the outset. As an example, consider the Romanian Constitutional Court’s decision concerning a tax on foreign travel.46 Under challenge here was the government ordinance of 12 August 1994 (subsequently approved by the Parliament) on the setting up of a tax for passing the frontier, the proclaimed purpose of which was to bring in some additional resources for welfare purposes. The Court established first that the regulation clearly constituted an impediment on the exercise of the constitutional right of free movement within the country and abroad (Art. 25 (1)). Thus, to be upheld, the restriction must conform to Article 49 of the Constitution, which contains the general conditions for restrictions of rights, including that such
274
CHAPTER 10
restrictions must be “absolutely unavoidable” to achieve a number of constitutional goals (such as the protection of citizens’ rights and freedoms) as well as the requirement of proportionality. The problem for the Constitutional Court was that an earlier Court decision47 on a very similar issue – also concerning the constitutionality of a foreign travel tax – was deemed to conform to the Constitution. This time, however, the Court found a relevant difference in the degree of specificity of the asserted purpose of the regulation. In the earlier case, the tax was imposed (temporarily) for the specifically asserted purpose of funding tenement heating assistance, during the period of November 1993 – April 1994. In the opinion of the Court, this was an exceptional situation resulting from the lack of funds necessary to guarantee a constitutional goal; namely, that “the State shall be bound to take measures of economic development and social protection, of a nature to ensure a decent standard of living for its citizens” (Art. 43 (1)). As the Court remarked, this measure was justified by an “exceptional situation resulting from the lack [sic] budget funds necessary for the institution of a measure for protection, on the grounds of Article 43 paragraph (1) of the Constitution”.48 However, the Court found that, in the present case, the law in question was not specific at all. The government simply claimed to be acting to advance “social protection rights” generally. This was not sufficient for the Court: a specific right that figures in the rationale for a restriction of another right must be identified so that the proportionality of the restriction to its aim may be scrutinised. Further, the Court denied the constitutionality of extending what had initially been a special and temporary measure for an indefinite period of time: it would mean in effect that the right to travel would be restricted on a permanent basis. It is significant that both these decisions of the Court turned out to be popular with the general public and were generally acclaimed: the decision upholding a special temporary measure (when the cost imposed upon the travellers to subsidise disadvantaged people in time of need was seen as socially just) and the decision rejecting an imposition of such redistributive transfers on a permanent basis.49 The theoretical force of this decision is, however, doubtful, and the problem goes back to the very issue of whether the goal of promoting a decent standard of living can be considered as a “right”, and therefore whether the earlier regulation could be subsumed under the rubric of protecting other people’s rights. (To be sure, article 43 concerning “living standards” is included in the chapter on “Fundamental Rights and Freedoms” of the Romanian Constitution, but the wording of section 1 of the article, which constituted the basis for the temporary tax, is clearly in terms of a directive for government policy). If it cannot be properly viewed as a right, then it can hardly figure in the justification for a restriction on a right, because no social goals listed in Art. 49(1) as justifying the rights restrictions51 fit the rationale given for the temporary tax on foreign travel. However, regardless of the force of the distinction between the earlier and the current case, the argument in the current case can easily be read as an objection to the undue vagueness of the asserted purpose of a regulation. It is the purpose that is faulty, and so the examination of the means-ends “proportionality” is no longer necessary. The last example of proportionality-based decisions raises slightly different issues. This decision of the Hungarian Constitutional Court concerned children’s
RESTRICTIONS OF RIGHTS
275
membership in associations of homosexuals.52 It was issued as an advisory opinion at the request of the President of the Supreme Court, and it was triggered by a court refusal to register an association campaigning for the rights of homosexuals because it had not excluded minors from its membership. The Court characterised the dilemma as a conflict between two constitutional rights: the right of a child to protection (art. 67) and to freely join associations (art. 63(1). In this context, the proportionality of a limit on the latter right (in the context of an age limit on joining an association for the protection of homosexuals) was considered. The Constitutional Court noted that the question was not about the constitutionality of the aim of an association (namely the protection of homosexuality) but rather about the restriction of the rights (of association) of the children as justified by their own interests. Rights may be restricted if to do so is necessary to protect another constitutional right and if the extent of the restriction is proportionate to the desired aim. Here, the other right was identified as the right of a child to be protected. The Constitutional Court interpreted Article 67 of the Constitution as including both protection against clearly harmful effects and also avoidance of serious risks to the child's personality. The Court concluded in this case that the state must prevent children from taking risks that, due to their age, they are not able to evaluate. The Court was careful not to say that becoming a homosexual would endanger the "moral development" of a child: this would amount to making a moral judgement about homosexuality. However, the Court asserted that the Constitution did protect a child's right to decide with full knowledge of the possibilities and the consequences. Further, there is a theme in the Court’s decision that suggests that declaring oneself to be gay in Hungarian society may harm a child because of the views currently prevailing in that society towards this minority. Joining such an association would increase the possibility that a child who is unsure about his or her sexuality would pre-empt his or her future, more mature decision. For these reasons, the Court decided that an age limit could be imposed on such associations, and that such limits were not unconstitutional. Note that this is a special case of the “proportionality” argument, which does not fit clearly either of the two “scenarios” identified above: it concerns neither the restriction of a right for the sake of the promotion of constitutionally mandated “goals” (as the purpose of the restriction is defined in terms of children’s “rights” to protection), nor for the sake of protection of other people’s rights.53 Rather, it is a typical example of paternalistic interference, even if the Court admitted this only indirectly:54 the restriction of a right is justified by the interests of those whose rights are being restricted. It is also a case of “direct” paternalism: the very person who is denied a particular liberty-right is the one whose interests are allegedly protected.55 For this reason, the language of proportionality (normally triggered by a conflict between a right and a general social goal, or a right and another person's right) seems to be inadequate here: it is rather about the suitability of a paternalistic measure for an asserted aim. Typically, such a direct paternalistic measure would be anathema to anyone endorsing a liberal conception of rights; here, however, the problem is mitigated by the fact that the putative right-holders are minors – and so paternalistic measures are, in principle, applicable to them.
276
CHAPTER 10
As this quick survey suggests, scrutiny of the proportionality of challenged measures to purported aims has been a frequently used, and powerful, device for controlling and sometimes displacing legislative choices by constitutional courts. It has been applied regardless of whether there are explicit grounds in the constitution of a given country for such a scrutiny, i.e., regardless of whether the constitution requires rights restrictions to be proportional to constitutional aims. This is an inevitable consequence of entrusting constitutional courts with the role of checking the constitutionality of legislation for when explicit grounds for permissible statutory restrictions are spelled out. Under such a constitutional design, the negative impact of the exercise of one right upon the exercise of another (or upon constitutionally recognised goals) means that the rights in question have to be compared with one another, and this calls for the balancing of respective interests. No such balancing, however, is inevitable when the constitution does not provide the grounds on which rights may be restricted, as in the case in the United States. The Supreme Court can, and often does, conduct a balancing operation, but this is by no means inevitable; indeed, it has not always been so, as T. Alexander Aleinikoff showed in his classic article on the subject.56 However, the continental European tradition, notably exemplified by the European Convention of Human Rights and, most recently, and even more explicitly, by the EU Charter of Fundamental Rights,57 provides a mandate for those charged with the task of scrutinising the constitutionality of statutes to engage in balancing of the sort described above. Inevitable or not, the application of the doctrine of proportionality shows what a broad scope there is for constitutional courts to play a role in the legislative process: proportionality requires balancing, and balancing relies on judgements about which reasonable people – equally committed to constitutional fidelity – may disagree. As Aleinikoff correctly points out, a balancing methodology expresses “the activist, policy-oriented approach to constitutional law”.58 If the balancing consists of comparing the costs of restricting a given right with the costs to other people's enjoyment of their rights, the argument is more palatable than when the balancing entails comparing a right with social goals, even if the latter are constitutionally proclaimed. But this distinction has been rarely acknowledged, and drawn upon, in the proportionality-based decisions of constitutional courts. In any case, balancing offers the illusion of a “scientific” assessment: a metaphor of mathematical operation in which some commensurable goods are put on scales, and compared. The fact that this is not the case, that no such “objective” measure is available to us (or to the courts) is not, per se, an argument against balancing itself: we engage in decision-making based on comparing incommensurable values all the time. It is, however, an indication of how close this type of methodology applied by constitutional courts comes to the realm of legislation, and how carefully it should be handled. 3. OTHER STANDARDS OF RIGHTS RESTRICTIONS Among the criteria of acceptable grounds for restrictions on rights listed by general constitutional clauses, the requirement that a limit must not strip away the essential
RESTRICTIONS OF RIGHTS
277
quality of the right has a special character. This “essential quality” may be constitutionally defined as “the essence” of a right,59 “the essence and significance” of the right;60 “the existence” of that right or liberty;61 or the “nature” of the right that must not be distorted.62 If one allows for the vagaries of translations into English, the differences of wording are insignificant. What is important is the very idea, influenced by the German approach, that there is a “core” of a right that must not be affected, even if the criteria for constitutionally permissible limitations are scrupulously observed. Not one of the constitutions spells out any further criteria as to what this “essence” consists of, and the task of fleshing out this notion is left to the official interpreters, and in particular, to the constitutional courts. Significantly, there is a partial overlap between those countries that have activist, strong constitutional courts and those whose constitutions have a “non-infringement of the essence of a right” clause. Notwithstanding certain examples to the contrary,63 it must be acknowledged that the “essence” clause has been a significant constitutional device in checking legislative discretion in restricting constitutional rights.64 As an example, consider an important decision of the Constitutional Court of the Czech Republic on conscientious objection.65 The Court struck down a provision of the Act on Civilian Service, which established that declarations of refusal to perform military service submitted after the deadline of 30 days after the date of conscription would not be taken into consideration. The case was brought to the Constitutional Court by a conscientious objector whose request for an alternative service was not taken into account because he had let the time limit of 30 days pass. He was subsequently prosecuted for his refusal to report to military service. The Court acknowledged that there is a legitimate aim to restrict the right to conscientious objection and to alternative civilian service: the army needs to have a reasonable degree of certainty as to the number of soldiers it will enlist. However, the Court further found that the provision under challenge totally precluded there being any possibility of exceptions to the 30-day rule, and that this thus violated the very essence of the right, as it provided for the total extinction of the right to conscientious objection after the passage of a certain amount of time. As another example, consider a 1999 decision by the Polish Constitutional Tribunal on the right of ownership.66 One aspect of the decision was the question whether a particular regulation of the law regulating construction (which authorised the state to demolish those buildings that were constructed without the appropriate permit) affected the essence of the right to ownership. In explaining the constitutional prohibition on affecting the “essence” of a right, the Tribunal held that it necessitates the identification of a certain inalienable core of the right in question, which should be free of legislative interference even if such interference was motivated by the promotion of explicitly mandated constitutional values. The Tribunal attempted to spell out what would actually amount to an infringement of the essence of a right, but, reading the Court’s opinion, it is difficult to escape the impression that it simply kept accumulating various synonyms of the word. It stated, for example, that such an infringement would take place if the legislative restrictions concerned the fundamental entitlements that, taken together, establish the substance of a given right, or if they prevented the right from performing its functions as envisaged by a legal system based on the principles established by the Consti-
278
CHAPTER 10
tution.67 This is very much an idem per idem explanation. On the basis of this explanation, the Tribunal concluded that, in this particular case, the core of the right to ownership was not affected by the regulation under challenge: the demolition order does not deprive the owner of the right to the real estate upon which the building has been constructed, or of the right to apply for a building permit in the future. The Tribunal added that the essence of the right would have been affected if, for example, the requirements for obtaining building permits were so rigorous as to actually render it impossible to undertake construction legally. Another requirement, that restrictions of rights can only be established by statute, thus excluding the possibility of restricting constitutional rights by executive action, is widely present in CEE constitutions68 and has served as an important basis for many constitutional court decisions. The Constitutional Court in Croatia invalidated, in 1998, a provision of the Law on Railroads that gave the Minister of Transport the right to determine whether the workers had to continue working during a strike; the Court found this provision to be contrary to the principle that restrictions on the right to strike must be specified by statute.69 The Constitution (article 60(2)) states that “The right to strike may be restricted in the armed forces, the police, government administration, and the public services specified by law”. The Constitutional Court interpreted this to mean that it is the statute that must be the instrument used to limit the constitutional right to strike. Indeed, the statute must also determine who else is must be involved in the decision-making process concerning a strike (e.g. trade unions) and also what the legal remedy against a Minister’s decision is. The Court further noted that restrictions by law cannot lead to a total prohibition on the right to strike. The Croatian Court took similar decisions with regard to analogous provisions on strikes in postal and telecommunications services70 and in electrical enterprises.71 The same Court decided, in 1999, that a provision of the Law on Public Assembly, which allowed local self-government bodies to designate places in which public meetings could be held, was unconstitutional, on the grounds that only the national legislator could restrict the constitutional right to public assembly.72 Otherwise, local self-government bodies could designate inconvenient places as being the ones at which public assembly could occur. This, in effect, would restrict the constitutional right to public assembly. The Constitutional Court accepted that the legislator may restrict the places allowed for such meetings; this, however, must be done by means of a law containing clear and objective criteria for the determination of places where public assemblies may not be held, having in mind the protection of the rights and freedoms of others, public order, morals and health. The legislator may not delegate such competencies which, according to the Constitution, must be regulated by national laws. The Constitutional Court of Lithuania invalidated a number of statutory provisions on the basis that they would allow for restrictions of constitutional rights by legal means other than statutes. The Court struck down a law on restoration of property that would have allowed the government to establish conditions for the restoration of the ownership of land;73 a law on state secrets that would allow the government to approve the list of state secrets,74 and certain laws on tobacco and alcohol control.75 In this last decision, the statute in question banned tobacco and
RESTRICTIONS OF RIGHTS
279
alcohol advertising, as well as advertising of certain products only indirectly linked to alcohol and tobacco. As the criteria for determining the list of such products were to be established by the government, the Court found that such a restriction on the right to free dissemination of information would not result from a statute but from an administrative decision; hence, the provision allowing for such a possibility was unconstitutional. Another example from Lithuania is provided by the Constitutional Court’s decision of 4 March 1999 on the “lustration” law of 16 July 1998, which provided for the restrictions on employment of former KGB employees.76 They were denied the right to be employed in (and if already employed there, were to be dismissed from) virtually all State institutions, as well as a number of different types of private institutions (e.g. in state and credit unions, security services, or as private lawyers or notaries, etc.) for ten years. The Court upheld the most important provisions of the law, arguing, among other things, that a restriction in employment for an ex-KGB official did not constitute a criminal punishment and therefore that the charge that only a court was competent to decide thereon, and also that the presumption of innocence should be observed, did not apply to this particular measure.77 However, the Court invalidated a provision of the statute (Article 3, Part 2), which provided for a procedure allowing a suspension of this lustration measure towards some of the ex-KGB employees under certain conditions: in particular, if they revealed all of the information about their former links with the KGB. In such a case, the law designed a procedure whereby the Centre for Research into the Genocide of the People and Resistance of Lithuania and the State Security Department would jointly adopt a recommendation for the suspension of lustration measures towards a particular person. Such recommendation was then to be considered, and a decision taken, by a three-person commission appointed by the President of the Republic; its decisions would be subject to confirmation by the President himself. The Constitutional Court found this procedure unconstitutional, because it effectively allowed the President “to form a commission which could decide whether to apply the restrictions to the right to choose an occupation” while “[t]he Constitution . . . does not provide that the President of the Republic may decide the questions of restriction of human rights and freedoms. . . .”.78 This was in contravention of the constitutional requirement that the issue of restrictions of constitutional rights must be determined by statutes. This type of general clause also been used, on occasion, against vague restrictions on rights, on the basis that vagueness in fact empowers an administrative agency to establish the actual content of the statutory restriction. The Lithuanian Constitutional Court considered, and invalidated, a provision of the Law on Officials that would restrict the right of public officials to criticise their employers.79 What the Court found particularly problematic in this statute was that it was formulated in such a vague way that the precise definition of criticism and disagreement were not given (no distinction was even made between constructive and destructive criticism). This contradicts the balancing requirement that has to be present in any restriction of a fundamental right (such as that of freedom of expression, or the right to criticise). This lack of clarity meant that the administrative body in question, when applying the norm, established the actual content of the norm itself - and this was particularly problematic given that the sanction was imperative, and harsh (resignation or
280
CHAPTER 10
dismissal). As the Court stated: “Such imprecise legal regulation creates preconditions for such cases when an administrative body, in applying this norm, establishes the content of the norm by itself”.80 Some constitutions state that limitations on rights and freedoms must be nondiscriminatory, and therefore must “apply in the same way to all cases that meet the specified condition”.81 This, one might think at first glance, is a redundant proviso, given that all of the constitutions have general prohibitions on discrimination anyway, and any statutory regulation that could be characterised as restricting a constitutional right in a discriminatory fashion violates ipso facto the constitutional ban on discrimination. The difference is that the general constitutional prohibitions on discrimination in these constitutions usually list a number of banned grounds of discrimination, that is, list a number of properties of individuals that cannot be used as grounds for invidious classifications.82 By contrast, the provision now under discussion is open-ended: any improper classification may be found discriminatory when figuring in restrictions on constitutionally guaranteed rights. As an example of a decision implementing such a rule, consider the decision of the Czechoslovak Constitutional Court of 26 November 1992, concerning those who collaborated with the old regime.83 The “lustration” law of 1991 established that those who had collaborated clandestinely with the old regime may not hold certain positions in governmental bodies and organisations. A group of 99 members of the Federal Assembly challenged this provision. The Constitutional Court found this law to be, in principle, valid: it declared that, in the light of the violations of human rights by a totalitarian regime, the state had the right to apply legal measures to protect democracy and avert the risk of a relapse into totalitarianism. There was a danger, according to the Court, that some people would try to place themselves in positions of power in the new democratic state solely in order to maintain the power of the previous ruling section of society. Further, the Court pronounced that, in establishing the criteria for filling certain top offices and administrative positions, the state should take into account “its own safety, the safety of its citizens, and . . . further democratic developments”.84 The legal restrictions under challenge were thus seen as reasonable, on the basis that they protected the constitutional system. However, certain details of the statute were found to be unconstitutional. One provision included restrictions on persons who were classified as “confidential affiliates”, “candidates for clandestine collaboration”, or “clandestine collaborators in confidential contact”. The details of what this category of persons did were never kept on file (for secrecy reasons) and many of them were merely being sounded out to find out whether they would be willing to provide information. The Court found that they may have had their names in the files without any written commitment on their part, and even without their knowledge. For this reason, it was impossible to ascertain reliably whether those concerned were actually conscious collaborators. The situation of these people was different from the situation of those to whom other articles of the statute applied, who had knowingly, and unquestionably, collaborated with the former regime’s security services. For these reasons, the Constitutional Court found this provision contrary to Article 4 (3) of the Charter on Fundamental Rights, which states that “any statutory limitation upon the fundamental rights and
RESTRICTIONS OF RIGHTS
281
basic freedoms must apply in the same way to all cases which meet the specified conditions”. Finally, one last test that often must be passed before a legislative action regulating the limits of constitutional rights will be found valid is that such rights must not be restricted because of the lawful exercise of another right.85 The Slovak Constitutional Court has established that the exercise of a (constitutional) right to conscientious objection must not result in the denial of the (sub-constitutional) right to obtain a gun license.86 Certain provisions of the 1995 Law on Weapons were challenged by a group of MPs: these articles stated that persons applying for a gun licence are obliged to give evidence that they have not refused to perform military service. If they later refuse to perform military service, the gun licence can be revoked. The Constitutional Court noted that the Constitution gives the right to refuse military service;87 the Constitution also proclaims that “No person shall suffer injury on his or her rights just because of exercising his or her fundamental right or freedom”.88 The law in question here provided that a person could only receive a gun licence if they had refused to exercise their constitutional right not to perform military service; the Court, however, held that it was constitutionally impermissible to exclude a person from obtaining some right only on the basis of the previous exercise of some other fundamental right or freedom by that same person. Furthermore, the Bulgarian Constitutional Court has established that the right to social security cannot be affected by one’s exercise of the right to work.89 In this case, two articles of the Pension Act, which stated that the pension entitlement would be withdrawn from all pensioners who had an earned income, were challenged. The Constitutional Court held that this was contrary to the right to social security, which is a totally separate right from the right to work. Thus, the right to social security cannot be made dependent on whether the right to work is being exercised or not. This is so even although retirement legislation may affect the labour market, as such facts have no direct Constitutional relevance. For these reasons, the section of the law challenged was held to be unconstitutional. As can be seen, the constitutional principle at work here is substantially equivalent to the U.S. doctrine of “unconstitutional conditions” which states that the government must not make the enjoyment of a benefit conditional on the sacrifice of a constitutional right. However, the doctrine itself has been eroded lately in the United States,90 largely on the basis of the ambiguity of the distinction between penalty and non-subsidy: it is one (relatively uncontroversial) thing to say that the government must not penalise a person for an exercise of her or his right, and another (more controversial) to say that a government must not withhold a subsidy from those who exercise their specific rights that may be relevant to the subsidy in question. The difference between these two situations is well illustrated by two leading cases in CEE (described immediately above) on the doctrine of nonrestriction of one right due to an exercise of another. If we translate the constitutional language of CEE into that of the United States, we may say that the first decision (the Slovak one) illustrates the imposition of a penalty for the exercise of one's constitutional right, while the second was an example of the withdrawal of an economic subsidy (pension) in the case of an exercise of a right (earning salary from work) substantially related to the benefit in question. The reason the latter case
282
CHAPTER 10
is less controversial has to do with the scarcity of the object of any given benefit; in circumstances of scarcity, some non-arbitrary criteria of elimination of beneficiaries must be chosen. In abstract terms, it certainly sounds objectionable to suggest that these criteria may include the exercise of a constitutional right; and yet, where the public sector is as large as it is in post-communist states, and where the range of constitutional rights is very broad, the possibility of conflict between a legislative measure and a constitutional right grows enormously whenever the government wishes to regulate access to socially scarce goods. As Kathleen Sullivan wrote, “The larger the public sector in relation to the private, the more conditions on benefits tend toward equivalence with regulation”.91 4. CONCLUDING REMARKS ON RIGHTS LIMITATIONS What conclusions can be drawn from our classification of constitutional approaches to statutory restrictions of constitutional rights, in section 1 of this chapter? Offhand, a purely textual analysis might suggest that constitutions in the first category (those with specific grounds for restrictions attached to particular rights, with no general clause on restrictions) and in the second category (where the general clause on restrictions applies only to those particular rights that explicitly allow for it) are preferable, in terms of protection of citizens’ rights, to constitutions in the third category (that contain a general clause specifying the grounds for restrictions that apply to all constitutional rights). This is because legislative discretion – seen usually as an important threat to individual rights, ceteris paribus – is more restricted in the first two types of constitutions than in the third one.92 It may be claimed that, when a legislator has a constitutional mandate to restrict any constitutional rights under the general and (of necessity) vaguely formulated grounds for restrictions, the entrenchment value of constitutional rights is largely illusory, creating merely a set of general guidelines, subject to further legislative elaboration based solely on the legislature’s views concerning the requirements of national security, public morals, etc. Moreover, it can be claimed that, ceteris paribus, constitutions in the first category offer superior protection of citizens’ rights than those in the second because, in the former, the list of grounds upon which those rights can be statutorily restricted is narrowly tailored to the specific subject-matter of a given right, thus limiting the risk of arbitrary and excessive restrictions thereon. Not surprisingly, the model of constitutional design of rights restrictions that corresponds to our third category has often been deplored by liberal-minded scholars. As an example, the Polish constitution has been criticised in this respect, as containing “a mechanism of limits on the application of human rights … [that] is misguided, defective and [able to] lead to deep uncertainties and even to a threat to . . . the effectiveness of constitutionally protected rights”.93 The author, Tadeusz Jasudowicz, traces these “defects” largely to the fact that a single general list of grounds for constitutionally permissible restrictions applies to all the rights. As one specific negative illustration of this approach, he cites the protection of environment as one of the goals that can legitimate rights restrictions. While recognising the great importance of this aim, he expresses strong reservations as to the listing of that goal
RESTRICTIONS OF RIGHTS
283
as a “justification for limiting each and every constitutional human right”, and asks rhetorically: “What if there is a conflict between a human right to life and an environmental need – should the life be defeated?”94 Strangely enough, our a priori speculation that the third model leads to the least rights-protective constitutional regime is not in fact confirmed by the constitutional reality of post-communist democracies in CEE. There is no discernible correlation between the typology of constitutions offered here and the level of rights protection under different constitutional regimes of allowing for legislative limitations of rights. The first and second categories of constitutions are found both in those states that have a reasonably high level of legislative protection of rights (Slovenia and Lithuania) and also in those that have a much less enviable legislative record (Serbia, Georgia). The same can be said about the third category, in which we find countries of widely differing standards of legislative rights protection. Overall, they are not significantly inferior to the standards in countries belonging to the first category. There are two explanations of this apparent anomaly. One is that the correlation would occur only if all other things (other, that is, than the constitutional design of permissible limitations on rights) were equal; obviously, however, they are not, and certain other factors are more significant than the structure of the constitution. These other factors include both the formal institutional mechanisms (the powers of and modes of activating the constitutional court) and the legal and political culture of the community in question. What is clear is that this particular constitutional variable (the design of permissible statutory restrictions on constitutional rights) turns out to be relatively insignificant in affecting the shape of the system of legislative protection of rights. The second explanation is more complex. Consider the first category of constitutions again: certain particular rights are accompanied by clauses concerning statutory limitations thereto; others not. It might appear that those rights that are framed as “absolute” must never be subject to statutory limitations. But this is obviously not the case; these rights are not absolute, any more than are those contained in the United States Bill of Rights (which, after all, do not have any grounds for permissible restrictions spelled out in the constitutional text). Judicial scrutiny of a statutory limitation on a right that is formulated in an absolutist manner cannot appeal to constitutional grounds for restrictions, but it does not follow that such scrutiny can never be undertaken, or that it can never conclude with the upholding of the statutory limitation in question. For instance, statutory restrictions on speech may be interpreted as not restricting the right to freedom of speech because the term “freedom” can be construed in a narrower sense than a license to speak what one wishes without any restraints. The more general an “unrestrictable” right is, the more obvious it becomes that it will need to be subject to some sort of restriction, and that reasonable people will disagree as to whether a particular regulation constitutes a restriction in the first place. Consider the list of rights that are solemnly declared to be beyond any restrictions whatsoever in the Russian Federation.95 On one hand, there are some specific rights that can be clearly declared unrestrictable, and in the case of which it is reasonably uncontroversial to establish whether a particular regulation constitutes a restriction; for example, the right to
284
CHAPTER 10
freedom from torture96 or the right to legal counsel.97 But some other rights, also covered by the “no restrictions” clause, raise some fundamental disagreement over what actually constitutes their restriction: consider the right to “the dignity of the person”,98 “the right to privacy, to personal and family secrets, and to protection of one’s honour and good name”,99 or the right to freedom of conscience and religion.100 It is simply unthinkable to consider these rights unrestrictable, unless, that is, we conceptualise them in a narrow fashion from the outset, so that certain exercises of the right to freedom of religion will be seen as falling outside the constitutionally guaranteed freedom of religion, narrowly construed. The “absoluteness” of these rights is made possible only by the narrowness of their scope.101 However, as reasonable people will disagree over the reasonable scope of any such right, the characterisation of a particular regulation as “restriction” will be eminently controversial in itself. It does not follow that the choice of formulation of statutory limits on rights opted for in a constitution is irrelevant. The real significance, however, lies not in the degree of “unrestrictability” of a right, but rather in the modes of argumentation regarding the constitutionality (or otherwise) of a statutory limitation on a constitutional right. The reasoning of legislators, and/or of constitutional courts, is more structured by constitution-makers when any limitation of a right must be matched to a standard provided by the constitution itself, such as “public security”, public health, etc, with additional requirements of proportionality, nondiscrimination, non-infringement of the “essence”, etc. If a constitution does not supply these yardsticks, much depends on the powers and characteristics of individual constitutional courts: when the constitutional court is weak and deferential, legislators have much more discretion in deciding on what restrictions on rights are appropriate than in the system where these standards are constitutionally determined. Where constitutional courts are activist and robust, the construction of restrictions clauses in the first two categories empowers the courts to a higher degree than is the case of the third: namely, in the case of rights that may seem constitutionally “unrestrictable” they are not strictly bound by constitutional guidelines as to how to reason about the plausibility of statutory restrictions on rights. Either way, the ultimate strength of protection of rights is only indirectly affected by the model chosen to approach the issue of statutory limits on constitutional rights. 5. POSTSCRIPT ON RIGHTS AND DUTIES A survey of post communist constitutions in CEE countries indicates that all but three102 of the constitutions of the region contain provisions placing affirmative duties upon citizens; something that is not unusual against the background of the European constitutional tradition.103 Whether citizens’ duties properly belong in a constitution depends upon a broader conception of what a constitution should be. If the main function of a constitution is seen as placing limits on the exercise of state powers, then the inclusion of citizens’ duties in its text is illogical and may even be harmful. If, however, the constitution is seen as a statement of the paramount moral
RESTRICTIONS OF RIGHTS
285
and political principles upon which the political system is built, and which are supposed to be subsequently articulated and concretised in ordinary laws, then citizens’ duties should be spelled out in the constitution alongside citizens’ rights, even if only to counteract what many deem an excessive emphasis on rights to the detriment of responsibilities in liberal democracies. Different answers to the question of the constitutional status of citizens’ duties can, therefore, be seen as resulting from different philosophies regarding the role of a constitution. These conclusions also result from different conceptions of the relationship between constitutions and sub-constitutional laws. If the relationship is seen as one of continuity between general and concrete legal provisions, then the constitution should spell out the bases for legislative regulation of individual duties. If, however, the relationship is seen as a tension between an act that constrains the legislative will (constitution) and that which gives expression to that will (ordinary legislation), then constitutional duties are redundant. By far the most frequently articulated constitutional duty is that of care for the quality of the environment (in sixteen constitutions of the region),104 closely followed by the duties of parents to nurture and educate their children (in fourteen constitutions),105 and the duty to perform military service (in thirteen constitutions,106 although in six of these this duty is modified by the right to conscientious objection107). The next most popular duty is the citizens’ obligation to contribute to public expenditures through the payment of taxes (twelve constitutions).108 In eight of the constitutions of the region there exists a child’s duty of care towards their old and infirm parents.109 Less common duties include assistance to the state and society in cases of natural or other disasters (three constitutions),110 and – perhaps most worryingly – a general duty of loyalty, either to the “constitutional system of government,”111 or to the state in general.112 Finally, in Bulgaria there is the unique, and ominous, duty of all to learn and use the official language.113 This requirement necessarily raises the issue of minority members’ rights. As this brief survey indicates, the duties included in post communist constitutions may be divided into two separate categories: (1) those that are capable of being further concretised into specific legal rules, which then become the actual basis for enforcement of specific obligations, and (2) those that are merely symbolic in character and that thus do not ground any specific, legally enforceable obligations. Regarding the first category, the duties to care for the environment (or, most specifically, not to pollute and damage the environment in any way), to perform military services, or to pay taxes are obvious examples of the kinds of obligations that a state can legitimately impose upon an individual. The constitutional duty per se, however, cannot be the grounds for the imposition of a specific obligation on citizens; rather any such obligation must be grounded on a clear statutory regulation. As such, the constitutional duty can be seen as authorising the legislator to issue specific statutory regulations in order to “operationalise” the duties. For example, no one can be forcibly enlisted in the army if there is nothing more than the constitutional statement requiring such a duty; a further, specific legislative act is required. Then, however, the usefulness of a statutory regulation may be put in doubt, because the implication may be that the legislator is empowered to impose upon individuals only those duties as are generally recognised in the constitution.
286
CHAPTER 10
This, clearly, is not the case. Consider, as an example, that a duty to pass on to the authorities any information pertaining to crimes committed exists in all criminal codes of the region, even though such a duty is not explicitly grounded in any constitution. As for the second category of constitutional duties, of which the duty of loyalty to the state is an example, they clearly have only an expressive value and are not enforceable in any way which can be operationalised by statute. They are, as some critics of these constitutional provisions observe, “legally meaningless”.114 For these reasons, the constitutional provisions on citizens’ duties have only been extremely rarely subject to interpretation by constitutional courts. One such exceptional case arose when the Polish Constitutional Tribunal was asked to consider the constitutionality of the so-called “lustration” law; specifically, the statute of 11 April 1997, which provided for sanctions for making false declarations about one's work for, or collaboration with, the security services under Communism.115 One of the many grounds of alleged unconstitutionality (admittedly, a relatively marginal one in the context of this important, long, and multi-faceted decision triggered by a group of left-wing MPs) was the constitutional duty to defend the country. More specifically, the Constitution proclaims duties of fidelity to, and defence of, the country,116 while the lustration statute used the concept of collaboration with operational or investigative units of the secret services in an undifferentiated way (according to the challengers), thus including in the concept also the implementation of constitutional duties. According to the petitioners, collaboration motivated by “constitutional civic duties” should be exempted from the definition of collaboration for the purposes of the lustration law.117 The Constitutional Tribunal rejected this objection by pointing out that the lustration law explicitly exempted from the notion of collaboration those actions that were done in fulfilment of legal duties under the law valid at the time, such as participation in compulsory military training, etc. No duties of work for, or collaboration with, security agencies stem from the constitutional duties of fidelity to the Republic.118 As one can see, the Tribunal – correctly, it should be added – found no operative meaning in the constitutional duty provisions, in that particular context at least. One can perhaps go a step further, and claim not merely that constitutional provisions on duties are innocuous redundancies, but that they can be harmful. Firstly, the provisions on duties diminish the overall civil-libertarian flavour of the constitution and introduce a statist, authority-respectful rhetoric. The message is that, while the state has some duties toward its citizens, the citizens have duties not just to one another, but also to the state. This message emerges, for instance, from article 82 of the Polish Constitution: “Loyalty to the Republic of Poland, as well as concern for the common good, is the duty of every Polish citizen”. Secondly, and more importantly, the inclusion of duties alongside rights might be seen as implying that the enjoyment of one’s rights is conditional upon the performance of one’s duties.119 This has never been explicitly stated in this manner, and none of the post communist CEE constitutions have adopted the old, pernicious principle of Soviet constitutionalism that “the exercise of rights and freedom is inseparable from the performance by the citizens of duties”.120 However, if one considers that there is no discernible good reason for having duties in the constitution, this dependence of rights enjoyment upon the fulfilment of duties may be one way of interpreting the
RESTRICTIONS OF RIGHTS
287
reason behind the inclusion of the latter in the first place. But such an implication is, of course, anathema in a liberal theory of citizens’ rights, according to which the existence and applicability of fundamental rights to all cannot be made conditional on the fulfilment of duties, constitutionally proclaimed or otherwise. Strictly understood, however, this suggested connection of rights and duties implies only that one cannot effectively enjoy one’s rights if others do not perform their correlative duties; it does not amount to the normative position that one must be denied the enjoyment of one’s rights as a result of the non-performance of one's duties (including the duty to respect other people’s rights). Whether such an implication is permitted by any given constitution is a matter of speculation. To some extent, this question depends upon the structure of the constitutional text: if citizens’ duties are grouped in a separate subdivision, as is the case in only three constitutions,121 then the thesis about rights’ enjoyment being contingent upon performance of duties is more difficult to sustain than when duties are spelled out together with specific rights. More importantly, the very nature of the rights and duties in question may render such an implication more or less plausible. For example, if parental duties and rights are closely connected in a constitutional text,122 then it is plausible to argue that such a connection implies that parents who are grossly negligent toward their children surrender the right to decide about their children’s education and upbringing. In this situation, the non-performance of a duty might result in a constitutional right being extinguished. A contrasting example is provided by rights and duties concerning a clean and healthy environment:123 the exclusion of a negligent citizen from the benefit of a public good such as clean air is virtually impossible. There is, however, a third category of cases in which the benefit protected by a right is not a public good and, therefore, exclusion is physically possible. In such instances, exclusion would be grossly unfair even if a person had not fulfilled his corresponding duties. For example, should a citizen who has not fully discharged her duties to “assist the state and society in the case of natural or other disaster”124 be denied state assistance when she falls victim to just such a disaster? Again, it is not suggested that the very fact of constitutionalising citizens’ duties implies that the drafters intended to make the enjoyment of rights contingent upon the fulfilment of duties. The fact, however, that such an interpretation is plausible, at least with regard to some pairs of rights and duties, may be seen as a cost that is not offset by any clear benefit gained from constitutional proclamations of citizens’ duties.
CONCLUSIONS Now that we have completed our overview of the ways in which constitutional courts in the post-communist states of Central and Eastern Europe understand and articulate constitutional rights, and impose, at times, their own understandings upon the political branches of government, can we concur in the magnificent praises that have been lavished upon them by outside observers and acknowledge that they have been “the flagships of the rule of law and constitutional faith in the emergent Eastern European democracies”.1 Can we conclude that they indeed became “courageous and vigorous defenders of constitutional principles and human rights, continuing the pattern shown elsewhere in the world”?2 No doubt, much will depend upon one’s assessment of the “pattern shown elsewhere in the world”; as already mentioned in the Preface, the scholarly support for systems of robust judicial review “elsewhere in the world” is far from unanimous. Important doubts and criticisms have been expressed, and the fact that constitutional judicial review in CEE has not thus far been the target of these may in itself be part of the problem: institutions that are insulated from principled critique are in danger of developing a culture of self-importance and conceit. This may be all the more troubling because, as our survey in Chapters 6-10 has shown, the record of constitutional courts in the region (including the most activist, independent and powerful of the courts) is far from unambiguously positive. While there is no doubt that they have improved a large number of statutes and invalidated numerous laws that were deeply objectionable from the point of view of liberal-democratic standards of individual rights, there have also been several more worrying cases: of missed opportunities to strike down objectionable laws, and even of invalidating certain rights-protective laws, and thus pushing the legal regime of their states in a less liberal direction. There have been also several examples involving legal regulations that were the subject of reasonable disagreement amongst people in good faith, where the courts replaced the choices of the political branches of government with their own choices and preferences. Undoubtedly, these negative categories (of missed opportunities and objectionable invalidations) are, on balance, less numerous and less weighty than the interventions of constitutional courts to remove excessive restrictions on and departures from constitutional rights, expansively interpreted. But this is not the end of the story; the calculus does not finish there. If the existence of robust constitutional review were cost-free, and any positive interventions by the courts could be seen as a “bonus”, then the enthusiasm of strongest supporters of those courts would be fully justified. However, as suggested in Chapter 5, the calculus has to be more nuanced, and thus the enthusiasm more qualified. As the existence of the constitutional courts is not cost-free, and the legitimacy of those courts in displacing the choices of the political branches of government is largely seen as resting upon their record in the protection of constitutional rights, it is perhaps proper to subject these institutions to a more severe scrutiny than has thus far been the case both within and outside the states in which they have been set up. In particular, it is important to go beyond the conclusion that, on balance, constitutional courts have saved the post-communist citizens from dangerous and oppressive laws – even although such a conclusion would be justified – and raise the question of the
290
CONCLUSIONS
connections between political legitimacy, institutional competence, and possible infringements of the political rights of citizens. These three dimensions are obviously implicated whenever the last word on issues of rights protection or policysetting are placed in the hands of a body that is not accountable to the electorate in the way parliaments (and governments controlled by the parliaments) are. At the very least, the positive contribution of these courts has to be compared to the negative consequences of elevating them to such a prominent role in the political system. In Chapter 5, I identified two possible negative consequences in the long run for establishing a system of judicial review of statutes in terms of constitutional rights: a possible societal apathy towards rights articulations, and the potential for legislative negligence towards rights. These two phenomena have to be taken seriously. Mark Tushnet has coined the concept of “democratic debilitation”, which he takes to mean that “the public and their democratically elected representatives cease to formulate and discuss constitutional norms, instead relying on the courts to address constitutional problems.”3 “Democratic debilitation” may lead to depoliticisation, apathy and withdrawal of the general public from the public discourse on policy proposals and law reform, and to the conversion of fundamental moral-political choices and dilemmas into the juridical language of the true meaning of the constitution. Here is how Jack Balkin describes the phenomenon (although not in the context of constitutional courts): People who immerse themselves in the Constitution and its traditions often bring the concepts of that tradition to bear in their other moral and political judgments. In this way the tools of constitutional thinking infect our attitudes towards basic questions of social justice and political philosophy. We find that when we discuss these questions, we turn to the language of the Constitution as second nature. It is a language that warps and limits our imagination about justice. 4
Balkin uses the loaded word “infect”; however, the same phenomenon can be described from an opposite value perspective, namely as a necessary and desirable consequence of the attempt to insulate the protection of those fundamental interests that are constitutionally identified from the realm of everyday politics, the latter characterised, as it often is, by populism, demagogy and intolerance. Indeed, it has been the positive perspective that has dominated the discourse both by, and about, the constitutional courts of post-communist states. Probably the most famous of all CEE constitutional judges, the chief justice of the most activist of these courts, László Sólyom, announced at the height of the most dynamic period of the life of his Courts: “The existence of the Constitutional Court during the transition … allowed the transformation of political problems into legal questions that could be addressed with final, binding decisions”.5 This, Sólyom took to be a matter of pride rather than embarrassment. However, “transforming political problems into legal questions” has its price, both in terms of the depth of societal discourse concerning those “political problems” once they have been framed in terms of the “true meaning” of complex constitutional texts, and of the institutional implications of such a translation. Significantly, these words of Judge Sólyom were quoted by the American legal scholar Ruti Teitel in the context of her discussion of the famous Zetenyi judgment of the Hungarian Constitutional Court, described here
CONCLUSIONS
291
in chapter 9; an important decision in which the Hungarian Court asserted for itself a central role in defining the terms of transition to a democratic system and, more importantly, the degree of continuity with the Communist system. Teitel observed: “The Zetenyi case stands for the proposition that the authority to asses the legality of the prior regime does not lie with Parliament, but instead with the Constitutional Court”, and she further hypothesised that (under some interpretations) this decision may have been “less about the rule of law than about institutional distrust”.6 As is well known, such “institutional distrust” has largely been justified: the performance of most of the parliaments in CEE has been an unwholesome display of demagoguery, intellectual incompetence and corruption. Of course, the reality was much more complex than this, and to what extent these characteristics were encouraged rather than countered by strong judicial review is a question that must be asked. There is a very real danger of producing pathologies in the incentive structures for the other political actors: when the legislative process occurs “in the shadow of judicial review”, there is a high likelihood that legislative irresponsibility will result. Kim Lane Scheppele, in her account of a series of decisions by the Hungarian Constitutional Court in 1995, which struck down the laws upon which the government’s austerity program relied (the Bokros package, discussed in Chapter 7), gives the example of one particular provision of this legislative package that raised the suspicion “that the Parliament . . . very much wanted the Constitutional Court to strike down this law”.7 This, in Scheppele’s account, is an isolated and non-typical episode; nevertheless it points to a much more serious threat to the integrity of the legislative process than many enthusiasts of constitutional judicial review are willing to concede,8 namely, the encouragement of legislative irresponsibility. This occurs when the very awareness of possible review makes legislators less attentive than they should be to constitutional rights. The result can be legislative apathy in the implementation of constitutional rights (along the lines of: “if something is wrong, the court will remind us of it”). An opposite, but equally plausible, result of judicial finality in the articulation of rights is legislative inaction: not enacting certain laws on the basis that the court is likely to invalidate them (“we would like to legislate but the court will not allow us ”).9 Either way, whether irresponsibility or apathy, pathology is introduced into the decision-making process of the legislature where the existence of a superior body (rather than the existence of constitutional norms) causes distortions in legislative choices. This is not to say that legislatures are necessarily made unhappy by these distortions; on the contrary, they may enjoy the fact that they have an excuse for doing (or not doing) something that might be politically (un)popular, and that they find, in constitutional courts, a convenient scapegoat. This is a phenomenon that has been observed by the students of Western judicial review. Keith Whittington notes that in the United Stated, where the link, in the eyes of the voters, between a parliamentary action and judicial conduct is very remote and indirect, the politicians may use the Supreme Court as a vehicle of achieving unpopular, or politically costly, results without risking political losses themselves: “Elected officials have an incentive to bolster judicial authority not only to encourage the judiciary to take independent action but also to weaken the voter’s ability to trace responsibility back to elected officials”.10 The example that Whittington gives is of civil rights for
292
CONCLUSIONS
African-Americans under the Warren Court: the Southern Democrats could denounce those decisions and claim credit in the eyes of the electorate, while Northern Democrats could achieve substantive political gains without having to take any action that could cause them electoral losses. As a result, “the Democratic Party was less likely to be electorally punished for decisions on desegregation . . . even though these decisions were consistent with basic party commitments”.11 More generally, Klaus von Beyme notes that constitutional courts, including those in CEE, perform a useful role of “blame-sharing”: “In democratic systems [constitutional courts] perform important functions in the area of “blame-sharing”. Government alone is not responsible for certain unavoidable but unpopular measures which have to be taken in order to consolidate democracy”.12 Such a system of “blame sharing” may not be, in itself, a bad thing; on the contrary, it may be a useful and generally beneficial strategy of taking away some of the populist political pressure from parliaments, if the blame can be attributed to those who do not have to pay the political price for “unpopular” (in the short term) decisions. This presupposes a certain division of labour in an overall system of government. It does not presuppose that the parliaments will be the weak link of the system, and it is not an automatic invitation to irresponsibility or apathy. If, however, the parliaments are weak and unpopular, the “blame sharing” strategy may backfire: it may weaken the legislature even further by taking away both their prestige (because they will be seen as less salient in the system) and their incentives for taking constitutional values seriously. As a Polish political philosopher, Marcin Król, caustically noted recently about the legislature in that country: Although the Parliament is a lawmaking body, therefore a very important one, nevertheless – in Poland – almost all important statutes are challenged either by the President or by the Constitutional Tribunal, and this permits the deputies to do what they like because their ideas will come under constraint.13
Król’s observation cannot be easily discarded; it shows that legislative irresponsibility is perceived as a real problem, and that it is directly linked to the probability that the constitutional court will counter the deputies’ worst legislative excesses. There has been an important trend in Western constitutional theorising that suggests that the creation of “external” scrutiny of constitutionality, in the form of bodies such as constitutional courts, has detrimental effects upon the seriousness with which the parliaments (and, consequently, the people) take constitutional norms; the argument is that there is an inverse relationship between the salience of constitutional considerations in parliamentary deliberations and the existence of external checks of the constitutionality of legislation by other bodies. This has been well expressed by Robert Dahl, who uses the concept of “quasi-guardians” to describe those external scrutinisers: Quasi guardianship may . . . require less self-restraint on the part of the demos and its representatives and more externally imposed restraint by judicial guardians. Over time, the political culture may come to incorporate the expectation that the judicial guardianship can be counted on to fend off violations of fundamental rights, just as greater self-restraint on the part of the demos and its representatives may become a stronger norm in the political cultures of polyarchies without judicial guardianship.14
CONCLUSIONS
293
Whether, and to what degree, such an effect has occurred in CEE is a matter for speculation only, and there are some who feel that the presence (and the activities) of the constitutional courts has actually strengthened the role of constitutional considerations in the lawmakers' motives. For instance, according to the President of the Czech Senate, the existence of the constitutional court “is a kind of conscience for the Senators… that guides their considerations when they vote for a particular law”.15 This is particularly important given that, according to the same politician, the Senators often wilfully ignore constitutional provisions and are prepared to endorse various legislative measures even if they are inconsistent with the Constitution. The existence of the Constitutional Court is, he argues, a factor that strengthens rather than weakens the significance of the constitutional considerations in the legislative process. Which effect in fact prevails, either that of emphasising the importance of constitutional consideration or of emptying the legislators’ minds of constitutional concerns, is something that cannot be determined with any precision. However the fact that the latter effect is present is beyond any reasonable doubt, and it is thus something that has to influence the overall score card for the constitutional courts in the region. Now if this latter effect does occur, and the existence of constitutional courts have a weakening effect upon the prestige, seriousness and the sense of responsibility of the parliaments, then one cannot fail to notice that those arguments in favour of robust judicial review that are based on the weakness and irresponsibility of parliaments put in motion a self-perpetuating mechanism: the presumed defects of parliaments used to justify judicial review in the first place become a self-fulfilling prophecy. True, parliaments in CEE enjoy very little social prestige16 and, as noted above, there is often more than enough reason for the high levels of social disrespect. And one can see why this happened, even quite apart from the – contingently true – low quality of the “political class” in post-communist democracies: the parliaments have carried the odium of the old, sham parliaments of the Communist era, while the constitutional courts have benefited from what Ruti Teitel called “the legitimacy of hope” granted to a new institution.17 However, to draw from the accumulated distrust felt generally towards parliaments a conclusion as to the need for a robust “oversight institution” in the shape of a constitutional court may be likened to finding a remedy that contributes to the illness. As Stephen Holmes and Cass Sunstein predict: “a court-centred democracy is unlikely to last. The greater power and prestige granted to the constitutional court, the more diminished may be the power and prestige of parliament, and the more difficult it may be to create legitimate and accountable authority through elections. . . .”.18 They offer this prediction not in order to fundamentally reject the very idea of constitutional courts in CEE, but rather in order to warn against allowing them to develop into something more than a secondary institution in a democratic political system. This also may be seen as one of the central themes of this present book. The experience with constitutional courts in CEE, and in particular the question of their monopoly on the conclusive articulation of the true meaning of constitutional rights, resonates with a more universal, world-wide questioning of the ramifications of judicial review. The fundamental question raised by judicial review under bills of rights may be formulated as: is judicial supremacy really essential to
294
CONCLUSIONS
“the maintenance of constitutionalism”, or is it the case that all of the main institutional actors, and in particular the parliaments, are responsible for taking the Constitution seriously and providing authoritative articulations of its meaning? As Alec Stone Sweet sensibly remarks in his recent article: It would be a mistake to dismiss parliamentary adjudication of rights as inherently less meaningful or less ‘judicial’ than the deliberations of a constitutional court. Parliament and the court are doing more or less the same thing, speaking in more or less the same language and working through more or less the same normative material.19
If the link between the very principle of constitutionalism and judicial supremacy is questioned, then the institutional problem of how best to guarantee the salience of constitutionalism in political life becomes precisely that: an institutional question, which has to be tackled in a pragmatic way, by studying the various incentives that act upon various institutions and that render them more or less likely to take the constitution seriously. As Keith Whittington argues: “Constitutional maintenance requires a political commitment to constitutional interpretation. Political actors must care what the Constitution means, and be willing to temper their immediate political desires accordingly”.20 The prevailing constitutional doctrine in CEE (as much as everywhere else in the world) portrays this interpretative task as a technical, arcane skill, where legal qualifications are key to the successful discharge of the task. But as soon as we start to view constitutional interpretation as a constructive and creative task of infusing the abstract constitutional clauses with political values in order to render them operative in specific cases, the claim that lawyers enjoy a privileged insight into their “real meaning” is undermined, and is revealed as a grab for power by the legal elites, particularly those represented in, or gravitating towards (as constitutional-law scholarship often does), constitutional courts. If we accept that all branches of government have a duty and the authority to provide binding articulations of constitutional texts (including of constitutional rights), then the issue of the relative powers of the legislature and the judicial (or quasi-judicial) bodies assumes a pragmatic-institutional character rather than a principled one: whether the constitutional courts should exist and should have the power of final decisions even if they do displace the choices of the legislature becomes a matter evaluating the incentive structure that this would produce for the overall system, including all other governmental bodies, in terms of taking the constitutional mandates seriously. As suggested above, it is far from obvious that judicial supremacy and finality must have a positive effect on such incentives. They may, and often do, create the perverse incentives of irresponsibility and/or apathy. Whether or not the latter prevail over the positive, constitution-respecting effects, one thing is clear: robust judicial review should not be based on a theory that assumes a judicial monopoly of access to “correct” constitutional meanings. The institutional theory endorsed here may be analogised to what one already quoted American scholar, Keith Whittington, calls “departmentalism”, that is “the doctrine that each branch of government has the authority to determine constitutional meaning independent of the judgements of the other branches”.21 It rejects the a priori assumption that, when the institutions clash as to the best constitutional interpretation, deference is owed automatically to the constitutional court. Instead, it
CONCLUSIONS
295
postulates that each the institutional political actors must follow their own understandings of the constitution, and calls for the various institutions to carefully take note of each other’s constitutional articulation, without necessarily assigning a privileged position to any of them. It does not reject the principle of constitutional review, because the courts should be able to set aside laws they deem unconstitutional, but it challenges the supreme power of the constitutional courts to define for all other institutions, in a final and authoritative way, what the true meaning of the vague rights clauses contained in the constitutions should be. This theory also encourages the search for institutional innovations that may reduce the force of the clash between the court and the legislature in cases of fundamental disagreement over articulation of rights. Perhaps there is a middle way between strong judicial review, in which the court simply strikes down a law in the process of abstract review at the behest of the parliamentary minority, and the immunisation of the parliament from any control under the rights provisions of the constitution. In a recent article, the American constitutional scholar Stephen Gardbaum has suggested that the emerging “Commonwealth model” of constitutionalism may offer just such a solution. 22 One example of this model is the Canadian system of judicial review under the Charter, in which the legislatures have the power to override the Charter rights (or, in practice, the courts’ interpretation of these rights) – as already discussed in Chapter 3 (3). Another is the New Zealand Bill of Rights Act, which cannot serve as a basis for the invalidation of legislation, but which provides a powerful instrument for judicial intervention in that all other statutes have to be interpreted in a manner consistent with it. Finally, the third example of the “Commonwealth model” suggested by Gardbaum is the United Kingdom Human Rights Act, which not only places on all courts the duty to interpret statutes consistently with the European Convention on Human Rights, but also gives them the power to declare the incompatibility of statutory provisions with that Convention, thus creating a strong expectation that the provision in question will eventually be repealed by the parliament. This expectation is supported by the fact that the parliament has a strong incentive to take such judicial declarations of incompatibility very seriously indeed: if proper remedial legislative action is not taken, then there is a high probability that the European Court of Human Rights would reach the same conclusion as the domestic UK court, thus leading to the considerable political embarrassment of having the United Kingdom lose a case in Strasbourg. All of these systems attempt, in different ways, to reconcile the task of the protection of rights with the requirements of democratic governance, and of transforming a judicial monologue into an inter-institutional dialogue. As Gardbaum puts it, the hybrid model exemplified by these three systems “decouples judicial review from judicial supremacy by empowering legislatures to have the final word”.23 Another American scholar attracted to such a model, Michael J. Perry, coined the term “judicial penultimacy”: a system that would grant the courts an important role in protecting human rights but that would, at the same time, observe the proposition that “with respect to the concrete, contextual meaning of entrenched, indeterminate rights, electorally accountable legislators should have the last word (even if their last word is to accept the court’s judgment)”.24 Whether any of these
296
CONCLUSIONS
institutional innovations are adaptable to the CEE context is a difficult matter to assess; the point, however, is that we must not confine ourselves to a simple and crude dichotomy: either full judicial supremacy on constitutional matters (subject only to constitutional amendment) or total legislative omnipotence on all matters pertaining to rights. But is it legitimate in the first place to draw an analogy between the constitutional courts in transitional states, such as those discussed in this book, and those in more “developed” and “mature” systems, and to simply import the arguments concerning the proper role of the judiciary as developed in those more mature systems? It is sometimes claimed that such an extrapolation is unjustified, and that the relevant differences between the states that have only just emerged from authoritarianism and those in established democracies should lead us to suspend the criticisms of strong judicial review, which are justified in the West. In a recent, wide-ranging article, the Chief Justice of the Israeli Supreme Court, Aharon Barak, stated that judicial “minimalism” may be appropriate in old and established democracies such as the United States, where “the main principles of the constitutional framework have already been established, and the judicial corrective – which assumes the existence of democracy – is limited in its role”; however, in “young and fragile democracies, such as many of the new democracies in Eastern Europe . . . . [m]inimalism is likely to be unsuitable”.25 This is a frequently made point, and it benefits from a degree of commonsensical support: the more democracy is vulnerable, the stronger the need seems to be for extra-political mechanisms (such as judicial review) to protect it from itself and from its enemies. A careful student of post-communist legal transformation, Ruti Teitel, has suggested that the usual doubts concerning the legitimacy of judicial lawmaking simply do not apply in a transitional legal environment: Our ordinary intuitions about the nature and role of adjudication relate to presumptions about the relative competence and capacities of judiciaries and legislatures in ordinary times that simply do not hold in unstable periods. … In periods of political change, the very concerns for legitimacy and democracy that ordinarily constrain activist adjudication may well support such adjudication as an alternative to more politicized uses of the law.26
In contrast to Barak and Teitel, I believe that there are good reasons to resist reliance on the exceptional nature of transitional states as a means of side-stepping the objections that we might raise elsewhere to the institutional anomalies dealt with in this book. For one thing, as the post-communist states of CEE become more mature and stable, so the objections against judicial lawmaking, which are pertinent elsewhere in the democratic world, gain force in this region also. In particular, there is little or no reason to suspend “our ordinary intuitions” about democracy and legitimacy with respect to countries such as Hungary, Poland, the Czech Republic or Slovenia, the very countries in which the constitutional courts are particularly activist in character. In many relevant respects these states fully resemble mature democracies, exhibiting as they do developed and pluralistic party systems, a free and diverse press, well-educated and politically aware electorates and independent judiciaries, etc.
CONCLUSIONS
297
Second, there would be a certain irony in reliance on the exceptional nature of transitional states in CEE to defend the role of the activist constitutional courts of the region, as some of the most activist courts themselves actually refer to the “normalcy” of the democratic systems in which they operate to justify some of their most activist decisions. The rhetoric of transition and extraordinariness is actually strongly resisted by the constitutional courts themselves. As we saw in Chapter 9, in an important decision of December 1994 declaring the “lustration” law (pertaining to the vetting of political figures) unconstitutional on various grounds, the Hungarian Constitutional Court relied partly on the argument that a successful transition to a democratic system had actually occurred without the need for restrictions on employment of the nature proposed; the upshot was that the alleged purpose of the challenged law (namely, to secure a successful transition to democracy) could not apply. The principles to be applied in assessing the lustration provisions were therefore to be those applicable to a democratic state based on the principle of the rule of law. The Court drew a clear contrast between the past and the present, holding that they were separated by “the transition as a historical fact”.27 It thus made it clear that the lawfulness of the “lustration” laws should be judged not by reference to the unusual circumstances of transition, but rather by appeal to the correct balancing of the rights and interests at issue. If anything, some of the courts in CEE use (at least rhetorically) the argument from transitionalism to strengthen the idea of deference to legislatures rather than for the purpose of self-reinforcement. The Constitutional Tribunal of Poland, in emphasising that it conducts the control of legality and not of wisdom of the statutes, has consistently asserted that the legislator should have a broad discretion in deciding on which measures to adopt in order to best achieved the declared aim. This principle, the Tribunal has occasionally noted, is particularly important in periods of major systemic transformations, as transition “from the authoritarian state to the state of law may exceptionally take forms that would be unjustified under normal conditions”.28 Put simply, the constitutional courts themselves either resist the rhetoric of exceptionalism altogether, or, when they do accept it, they use it for purposes opposite to those suggested by the defenders of strong judicial review in non-consolidated democracies. Third, exceptionality arguments resonate dangerously with the view that Central and Eastern European societies are as yet too immature and not ready for democracy,29 and that, rather than aiming at fully-fledged democratic institutions, the post-communist societies of CEE should aspire at no more than “authoritarian political institutions, buttressed by indigenous cultural traditions, [which] seem to offer the best matrix for the emergent civil society”.30 This type of prescription is usually based on the diagnosis that a decisive role in shaping the political life in these societies is played by pre-Communist traditions which are “hardly those of Western liberal democracy”.31 The diagnosis is, however, faulty, and so is the prescription. At least some of these societies had a pre-Communist past not less democratic than the pre-World War II systems of many Western European societies that now have unimpeachable democratic credentials. Moreover, the democratic aspirations of peoples in the countries of the region have been greatly influenced by the near-universal rise of democratic beliefs in recent decades – to the extent that the explanatory power of the pre-Communist past has now only a very limited appli-
298
CONCLUSIONS
cation. However, the greatest concern that I have with regard to exceptionality theories of transitional constitutionalism is that they have the potential to become self-perpetuating truths. By basing our actions and institutions upon them we may actually postpone the advent of “normalcy”, and unnecessarily prolong the suspension of our usual intuitions regarding the principles of the separation of powers and democratic legitimacy, such as those encapsulated in the simple maxim from a classic article by Robert Dahl: “a system in which the policy preferences of minorities prevail over majorities is at odds with the traditional criteria for distinguishing a democracy from other political systems”32. Suppose that we agree that objective social and political factors weaken the argument for parliamentary supremacy in transitional democracies, due to the weak civil society and underdeveloped party political system – two factors among a number of conditions that are usually marshalled in support of the “exceptionality” arguments for a courtdriven democracy.33 By acting on this diagnosis, we contribute to the perpetuation of this state of affairs by further disabling the parliaments, and by entrenching a strong role for constitutional courts. By doing this, we institutionalise those very conditions that led us to doubt the role and legitimacy of parliaments in the first place. For all of these reasons, the temptation of relying on exceptionably, and of exempting the post-communist constitutional justice from the usual democratic insights and critiques applicable elsewhere, has been consistently resisted here. An opposite approach has been recommended: that the very newness of democracy after the decades of authoritarianism requires a strong emphasis on participatory, democratic political mechanisms, which foster rather than marginalise public discourse. A system of strong judicial review of statutes under constitutional rights can have a negative educational effect; it may help to generate the perception that rights discourse in general is an obscure activity reserved for lawyers, and that deliberation on the political values that give rise to specific articulations of rights is something over which neither the population nor its elected representatives have any control. Constitutionalism is weakened rather than enhanced when the understanding emerges amongst civilians that a broad range of issues are decided on “legal” (constitutional) grounds, with qualified lawyers enjoying a privileged understanding of the ”true” meaning of those grounds because “legal craftsmanship is essential to decode the meaning” of the Constitution.34 When, to take an extreme example, the legal permissibility of abortion is decided by a contest between different legal understandings of the concept of Rechtsstaat, rather than by public debate about what the constitutionally proclaimed values of life, privacy, non-discrimination, dignity etc. dictate, then obviously the incentives for the general public to argue in constitutional terms are eroded. Constitutional arguments are then seen as an external constraint upon public discourse (a constraint policed by legal experts sitting on the constitutional court) rather than as an integral part of that discourse. The general public and its representative, the parliament, are then socialised into the rules of the division of labour in which they are expected to argue in terms of their interests and preferences, while the court looks after constitutional standards. The correlative implication is that these other political actors, and the general public, are relieved of a duty to take constitutional concerns into account, and even if they
CONCLUSIONS
299
occasionally do, they may be proved mistaken by a small body composed of constitutional experts. This would be a fiasco rather than a triumph of constitutionalism. The experience with constitutional courts in CEE shows that this side of constitutional maintenance is at least as important as the undeniable gains made in terms of having statutes monitored for their compliance with constitutional rights by specialised constitutional guardians.
ENDNOTES Preface 1
Klaus von Beyme, “The Genesis of Constitutional Review in Parliamentary Systems”, in Christine Landfried, ed., Constitutional Review and Legislation (Nomos: Baden-Baden, 1988): 21-38 at 37. See, e.g., Martin Krygier, “Poland: Life in an Abnormal Country”, National Interest 18 (Winter 1989/90): 55-64. 3 E.g., with respect to the Constitutional Court of the Czech Republic, Pavel Holländer reports: “The scope of the Constitutional Court’s powers, as defined by the Constitution, is not subject of a discussion in legal theory”, “The Role of the Czech Constitutional Court: Application of the Constitution in Case Decisions of Ordinary Courts”, Parker Sch. J.E.Eur. L. 4 (1997): 445-65 at 447. This observation applies very much to most of the constitutional courts in the region. 4 Princeton University Press: Princeton, 1999. 5 Routledge: London, 2003. 6 Washington v. Glucksberg, 521 U.S. 702, 720 (1997) (quoting Collins v. Harher Heights, 503 U.S. 115, 125 (1992)). 7 Judy Fudge, “The Canadian Charter of Rights: Recognition, Redistribution, and the Imperialism of the Courts”, in Tom Campbell, K. D. Ewing & Adam Tomkins, eds, Sceptical Essays on Human Rights (Oxford University Press: Oxford, 2001): 335-58 at 336. 8 Id. at 336. 9 The contrast is not as sharp as this sentence suggests, regarding, for instance, France, where the dramatic extension of the authority of the Constitutional Council was a result of decisions of the Council itself rather than of prior constitutional design. In a watershed decision of 1971, the Council “incorporated” the set of unwritten principles into a constitutional package (known by the doctrine as le bloc de constitutionnalité on the basis of which the constitutional review of laws is conducted). The Council did this by announcing its reliance on, among other things, “les principes fondamentaux reconnus par le lois de la République” (P.F.R.L.R.) [Fundamental Principles of the Laws of the Republic] even though the P.F.R.L.R. were not even mentioned, much less specified, by the Constitution in force (of 1958) at the time. However, the Council gave legal weight to the PFRLR because they were mentioned in the Preamble to the 1946 Constitution, which, in turn, was mentioned by the Preamble to the 1958 Constitution. These principles, the value of which was found to be equal to those of the Declaration of Rights of Man and Citizen of 1789, were said to be discernible (by the Council itself, naturally) in the legislation in place up to the fall of the republican system in France, i.e. up to July 1940. See Georges Burdeau, Francis Hamon & Michel Troper, Droit constitutionnel, vol. 26 (Paris: L.G.D.J. 1999) at 714-15. 10 Herman Schwartz, “The New East European Constitutional Courts”, in A. E. Dick Howard (ed.), Constitution Making in Eastern Europe (Woodrow Wilson Center Press: Washington, D.C., 1993): 163-208 at 165, footnote omitted. 11 Edouard Lambert, Le gouvernement des juges et la lutte contre la législation sociale aux Etats-Unis; l'expérience américaine du contrôle judiciaire de la constitutionalité des lois (Giard: Paris, 1921). 12 Bernard Chantebout, Droit constitutionnel et science politique (A Colin: Paris, 1997, 14th ed.), at 60, quoted by Michel Troper, La théorie du droit, le droit, l’État (Presses Universitaires de France : Paris, 2001) at 237 [The translation of this passage from French is mine – WS]. 13 Martin Shapiro, “The Success of Judicial Review”, in Sally J. Kenney, William M. Reisinger & John C. Reitz (eds), Constitutional Dialogues In Comparative Perspective (Macmillan: London, 1999): 193219 at 214. 2
302
ENDNOTES
14
Bernhard Schlink, “German Constitutional Culture in Transition”, in Michel Rosenfeld (ed.), Constitutionalism, Identity, Difference, and Legitimacy: Theoretical Perspectives (Duke University Press: Durham, 1994): 197-222 at 219. 15 Id. at 220. 16 For example, in Poland, by 1998, out of thirty-four judges of the Constitutional Tribunal, only seven had not been academics; see Schwartz supra note 1 at 261 n. 26. Bulgaria is the exception to this general rule; Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia, 11 May 2001. 17 For empirical support for this proposition, see James L. Gibson, Gregory A. Caldeira & Vanessa A. Baird, “On the Legitimacy of National High Courts”, American Political Science Review 92 (1998): 343-58. 18 George Fletcher, Basic Concepts of Legal Thought (Oxford University Press: New York, 1995) at 25. 19 See, similarly, Troper, supra note 12 at 244-45. 20 See id. at 245. 21 This is notwithstanding some excellent work published recently in English, in particular Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago, 2000), who discusses in great detail constitutional courts in Russia, Poland, Slovakia, Hungary and Bulgaria, and Radoslav Prochàzka, Mission Accomplished: On Founding Constitutional Adjudication in Central Europe (CEU Press: Budapest, 2002), who discusses constitutional courts in Poland, Hungary, Czech Republic and Slovakia. Prochàzka’s volume unfortunately became known to me too late to be considered in any depth in the present book. 22 Art. 6.2 of the Treaty on European Union proclaims, inter alia, that the “Union shall respect fundamental rights . . . as they result from the constitutional traditions common to the Member States ….”. The concept of “constitutional traditions common to the Member States” features prominently in the case law of the European Court of Justice, see generally Alessandro Pizzorusso, Il patrimonio costituzionale europeo (Il Mulino: Bologna, 2002) at 7-12, and also the Preamble to the Charter of Fundamental Rights of the European Union (fifth paragraph). 23 Wojciech Sadurski, “Charter and Enlargement”, European Law Journal 8 (2002): 340-62.
Chapter 1 1
It was actually in 1982 that the constitutional amendment creating the Polish Constitutional Tribunal was passed, but the statute on the Constitutional Tribunal, which established a specific basis for that body, was only enacted in 1985. The Tribunal began its operations in January 1986. For the sake of completeness, mention should also be made of the Czechoslovakian Constitutional Court of the interwar period, although it was a rather feeble institution; see Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago, 2000) at 2930. 2 Darina Malová, “The Role and Experience of the Slovakian Constitutional Court”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 349-72 at 351. 3 Id. at 351. 4 See Wojciech Sokolewicz, “Sąd Konstytucyjny w Rumunii”, in Janusz TrzciĔski, ed., Sądy konstytucyjne w Europie (Wydawnictwa Trybunaáu Konstytucyjnego: Warszawa 1997), vol. 2: 145-74 at 146; see also Renate Weber, “The Romanian Constitutional Court: In Search of Its Own Identity”, in Sadurski, supra note 2: 283-308 at 284-85. 5 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (199194), Sofia, 10 May 2001. 6 Stefan Rozmaryn, quoted in Mirosáaw Granat, “Droga do sądownictwa konstytucyjnego w paĔstwach Europy ĝrodkowej i Wschodniej”, PaĔstwo i Prawo no. 12/2001 (vol. 56): 15-24 at 15. 7 Id., quoted in note 6 at 16. 8 Id. at 17. 9 Id. at 19. 10 Id. at 20. 11 Miroslav Cerar, “Slovenia’s Constitutional Court within the Separation of Powers”, in Sadurski, supra note 2: 213-46 at 213.
ENDNOTES 12
303
Described by JiĜí PĜibáĖ, “Judicial Power vs. Democratic Representation: The Culture of Constitutionalism and Human Rights in the Czech Legal System”, in Sadurski, supra note 2: 373-94 at 374-75. 13 Discussed in Chapter 9 in this book pp. 237-39. 14 See Kataryna Wolczuk, “The Constitutional Court of Ukraine: The Politics of Survival”, in Sadurski, supra note 2: 327-48 at 328-31. 15 See http://www.satv.tiesa.gov.lv/Eng/ievads.htm. 16 For details, see http://ccbh.ba/home/?lang=e. 17 The account in this paragraph is based on Alexander Vashkevich, “The Republic of Belarus: The Road from Past to the Past”, in Andras Sajó, ed., Out of and Into Authoritarian Law (Kluwer Law International: The Hague, 2003): 265-98. 18 The exceptions are Moldova, Montenegro, Serbia, State Union of Serbia and Montenegro, and the Ukraine; in these states, solely abstract control by the Constitutional Court is envisaged. In Latvia, the possibility of concrete review has only recently been established through the constitutional amendments of 30 November 2000, and, to my knowledge, this has not been taken advantage of by the courts as yet. In Bulgaria, concrete review can only be initiated by the two top ‘regular’ courts; the Supreme Court of Cassation and the Supreme Administrative Court; see Art. 150 (2) of the Bulgarian Constitution. A quasi-concrete control, nevertheless departing from the standard form as characterised in the main text, exists in Belarus, where any court encountering a statute deemed by it to be unconstitutional has a duty to decide the case on the basis of the Constitution and then to petition the Supreme Court, which, in turn, has a duty to petition the Constitutional Court regarding the (un)constitutionality of the statute. In the Ukraine, there is also an element of “concrete” control; under Art. 83 of the Statute on the Constitutional Court, when an ordinary judge believes that a law that he must apply is inconsistent with the Constitution, he should address the Supreme Court, which may subsequently lodge a formal challenge to that law with the Constitutional Tribunal. There is, however, no stay of the proceedings granted by the “ordinary” judge, no obligation on the part of the Supreme Court to lodge a challenge, and the subsequent consideration by the Constitutional Court is identical to that conducted in the case of any other abstract review. Nevertheless, in the perception of several judges of the Constitutional Court, this amounts to a form of “concrete” control; interview with Professor Wolodymir Tychyj, Judge of the Constitutional Court of Ukraine, Kiev, 22 November 2002. 19 The exception is Lithuania, where the President can only challenge the constitutionality of Government acts but no other laws (in particular, he cannot challenge the constitutionality of parliamentary statutes). In Estonia and Romania the President’s power to initiate review is limited to only preventative control of statutes, and cannot therefore be used after they enter into force. 20 The exceptions are Croatia (where the government can initiate review of only sub-statutory laws), the Czech Republic (as in Croatia), and Georgia and Estonia (where the government has no role in the process of constitutional review). 21 The exception is Estonia. 22 In Bulgaria, Latvia, Moldova, Poland and Slovakia. 23 In Albania, Latvia, Poland, the Ukraine. In Estonia, the ex post abstract review can be initiated only by the Legal Chancellor, who is responsible for monitoring legal acts in the country from the point of view of their constitutionality. The Legal Chancellor, an office similar to the one in Finland, either calls on the body that passed the (in his view) unconstitutional legislation to rectify it, or he files a petition to the Constitutional Review Chamber for the act in question to be annulled. In addition, Estonian constitutional review can also be initiated by the President (as an ex ante review) or, for concrete review, by ordinary courts. 24 In Belarus, Bulgaria, Latvia, Moldova, Poland, Russia and the Ukraine. 25 In Albania and Poland. 26 In Russia. In Ukraine, the Parliament of the Autonomous Republic of Crimea also has such a power. 27 In Slovenia and Ukraine. 28 In Slovenia. 29 Decision 23/1990 of 31 October 1990. 30 Decision 15/1991 of 13 April 1991. 31 Decision 36/1994 of 24 June 1994.
304 32
ENDNOTES
See Georg Brunner, “Structure and Proceedings of the Hungarian Constitutional Judiciary”, in László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 65-102 at 99 n. 40. 33 Oral remarks by Professor László Sólyom, former Chief Justice of the Hungarian Constitutional Court, Workshop on Constitutional Adjudication in Southern and Western Europe, Fondazione Adriano Olivetti, Rome, 26 March 2002. 34 Gábor Halmai, “The Hungarian Approach to Constitutional Review: The End of Activism? The First Decade of the Hungarian Constitutional Court”, in Sadurski, supra note 2: 189-212 at 193. 35 Id. at 81-82. 36 Interview with Professor Vojtech Cepl, Justice of the Constitutional Court of the Czech Republic, Prague, 22 March 2002. 37 Leszek Garlicki, “Orzecznictwo Trybunaáu Konstytucyjnego w 2000 roku”, Przegląd sądowy (2001:9): 77-105 at 85-86. 38 Leszek Garlicki, “Orzecznictwo Trybunaáu Konstytucyjnego w 1998 roku”, Przegląd sądowy (1999:6): 104-28 at 113. 39 Garlicki, supra note 37 at 86. 40 See Maágorzata Masternak-Kubiak, Ustawa o Trybunale Konstytucyjnym (Wydawnictwa Prawnicze PWN, Warszawa 1998): 48-49. 41 “Trybunaá buduje praworządną Rzeczpospolitą”, Rzeczpospolita, 20th March 2002 at C-2 (remarks by the Chief Justice of the Constitutional Tribunal, Professor Marek Safjan). 42 Id. (remarks by the Chief Justice of the Supreme Court, Prof. Lech Gardocki). 43 Brunner, supra note 32 at 84. He adds, however, that in some exceptional circumstances, constitutional complaint may be the only avenue available to a citizen, see id. at 84. 44 Halmai, supra note 34 at 204. 45 For a description of this case, and the ensuing attempt to find an institutional compromise, see id. at 204-206. 46 Interview with Professor Boris A. Strashun, of the Centre for Analysis of Constitutional Justice at the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 47 Suren Avanesyan, “Constitutional Protection for Human Rights in the Russian Federation”, Journal of East European Law 6 (1999): 437-68 at 462. 48 In Albania, Croatia, the Czech Republic, Hungary, Latvia (since 1 July 2001), Macedonia, Poland, Russia, Slovenia and Slovakia (although, in the latter, the competencies to consider constitutional complaints are restricted only to those matters that do not fall under the powers of other courts). 49 Interview with Prof. Lucian Mihai, President of the Constitutional Court of Romania, Bucharest, 9 March 2001. 50 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia, 11 May 2001. 51 Interview with Judge Pavlo Jevgrafov of the Constitutional Court of the Ukraine, Kiev, 22 November 2002. 52 Apart from Hungary (and in the past, Poland and Russia), such a possibility exists in the statutes on constitutional courts in Albania, Montenegro and Serbia. 53 Decision no. 37/1992 (V.10) AB of 8 June 1992, reprinted in East European Case Reporter of Constitutional Law 2 (1995): 27-35. 54 See Brunner, supra note 32 at 85-6. 55 Zdzisáaw Czeszejko-Sochacki, Leszek Garlicki & Janusz TrzciĔski, Komentarz do ustawy o Trybunale Konstytucyjnym (Wydawnictwo Sejmowe, Warszawa 1999) at 202. 56 For example, in its Decision K 19/96 of 24 February 1997 the Constitutional Tribunal declared: “Deciding on its own initiative has an exceptional character and can take place only in cases of particularly flagrant breaches of the Constitution”, Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1997 (C.H. Beck: Warszawa 1998), item 6: 65-77 at 72. 57 Decision U-I-391/96 of 11 June 1998, translation in http://www.us-rs.si/en/casefr.html, Part B.-III. 58 Constitutional Review Court Procedure Act of 1993, Section 4.3, translation in http://www.nc.ee/ english. 59 Decision No. 20 of 10 May 1996, reprinted in the East European Case Reporter of Constitutional Law 4 (1997): 57-63, see also http://www.nc.ee/english/const/96/4a9601i.html.
ENDNOTES 60
305
Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia, 11 May 2001. 61 Interview with Dr Vladimir G. Yaroslavtsev, Justice of the Constitutional Court of the Russian Federation, Moscow, 19 November 2001 62 Interview with Professor Cristian Parvulescu, Bucharest, 8 March 2001. 63 Id. 64 Interview with Professor Mihai Constantinescu, Bucharest, 9 March 2001. 65 Id. 66 Id. 67 Interview with Professor Lucian Mihai, President of the Constitutional Court, Bucharest, 9 March 2001. 68 Id. 69 Interview with Mr Horiatiu Dumitru, Bucharest, 10 March 2001 70 Id. 71 Garlicki supra note 37 at 84. 72 Decision K 37/97 of 6 May 1998, Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1998 (C.H. Beck: Warszawa 1999): 167-75 at 172. 73 Decision 17/95 of 3 October 1995. The background of the decision is well described in Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago 2000) at 176-77. 74 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia, 11 May 2001. 75 In Albania, Bulgaria, Hungary, Russia, Slovakia, the Ukraine. 76 In Poland, until the 1997 Constitution removed this authority. 77 For example, in Hungary the following bodies can make such a request: the Parliament, the President, any parliamentary committee, the Government, the Chief Justice of the Supreme Court, the Head of the State Audit Office and the Attorney General. In Russia the range is somewhat narrower: the President, either chamber of Parliament, the Government and the legislative bodies of the entities that form part of the Federation. 78 For instance, the German Federal Constitutional Court decides on the interpretation of the Basic Law, but only in the event of disputes about the competencies of the highest federal bodies. 79 Decision 128/1990 of 18 December 1990. 80 Decision 5/1995 (V.10) AB hat., discussed in “Constitution Watch”, East Europ. Constit. Rev. 4:3 (Summer 1995): 10 at 10-11. 81 Brunner, supra note 32 at 80. 82 For example, on 20 February 1996 the Constitutional Court issued, at the request of President Yeltsin, an interpretation of the notion of parliamentary immunity. It established that such immunity did not release parliamentarians from liability for any violations of the law not connected with their official duties, see “Constitution Watch: Russia”, East Europ. Constit. Rev. 5:1 (Winter 1996): 21-25 at 24. 83 E.g. decision N 1-6/99 of 14 December 1999 regarding the interpretation of the constitutional provision that Ukrainian is the state language; a decision widely seen as adversely affecting the interests of the Russian-speaking minority; see Wolczuk, supra note 14 at 338-39. 84 For example, Decision 7/96 of 4 June 1996, summarised in Bull. Const. Case L. 1996 (2): 187-89, concerning the interpretation of freedom of expression, freedom of the press, and freedom of access to information; see the discussion in Chapter 6 of this book. 85 Venelin I. Ganev, “The Bulgarian Constitutional Court, 1991-1997: A Success Story in Context”, Europe-Asia Studies 55 (2003): 597-611 at 600. 86 Interview with Professor Petro F. Martinienko, Dean of the Faculty of Law, International Solomon University in Kiev, a former judge of the Constitutional Court of the Ukraine, Kiev, 22 November 2002. 87 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (199194), Sofia, 10 May 2001. 88 Id. 89 Interview with Professor Lucian Mihai, President of the Constitutional Court, Bucharest, 9 March 2001. 90 Weber, supra note 4 at 293. 91 Id. at 293. 92 Decision W 4/93 of 16 June 1993.
306 93
ENDNOTES
See Anna M. Ludwikowska, Sądownictwo konstytucyjne w Europie ĝrodkowo-Wschodniej w okresie przeksztaáceĔ demokratycznych (TNOiK, Torun 1997) at 94. 94 See id. at 97. 95 The exceptions are the courts in Belarus, the Czech Republic, Slovakia and Romania. 96 In Moldova and the Ukraine. In Moldova, all revisions of the Constitution must be first endorsed by the Constitutional Court before being submitted to the legislature (Art. 141 of the Constitution). In the Ukraine, the Constitutional Court considers proposed amendments to the Constitution (in an ex ante review) in terms of whether they restrict constitutional rights and freedoms or the independence and territorial integrity of the Ukraine, because such amendments are prohibited (arts. 157 and 159 of the Constitution). 97 In Albania, Bulgaria, the Czech Republic, Hungary, Poland (but only after the 1997 Constitution), Russia, Slovakia, and Slovenia. 98 In Bulgaria, also the Vice-President. 99 In Bulgaria, the Czech Republic, Hungary, and Slovakia. The procedure of impeachment may only be initiated by a constitutionally designated body, the Senate (in the Czech Republic) or the lower chamber of Parliament. In Russia, the Constitutional Court only decides on the legality of the preliminary phase of the impeachment process. In the Ukraine, the Constitutional Court can only declare the completion of the constitutional process of impeachment of the President, and in Romania the Constitutional Court only has a consultative role in the impeachment process. 100 In Albania, Bulgaria, the Czech Republic, Poland, Romania and Slovakia. In contrast, in Hungary a party may be prohibited by a local court (upon application by the public prosecutor), with an appeal to the Supreme Court. In Russia this power belonged to the Court on the basis of a constitutional amendment of the 2 April 1992, and was discontinued after the dissolution of the “First Court” in 1993. 101 More on this, in Chapter 3(2). 102 There will be further discussion on this in Chapter 3(3). 103 Different limits may, however, apply: 6 years in Moldova, 8 years in Croatia, 10 years in the Czech Republic, Latvia and Georgia, 12 years in Slovakia, 15 years in Russia. The 15-year term for Russian constitutional judges is a recent innovation. During the first stage of the Court's existence, judges were appointed for an unlimited term with compulsory retirement at the age of 65; in the second stage, a (non-renewable) tenure of 12 years was established, and extended by a new law of 2001 to 15 years. 104 In Bulgaria, Latvia, Lithuania, Poland, Romania, Russia, Slovakia (although only since a constitutional amendment in 2001), Slovenia and the Ukraine. 105 Eli M. Salzberger & Stefan Voigt, “On the Delegation of Powers: With Special Emphasis on Central and Eastern Europe”, Constitutional Political Economy 13 (2002): 25-52 at 38. 106 Id. at 38-39. 107 With the exception of Estonia, for the reasons mentioned above, and the “first Court” in Russia, which now can be seen as an aberration. 108 Leszek Lech Garlicki, “The Experience of the Polish Constitutional Court”, in Sadurski, supra note 2: 265-82 at 271. 109 Interview with Mr Mark Gillis, Prague, 21 March 2002. (Mr Gillies has taught law for several universities in Prague; from 1999 until 2001, he was the head of the Czech Supreme Court's Department of Foreign Relations). 110 Interview with Mr Petr Pithart, President of the Senate of the Czech Republic, Prague, 21 March 2002. 111 Interview with Professor Vojtech Cepl, Justice of the Constitutional Court of the Czech Republic, Prague, 22 March 2002 112 Garlicki, supra note 108 at 271. 113 As Professor Nenovsky observes, the number of academics has gradually decreased in the Bulgarian Constitutional Court; in the first Term it had three law professors and two doctors of law, while by 2001 it counted only one law professor (Todor Todorov). Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (1991-94), Sofia, 10 May 2001. 114 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia, 11 May 2001. 115 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (1991-94), Sofia, 10 May 2001.
ENDNOTES 116
307
Interview with Professor Volodymir Tychyj, judge of the Constitutional Court of the Ukraine and former Vice-President of the Court, Kiev, 22 November 2002. 117 In Russia, the Federation Council elects the judges from a list submitted by the President; in Slovenia and Serbia the parliament elects the judges from a group of candidates nominated by the President. 118 In Albania, the President elects judges with the consent of the Assembly; in the Czech Republic, the President appoints judges with the consent of the Senate. 119 In Bulgaria, four judges are appointed by the National Assembly, four by the President, and four by a joint meeting of the justices of the Supreme Court of Cassation and the Supreme Administrative Court; in Latvia, three are appointed by Parliament, two by the Government and two by the Plenum of the Supreme Council, and, in addition, all have to be confirmed by the Parliament; in Lithuania, the process of nomination is divided between the President, the Parliament and the Chair of the Supreme Court, and then three Parliament chooses three from each nominator; in Moldova, two are elected by Parliament, two by the President and two by the Supreme Council of the Magistracy; in Romania, three are appointed by the lower chamber of Parliament, three by the Senate and three by the President; in the Ukraine, six are appointed by the President, six by Parliament and six by the Assembly of Judges. 120 In Croatia (where the lower chamber elects judges by an absolute majority of votes, on the recommendation of the upper chamber of Parliament), Hungary (where the Parliament elects the judges by a two-thirds majority from among the candidates nominated by the parliamentary nominating committee, in which each parliamentary faction has one candidate) and Poland (where the lower chamber of Parliament, the Sejm, elects the judges by an absolute majority of votes). 121 Leszek Garlicki, “Trybunaá Konstytucyjny w projekcie Komisji Konstytucyjnej Zgromadzenia Narodowego”, PaĔstwo i Prawo 51:2 (1996): 3-19 at 6. 122 Garlicki, supra note 108 at 268. 123 Id. at 268. 124 Halmai, supra note 34 at 191-92. 125 Sanja Baric, “The Constitutional Court of the Republic of Croatia: Its Institutional Role Within the System of Government”, in Giuseppe di Vergotini, ed., Giustizia costituzionale e sviluppo democratico nei paesi dell’Europa Centro-Orientale (G. Giappichelli Editore: Torino, 2000): 115-25 at 117. 126 Interview with Professor Siniša Rodin of the Faculty of Law, University of Zagreb, Zagreb, 7 April 2000. 127 For a detailed account of the deal, see “Constitution Watch”, East Europ. Constit. Rev. 9:1/2 (Winter/Spring 2000) at 12. 128 Interview with Professor Yurii S. Shemshuchenko, Director of the Institute of State and Law, National Ukrainian Academy of Sciences, Kiev, 21 November 2002. 129 Interview with Judge Pavlo Jevgrafov, judge of the Constitutional Court of the Ukraine and former Vice-President of the Court, Kiev, 22 November 2002. 130 Interview with Professor Petro F. Martinienko, Dean of the Faculty of Law, International Solomon University in Kiev, former judge of the Constitutional Court of the Ukraine, Kiev, 22 November 2002. 131 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia, 11 May 2001 132 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (1991-94), Sofia, 10 May 2001. 133 Id. 134 Id. 135 Ganev, supra note 85 at 600-601. 136 Interview with Professor Cristian Parvulescu (Professor of Political Science), Bucharest, 8 March 2001. Although Professor Parvulescu stresses that all the judges of the Constitutional Court have high legal qualifications, in his opinion “they are not politically independent”, and the process of appointments is largely controlled by the office of the Presidency and also by the Minister of Justice. The impact of this is somewhat lower when the Government is based on a coalition (as in 1996-2000) than when there is a mono-party Government; in the former case, the President has to conduct negotiations within the coalition. According to Professor Parvulescu, one can identify the “Iliescu judges” and the “Constantinescu judges” within the current composition of the Constitutional Court
308
ENDNOTES
(referring, respectively, to Presidents Ion Iliescu, 1990-1996, and Emil Constantinescu, 1996-2000). It is also significant for him that the current (at the time of my interview) President of the Court, Professor Lucian Mihai, was, before his appointment in 1998, a secretary-general of the Chamber of Deputies (1996-98), “a political function” (in the opinion of Prof. Parvulescu) to which he was appointed by the then governing National Liberal Party. 137 Leigh Sprague, “The Russian Constitutional Court”, Parker Sch. J.E. Eur. L. 4 (1997): 339-56, at 345. 138 Interview with Dr Vladimir G. Yaroslavtsev, Justice of the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 139 Id. 140 He was invited to Moscow (prior to the appointment he was a professional judge in St. Petersburg) to a series of special meetings with high officials in the personnel office, and was then interviewed by a “personnel panel”. After his return from holiday he was urged to call immediately a “special number” at the Kremlin to contact, in the middle of the night, the “Head of Yeltsin’s administration”, who then announced to him that he was the President's choice; id. 141 “Neutral” is a self-description chosen by Justice Yaroslavtsev when hypothesising about some of the reasons for his successful appointment, id. 142 See Ludwikowska, supra note 93 at 186. Regarding Hungarian courts, see Brunner, supra note 32 at 82. 143 See footnote 18 above. 144 Tom Ginsburg, “Economic Analysis and the Design of Constitutional Courts”, Theoretical Inquiries in Law 3 (2002): 49-85 at 59. 145 Gábor Halmai, “Who is the Main Protector of Fundamental Rights in Hungary? The Role of the Constitutional Court and the Ordinary Courts”, in JiĜí PĜibáĖ, Pauline Roberts & James Young, eds., Systems of Justice in Transition: Central European Experiences since 1989 (Ashgate: Aldershot 2003): 50-73 at 66. 146 There are exceptions, however. In Hungary, the Supreme Court actually resisted a proposition (pressed upon it by the state prosecutor in a specific case before it) that it should consider all legal instruments relevant to the case, including the Constitution, and that in the process it should be able to set aside the sub-constitutional provisions inconsistent with the Constitution, see Halmai supra note 146 at 6566. The Supreme Court announced on this occasion: “Since no other agency is entitled by the Constitution to do so, only the Constitutional Court is entitled to state whether a given provision of substantive or procedural law valid at the time is in contradiction with the Constitution or not”, cited id. at 66. 147 Discussed in detail by Angela Di Gregorio, “The Evolution of Constitutional Justice in Russia: Normative Imprecision and the Conflicting Positions of Legal Doctrine and Case-Law in Light of the Constitutional Court Decision of 16 June 1998”, Review of Central and East Europ. Law 24 (1998): 387-419 at 389-96. 148 Quoted id. at 390. 149 For a discussion of this decision, see id. at 398-401. 150 See the opinions of Constitutional Court Justice Morshchakova expressed in legal periodicals, summarised by Di Gregorio, id. at 394-95. 151 Interview with Professor Boris A. Strashun of the Center for Analysis of Constitutional Justice at the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 152 Interview with Dr Vladimir G. Yaroslavtsev, Justice of the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 153 Quoted in Di Gregorio, supra note 147 at 403. 154 Oral remarks by Pedro Cruz Villalon, Justice of the Spanish Constitutional Court, Workshop on Constitutional Adjudication in Southern and Western Europe, Fondazione Adriano Olivetti, Rome, 25-26 March 2002. 155 According to the Constitution, the justices of the Constitutional Court are appointed by the President of the Republic with the consent of the Senate. However, the Senate had not been set up by 1996, and so until it was its role was played by the Chamber of Deputies. 156 Interview with Professor Vojtech Cepl, Justice of the Constitutional Court of the Czech Republic, Prague, 22 March 2002.
ENDNOTES
309
157
For an account of this controversy, see Pavel Holländer, “The Role of the Czech Constitutional Court: Application of the Constitution in Case Decisions of Ordinary Courts”, Parker Sch. J.E.Eur. L. 4 (1997): 445-65. 158 Decision IV.US 81/95 of 18 September 1995, http://www.concourt.cz/angl_verze/doc/4-81-95.html. 159 See Holländer, supra note 157 at 454. 160 Interview with Dr Vladimir Sladecek of the Law Faculty, Charles University of Prague, Prague, 23 March 2002 161 Holländer, supra note 157 at 450-52. 162 Id. at 452. 163 Id. at 456-57. 164 Garlicki, supra note 37 at 89. 165 Decision P. 8/00 of 4 October 2000, full text on file with the author. 166 Janusz TrzciĔski, “Orzeczenia interpretacyjne Trybunalu Konstytucyjnego”, PaĔstwo i Prawo 57:1 (2002): 3-14 at 9. 167 Garlicki, supra note 37 at 89 n. 17. 168 Id. at 89. 169 For a description of this “legal question” and of the Supreme Court’s response (in its decision of 4 July 2001), see TrzciĔski, supra note 166 at 12. 170 See “Ustawa zasadnicza w sądach powszechnych”, Rzeczpospolita (Warszaw) 5 July 2001 at C-1. 171 Lech Gardocki, “OsiągniĊcia i spory”, Rzeczpospolita (Warsaw) 15 April 2002 at C-2. 172 Roman Hausner, “Zapytajcie Trybunaá”, Rzeczpospolita (Warsaw) 18 March 2002, http://www.rp.pl/ gazeta/wydanie_020318/prawo/prawo_a_2.html. 173 Lech Gardocki, “Czy potrzebna jest wykáadnia Trybunaáu Konstytucyjnego”, Rzeczpospolita (Warsaw) 9 July 2001 at C-1. 174 See Marek Safjan, “Wykáadnia prawa - uĪyteczny instrument eliminowania niepewnoĞci”, Rzeczpospolita (Warsaw) 4 June 2001 at C-1. 175 Statement of the Chief Justice of the Polish Supreme Court, Lech Gardocki, quoted in “Trybunaá buduje praworządną Rzeczpospolitą”, Rzeczpospolita (Warsaw) 20 March 2002 at C-1; see also “Trybunaá przed dorocznym podsumowaniem”, Rzeczpospolita (Warsaw)19 March 2002 at C-2. 176 Id. 177 Gardocki, supra note 171 at C-2. 178 Czeszejko-Sochacki et al, supra note 55 at 155-56.
Chapter 2 1
By “quasi guardians” Dahl means the officials charged with the protection of fundamental rights and interests who are not themselves democratically controlled – such as the judges endowed with the power to declare legislation unconstitutional. 2 Robert A. Dahl, Democracy and Its Critics (Yale University Press: New Haven, 1989) at 188. 3 Michael S. Moore, “Law as a Functional Kind”, in Robert P. George, ed., Natural Law Theory: Contemporary Essays (Oxford University Press: Oxford, 1992): 188-242 at 229. 4 Id. 229. 5 Id. 228 6 The 1997 Constitution provided a two-year transitional period during which the decisions of the Constitutional Tribunal on the unconstitutionality of laws enacted under the old Constitution could be overridden by parliament; this possibility expired definitively on 17 October 1999. 7 Marek Safjan, “Epitafium dla nieostatecznych orzeczeĔ”, Rzeczpospolita (Warsaw) 4 October 1999 at C-2 8 Elsewhere, but still in the context of the same debate, Chief Justice Safjan claimed that the decisions of the Constitutional Tribunal conclusively put an end to emotional and politically charged debates “by appealing to objectified legal reasons, not to political criteria”, Marek Safjan, “Sąd ostateczny”, Wprost 17 October 1999 at 8 (emphasis added). 9 Quoted by Robert Alexy in “Constitutional Rights, Balancing, and Rationality”, Ratio Juris 16 (2003): 131-40 at 133. 10 James L. Gibson, Gregory A. Caldeira & Vanessa A. Baird, “On the Legitimacy of National High Courts”, American Political Science Review 92 (1998): 343-58 at 344-45.
310 11
ENDNOTES
Id. at 345, emphasis added. Id. at 345. 13 See e.g. John Hart Ely, Democracy and Distrust: A Theory of Judicial Review (Harvard University Press: Cambridge Mass., 1980): 135-79. 14 See Jeremy Waldron, “Precommitment and Disagreement”, in Larry Alexander, ed., Constitutionalism: Philosophical Foundations (Cambridge University Press: Cambridge, 1998): 271-99 at 280-81. 15 For an impressive statement and elaboration of the “comparative institutional” thesis, see in particular Neil K. Komesar, Imperfect Alternatives (The University of Chicago Press: Chicago, 1994). 16 There are also significant cultural factors, eg, the dominant social expectations concerning certain types of people who are encouraged to stand for election, or to apply for nomination to certain bodies. These cultural expectations are of course, themselves, partly determined by institutional factors (for example, by the procedures and formal criteria for election or nomination). 17 Philip Pettit, Republicanism (Oxford University Press: Oxford, 1997): 215-30. 18 See Jon Elster, “Majority Rule and Individual Rights”, in Stephen Shute and Susan Hurley (eds.), On Human Rights (Basic Books: New York, 1993): 175-216 at 179-80, 192-93. 19 See Jeremy Waldron, Liberal Rights (Cambridge University Press: Cambridge, 1993): 392-421. 20 Often this is the only sense in which “legitimacy” is used, especially when legitimacy of constitutional (and other) courts is discussed by political scientists; see, e.g., Gibson et al., supra note 10. 21 Bruce Ackerman, The Future of Liberal Revolution (Yale University Press: New Haven, 1992) at 109. Note that this is Ackerman’s wording, not Sólyom’s. 22 Id. at 143 n. 21 23 See Waldron, supra note 14 at 272-73. 24 Id. at 273. 25 Martin Shapiro, Courts: A Comparative and Political Analysis (University of Chicago Press: Chicago, 1981). 26 Id., chapter 1. 27 Id. at 8. 28 Id. at 1. 29 Id. at 8. 30 Jürgen Habermas, Between Facts and Norms, trans. William Rehg (Polity Press: Cambridge, 1996) at 262, emphasis in original. 31 As is the case, e.g., in Poland and Hungary. 32 As is the case in all other CEE countries, with the exception of Estonia. For a discussion of the selection of judges, see Chapter 1(3). 33 For a similar argument with respect to the Supreme Court of the United States, see Christopher L. Eisgruber, Constitutional Self-Government (Harvard University Press: Cambridge Mass., 2001) at 6466. 34 For such a conception of the role of the Supreme Court of the United States, see Richard H. Fallon, “The Supreme Court, 1996 Term – Foreword: Implementing the Constitution”, Harvard Law Review 111 (1997): 54-152 at 144-145. 35 Id. at 145, footnotes omitted, emphasis in the original. 36 See, generally, Wojciech Sadurski, “Conventional Morality and Judicial Standards”, Virginia Law Review 73 (1987): 339-97. 37 Burt Neuborne, “Judicial Review and Separation of Powers in France and the United States”, N.Y.U. Law Review 57 (1982): 363-442 at 368. 38 Part 7 of the Constitution of the Slovak Republic. 39 Art. 124 of the Constitution of the Slovak Republic. 40 See Zdzislaw Czeszejko-Sochacki, Leszek Garlicki & Janusz TrzciĔski, Komentarz do Ustawy o Trybunale Konstytucyjnym (Wydawnictwo Sejmowe: Warszawa, 1999) at 8, who state that, in Poland, the majority of authors consider the Constitutional Tribunal as belonging to the judicial branch. 41 Leszek Garlicki, ed., Konstytucja Rzeczypospolitej Polskiej: Komentarz (Wydawnictwo Sejmowe: Warszawa, 1999) (loose-leaf edition), chapter 8 at 10. 42 Owen Fiss, “Judiciary Panel: Introductory Remarks”, 19 Yale J. Int. L. (1994): 219-221 at 220. 43 Ruti Teitel, “Post-Communist Constitutionalism: A Transitional Perspective”, Columbia Human Rights Law Review 26 (1994): 167-190 at 178. 12
ENDNOTES 44
311
Ruti Teitel, “Transitional Jurisprudence: The Role of Law in Political Transformation”, Yale Law Journal 106 (1997): 2009-2080 at 2033 (footnote omitted). 45 “Kelsenian” is herein used as a short-hand to describe the Continental model of abstract and centralised review. I am however conscious that the model that emerged in Europe after the Second World War, in particular in Germany, but also in Italy, Spain, France etc, is not a purely “Kelsenian” model, because it envisaged, among other things, a rights-based scrutiny of constitutionality of laws, and contained important elements of “positive” legislation. In both these respects, Hans Kelsen expressed the opposite views when he advocated the establishment of the constitutional court in Austria. 46 Under the 1975 Constitution of Greece (art. 95), all courts have the power not to apply legal provisions that they consider to be contrary to the Constitution. A diffuse system exists also to a certain degree in Switzerland (although only the laws of the Cantons, not the federal ones, can be judicially reviewed) and in Portugal. 47 Allan Randolph Brewer-Carías, Judicial Review in Comparative Law (Cambridge University Press: Cambridge, 1989) at 128-131. 48 See Alec Stone Sweet, Governing with Judges (Oxford University Press: Oxford, 2000) at 120-21. 49 More on this in Chapter 1(4). 50 Wiktor OsiatyĔski, “Paradoxes of Constitutional Borrowing”, I.CON 1 (2003): 244-68 at 260. 51 See Ackerman, supra note 21 at 108-9. 52 Bruce Ackerman, “The Rise of World Constitutionalism”, Virginia Law Review 83 (1997): 771-797 at 776. 53 The United Kingdom and the Netherlands have no judicial constitutional review at all, while Denmark, Ireland, Greece and Sweden have adopted systems bearing resemblance to the US-style model of decentralised judicial review. 54 Louis Favoreu, “American and European Models of Constitutional Justice”, in David S. Clark, ed., Comparative and Private International Law: Essays in Honor of John Henry Merryman on His Seventieth Birthday (Duncker u. Humblot: Berlin, 1990), p. 110. 55 See, e.g., John Ferejohn & Pasquale Pasquino, “Constitutional Courts as Deliberative Institutions: Towards an Institutional Theory of Constitutional Justice” in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 21-36 at 31. A leading Russian constitutional expert used a similar argument when explaining to me why the decentralised system of review would not work in Russia: the majority of judges, he asserted, are old-fashioned and simply “do not know how to apply the Constitution”. Interview with Professor Boris A. Strashun, of the Center for Analysis of Constitutional Justice at the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 56 Personal conversation with a judge of the Polish Supreme Court, 16 July 2002. 57 Teitel, supra note 44 at 2032. 58 Stone Sweet, supra note 48 at 37; see also Stephen M. Griffin, American Constitutionalism (Princeton University Press: Princeton, 1996) at 121. 59 See, e.g., Andrzej Wasilewski, “Przedstawianie pytaĔ prawnych Trybunaáowi Konstytucyjnemu przez sądy (art. 193 Konstytucji RP)”, PaĔstwo i Prawo 54:8 (1999): 25-39 at 29; Anna M. Ludwikowska, Sądownictwo konstytucyjne w Europie ĝrodkowo-Wschodniej w okresie przeksztaáceĔ demokratycznych (TNOiK: Torun, 1997) at 21. 60 On “new constitutionalism” in Europe, contrasted to pre-World War II European constitutionalism, see Stone Sweet, supra note 48 at 31 and 37-8. 61 Interview with Professor Boris A. Strashun, of the Centre for Analysis of Constitutional Justice at the Constitutional Court of the Russian Federation, Moscow, 19 November 200. 62 This is not merely a theoretical possibility. Consider the current status of affirmative action, one of the most contentious issues in American constitutionalism. In 1996 the Court of Appeals for the 5th Circuit invalidated an affirmative action plan implemented by the University of Texas Law School and held that the use of race as a factor in university admissions was constitutionally proscribed; Hopwood v. Texas, 78 F.3d 932 (5th Cir. 1996). The other circuits follow the 1978 Supreme Court’s decision Regents of University of California v. Bakke, 438 U.S. 265 (1978), which explicitly permitted certain forms of race-based preferences in admissions. The Hopwood court argued that it was not bound by the Bakke precedent because Justice Powell’s opinion (according to the Court) did not garner a majority (in fact, the central part of Powell’s opinion, though not an opinion in its entirety,
312
ENDNOTES
was joined by the majority of judges). The Supreme Court denied certiorari in Hopwood, 518 U.S. 1033 (1996). I am grateful to Robert Post for pointing this out to me. 63 Michel Troper & Christophe Grzegorczyk, “Precedent in France”, in D. Neil MacCormick & Robert S. Summers, eds., Interpreting Precedents: A Comparative Study (Dartmouth: Aldershot, 1997): 103140 at 112-13 and 117. 64 See, e.g., Lech Morawski & Marek Zirk-Sadowski, “Precedent in Poland”, in MacCormick & Summers, supra note 63 at 219-58. 65 Stone Sweet, supra note 48 at 40. 66 Fiss, supra note 42 at 219. 67 For a characterisation of the Japanese system of constitutional review as “modelled very much after the American system of judicial review”, see Itsuo Sonobe, “Human Rights and Constitutional Review in Japan”, in David M. Beatty, ed., Human Rights and Judicial Review (Kluwer: Dordrecht, 1994): 135174 at 138. 68 I develop this argument in chapter 3(1). 69 John C. Reitz, “Political Economy and Abstract Review in Germany, France and the United States”, in Sally J. Kenney, William M. Reisinger & John C. Reitz, Constitutional Dialogues in Comparative Perspective (Macmillan: London, 1999): 62-88 at 74-84. 70 In contrast, such a deadline regarding a challenge initiated in the course of concrete review (but not constitutional complaint) that is, occasioned by a concrete litigation, would clearly be pernicious. A person has no control over when she can be brought to court under a particular law that she can then claim unconstitutionally violates her rights! 71 As an example of such a time limit, one might mention the rule in Poland until 1997 that abstract review of statutes applied only to statutes enacted no earlier than 5 years before the date of the Constitutional Tribunal’s decision (Art. 24 of the Law of 29 April 1985 on Constitutional Tribunal). This limit has been abandoned by the new statute on Constitutional Tribunal, adopted 1 August 1997. One may hypothesise that one reason why this provision was dropped had to do with its very low practical relevance: in a system of predominantly abstract review, where challenges to laws are most likely to be launched by the defeated parliamentary minority, it is highly unlikely that laws that have been on the books for a very long time will be called into question. 72 Reitz, supra note 69 at 80-81. 73 Id. at 81. 74 Id. at 81. See also Lea Brilmayer, “The Jurisprudence of Article III: Perspectives on the ‘Case or Controversy’ Requirement”, Harvard Law Review 93 (1979): 297-321 at 313. 75 See Stone Sweet, supra note 48 at 80-83 76 See id. at 66-8. 77 The words in quotation marks are from Lea Brilmayer, “A Reply”, Harvard Law Review 93 (1980): 1727-33 at 1732, and they apply not so much to an abstract review initiated by political bodies but to the idea of public interest litigation launched by “altruistic plaintiffs”. 78 Robert Badinter, quoted in Jean Gicquel, Droit constitutionnel et institutions politiques, 14th edn (Montchrestien: Paris, 1995), at 767 (emphasis in original). 79 See Reitz, supra note 69 at 81-84. 80 Robert A. Dahl, “Decision-Making in a Democracy: The Supreme Court as a National Policy-Maker”, Journal of Public Law 6 (1957): 279-295. 81 See Hunt v. Washington Apple Advertising Commission, 432 U.S. 333, 343 (1977), discussed by Brilmayer, supra note 77 at 318-19. In this decision, the Supreme Court unanimously accepted the standing of a state governmental commission composed of representatives of the apple industry (thus treating it as analogous to a voluntary association) to challenge the constitutionality of a statute regulating the packaging of apples. This is as clear a case as they have produced in terms of using concrete review in order to change economic policy. 82 Minister of Justice of Canada v. Borowski, [1981] 2 S.C.R. 575, 598, emphasis added. 83 See Charles R. Epp, The Rights Revolution (University of Chicago Press: Chicago, 1998) at 86. 84 Interview with Laszló Sólyom, East Europ. Constitut. Rev. 6:1 (Winter 1997): 71-77 at 72. 85 Andrew Arato, “Constitution and Continuity in the Eastern European Transitions: The Hungarian Case (part two)”, in Irena Grudzinska-Gross (ed.), Constitutionalism & Politics (Slovak Committee of the European Cultural Foundation, Bratislava 1993): 271-87 at 271. 86 Interview with Boris Ebzeev, East Europ. Constit. Rev. 6:1 (Winter 1997): 83- 88 at 86.
ENDNOTES 87
313
Martin Shapiro, “The Success of Judicial Review”, in Kenney, Reisinger & Reitz, supra note 69: 193219 at 205. 88 See Arato, supra note 85 at 272-3. 89 Spencer Zifcak, “Hungary’s Remarkable, Radical, Constitutional Court”, Journal of Constitutional Law in Eastern and Central Europe 3 (1996): 1-56 at 27. 90 See, similarly, Wiktor Osiatynski, “Rights in New Constitutions of East Central Europe”, Colum. Hum. Rts. L. Rev. 26 (1994): 111-166 at 151 n. 185. 91 Lee Epstein, Jack Knight & Olga Shvetsova, “The Role of Constitutional Courts in the Establishment and Maintenance of Democratic Systems of Government”, Law & Society Review 35 (2001): 117-63. 92 For discussion of this case, see Chapter 9, p. 248. 93 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (in 1991-94), Sofia, 10 May 2001. 94 According to Jonathan Macey, structural constitutional rules are “self-executing”, in contrast to “directives that forbid government officials from doing certain things” (such as, infringing individual rights), which “rely on an allegiance to vague constitutional principles”; Jonathan R. Macey, “Transaction Costs and the Normative Elements of the Public Choice Model: An Application to Constitutional Theory”, Virginia Law Review 74 (1988): 471-518 at 503. 95 “Constitution Watch”, East Europ. Constitut. Rev. 9:1/2 (Winter/Spring 2000) at 23. 96 Id. at 27. 97 See Neuborne supra note 37 at 369, footnote omitted. 98 See Cass Sunstein, “Introduction: The Legitimacy of Constitutional Courts: Notes on Theory and Practice”, East Europ. Constitut. Rev. 6:1 (Winter 1997) at 61-63. 99 The Polish Constitutional Tribunal recently struck down an amendment to the so-called “lustration law” on the basis that significant changes to the bill were made by the Senate, which had thus overstepped its law-making powers; see Decision K 11/02 of 19 June 2002, http://www.trybunal.gov.pl/ OTK/teksty/otkpdf/2002/K_11_02.pdf, discussed in Chapter 9, p. 247. 100 See Ely, supra note 13. 101 Neil K. Komesar, “Taking Institutions Seriously: Introduction to a Strategy for Constitutional Analysis”, Univ. of Chicago Law Rev. 51 (1984): 366-446 at 386; see also generally William W. van Alstyne, “A Critical Guide to Marbury v. Madison”, Duke Law Journal (1969): 1-47 at 23-4. 102 See text accompanying footnote 84 above. 103 Zifcak, supra note 89 at 27, footnote omitted. 104 Arato, supra note 85 at 272. 105 See Ely, supra note 13. 106 Decision K. 3/98 of 24 June 1998, Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1998 (C.H. Beck: Warszawa 1999), item 19: 308-71 at 353, translation in East European Case Reporter of Constitutional Law 6 (1999): 130-211. References here are to the Polish text. 107 Id. at 353. See, similarly, cases cited by Jerzy Oniszczuk, Konstytucja Rzeczypospolitej Polskiej w orzecznictwie Trybunaáu Konstytucyjnego (Zakamycze: Kraków, 2000) at 161, translation in East European Case Reporter of Constitutional Law 6 (1999): 130-211. 108 See Eisgruber, supra note 33 at 165-167 and 169-175. 109 Decision K. 3/98 supra note 106 at 353. 110 Id. 111 Id. at 354. 112 Dissenting opinions by Justice Rymarz, at 363-64, and Justice Zdyb, at 365-71. 113 Dissenting opinion by Justice Zdyb, at 370. 114 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (in 1991-94), Sofia, 10 May 2001. 115 See, e.g. Jean-Pierre Massias, Droit constitutionnel des États d’Europe de l’Est (Presses universitaires de France: Paris, 1999) at 163. 116 Jürgen Habermas, The Postnational Constellation: Political Essays, trans. Max Pensky (Polity: Cambridge, 2001) at 122. 117 Jeremy Waldron, “Rights and Majorities: Rousseau Revisited”, in John W. Chapman & Alan Wertheimer, eds, Majorities and Minorities: Nomos XXXII (New York University Press: New York, 1990): 44-75 at 59, footnote omitted. 118 Griffin, supra note 58 at 116, footnote omitted.
314
ENDNOTES
119
Id. at 116, emphasis added. Id. at 116. 121 Dahl, supra note 80 at 282. 122 Alec Stone Sweet, “Constitutional Dialogues: Protecting Rights in France, Germany, Italy and Spain”, in Kenney, Reisinger & Reitz, supra note 69: 8-41 at 27. 123 Of course, for the purposes of the theory of minority protection against the tyranny of majority it is not necessary (or even proper) to understand “minority” in statistical terms but rather in terms of underrepresentation of a particular category of citizens in the political system. 124 Bulgarian Constitutional Court decision of 22 April 1992, discussed in Chapter 8, pp. 220-22. 125 Vello Pettai, “Democratic Norm Building and Constitutional Discourse Formation”, paper presented at the workshop “Rethinking the Rule of Law in Post-Communist Europe”, European University Institute, Florence, 22-23 February 2002. More on this in Chapter 8, pp. 216-18. 126 See Kataryna Wolczuk, “The Constitutional Court of Ukraine: The Politics of Survival”, in Sadurski, supra note 55: 327-48 at 338-39. 127 See, respectively, Decision no. K. 26/96 of 28 May 1997 (abortion), Decision K. 11/90 of 30 January 1991 (religious teaching in schools), and Decision K. 17/93 of 7 June 1994 (broadcast law). All these three decisions are discussed in Chapter 6. 120
Chapter 3 1
See Chapter 1.2. Oral remarks by Professor László Sólyom, former Chief Justice of the Hungarian Constitutional Court, Workshop on Constitutional Adjudication in Southern and Western Europe, Fondazione Adriano Olivetti, Rome, 26 March 2002. 3 John C. Reitz, “Political Economy and Abstract Review in Germany, France and the United States”, in Sally J. Kenney, William M. Reisinger & John C. Reitz, Constitutional Dialogues in Comparative Perspective (Macmillan: London, 1999): 62-88 at 67. 4 Laurence H. Tribe, American Constitutional Law, 2nd edition (The Foundation Press: Mineola N.Y., 1988) at 153, footnote omitted. 5 Chicago & Grant Trunk Ry. V. Wellman, 143 U.S. 339 (1892). This maxim was established in the context of expressing the Court’s antipathy to “friendly suits”, the sole purpose of which is obtaining a judicial opinion. 6 Art. 5 of the Law on Constitutional Review Chamber Procedure of 5 May 1993. 7 Bruce Ackerman, The Future of Liberal Revolution (Yale University Press: New Haven, 1992) at 108. 8 Jürgen Habermas, Between Facts and Norms, trans. William Rehg (Polity Press: Cambridge, 1996) at 261, italics in original. 9 Id. 10 5 U.S. (1 Cranch) 137 (1803). 11 Stephen M. Griffin, American Constitutionalism (Princeton University Press: Princeton, 1996) at 96. 12 5 U.S. 177. 13 Neal Devins, Shaping Constitutional Values (Johns Hopkins University Press: Baltimore, 1996) at 13, italics added. 14 Martin Shapiro, “The Success of Judicial Review”, in Sally J. Kenney, William M. Reisinger & John C. Reitz, eds., Constitutional Dialogues in Comparative Perspective (Macmillan: London, 1999): 193219 at 211. 15 See William W. van Alstyne, “A Critical Guide to Marbury v. Madison”, Duke Law Journal (1969): 147 at 23-24. 16 As Jeremy Waldron recently observed, see “Eisgruber’s House of Lords”, University of San Francisco Law Review 37 (2002) 89-114 at 89. 17 See, e.g. Dickerson v. United States, 530 U.S. 428 (2000). In this decision, the Court found unconstitutional, by a 7-2 majority, a provision adopted by Congress that the admissibility of suspects’ statements in criminal cases turns solely on whether they were made voluntarily, regardless of whether the suspects had received so-called Miranda warnings before being interrogated. In his dissent, Justice Scalia warned that, until this law is repealed [by Congress], he would “continue to apply it in all cases where there has been a sustainable finding that the defendant’s confession was voluntary”, id. at 434. 2
ENDNOTES 18
315
Alexander M. Bickel, The Least Dangerous Branch (Bobbs-Merrill: Indianapolis, 1962) at 113-16. Id. at 114, emphasis added. 20 See Alec Stone Sweet, Governing with Judges (Oxford University Press 2000) at 90. 21 Frothington v. Mellon, 262 U.S. 447, 488-9 (1923). In this case, a taxpayer challenged the constitutionality of a statute (Maternity Act of 1921) that provided federal funds to state programs to reduce infant mortality, on the basis that it harmed her by increasing her tax liability. The Supreme Court held the action non-justiciable. 22 Fritz W. Scharpf, “Judicial Review and the Political Question: A Functional Analysis”, Yale Law Journal 75 (1966): 517-97 at 532-33, footnote omitted. 23 Who does not necessarily subscribe to the general position that I defend in this and in the following paragraphs. 24 Pasquale Pasquino, “Lenient Legislation: The Italian Constitutional Court” (unpublished manuscript, June 1999). It should be added that the Italian Constitutional Court also conducts abstract review only with respect to the regional authorities’ challenges to national laws and, vice versa, the national authorities’ objections to regional laws. However, an overwhelming majority of Court decisions are taken in the course of a concrete review, and Pasquino’s argument described in the main text refers only to this form of review. 25 Id. at 8. 26 Id. at 6. 27 Bickel supra note 18 at 115, emphasis added. 28 Id. at 115. 29 Michael S. Moore, “Law as a Functional Kind”, in Robert P. George, ed., Natural Law Theory: Contemporary Essays (Oxford University Press: Oxford, 1992): 188-242 at 230. 30 For criticism of this view, see Waldron, supra note 16 at 100-103. 31 Tom Ginsburg, “Constitutional Courts in New Democracies: Understanding Variation in East Asia”, Global Jurist Advances 2 (2002), available at http://www.bepress.com/gj/advances/vol2/iss1/art4/ at 17. 32 Oral remarks by Pedro Cruz Villalon, Justice of the Spanish Constitutional Court, Workshop on Constitutional Adjudication in Southern and Western Europe, Fondazione Adriano Olivetti, Rome, 25-26 March 2002. 33 Art. 144 (a) of the 1991 Constitution. 34 See Wojciech Sokolewicz, “Sąd konstytucyjny w Rumunii”, in Janusz TrzciĔski, ed., Sądy konstytucyjne w Europie, vol. 2 (Wydawnictwa Trybunaáu Konstytucyjnego: Warszawa, 1997): 14574 at 161. 35 Id. at 166. 36 Art. 107 of Estonian Constitution. 37 Art. 142. 38 See Leszek Garlicki, “Orzecznictwo Trybunaáu Konstytucyjnego w 1998 roku”, Przegląd sądowy (1999:6): 104-28 at 109. 39 Decision No. 71-44 DC. 40 See Alec Stone, The Birth of Judicial Politics in France (Oxford University Press: New York, 1992) at 209-10. 41 For example, in 1982, the Council struck down the Socialist government’s nationalisation bill on the grounds that its provisions for compensation violated property rights; it went on to state ways in which the government could save the bill by employing different formulas for the valuation of the companies concerned; the government then wrote the formulas into the law, and eventually the revised bill survived a second referral. The revision raised the cost of the nationalisations by 25 percent; see Stone supra note 40 at 241. 42 In one of its most famous decisions, in 1986, the Council invalidated a proposed Press Law on the basis that it provided insufficient guarantees for pluralism in media (a concept not mentioned in any constitutional text) and thwarted the Chirac government’s attempt to repeal the limits on press ownership; Decision no. 86-210 DC of 29 July 1986, see John Bell, French Constitutional Law (Oxford University Press: Oxford, 1992) at 327-30. 43 Laws may be referred to the Council by the President, the Prime Minister, the Presidents of the chambers of the Parliament and sixty members of either chamber of the Parliament. In practice, however, it is almost solely the instrument of parliamentarians. From 1974 (when the constitutional 19
316
ENDNOTES
amendment expanded the right of referral to any sixty deputies or senators) until 1987, out of 202 referrals, 196 were made by parliamentarians, and only 6 by other authorised persons; see Stone, supra note 40 at 58. 44 Stone, supra note 40 at 244. 45 Art. 128 (2) of the Constitution. 46 See art. 4(2) of the 1993 Law on Constitutional Review Court Procedure. 47 See Attila Agh, “The Permanent ‘Constitutional Crisis’ in the Democratic Transition: The Case of Hungary”, in Joachim Jens Hesse & Nevil Johnson, eds, Constitutional Policy and Change in Europe (Oxford University Press: Oxford, 1995): 296-326 at 316. 48 Andras Mink, “Interview with László Sólyom, President of the Hungarian Constitutional Court”, East Europ. Constit. Rev 6:1 (1997): 71-77 at 72 49 Decision 16/1991 of 20 April 1991, reprinted in László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 151-158, at 155. 50 See Anna M. Ludwikowska, Sądownictwo konstytucyjne w Europie ĝrodkowo-Wschodniej w okresie przeksztaáceĔ demokratycznych (TNOiK: Torun, 1997) at 60; Gabor Halmai, Comment, “The Constitutional Court of the Republic of Hungary”, East European Case Reporter of Constitutional Law 1 (1994) 116 at 118. 51 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (in 1991-94), Sofia, 10 May 2001. 52 See Mink, supra note 48 at 72. 53 Agh, supra note 47 at 316. 54 Interview with Professor Vojtech Cepl, Justice of the Constitutional Court of the Czech Republic, Prague, 22 March 2002. 55 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia, 11 May 2001. 56 See William L. Landes & Richard A. Posner, “The Economics of Anticipatory Adjudication”, Journal of Legal Studies 223 (1994): 683-719 at 685. Note that Landes and Posner write about binding adjudication in concrete cases (in which the courts are asked to pronounce on the legality or otherwise of a proposed course of action by an individual party) but their argument applies, mutatis mutandis, to any advisory role of courts in law-making. 57 See id. 686. 58 Stone Sweet, supra note 20 at 75-79. 59 This concern was expressed by Vice-President of the Constitutional Court of the Russian Federation, see Tamara Morshchakova, Panel discussion, in Konstytucja w sáuĪbie demokracji; Constitution in Service of Democracy (The International Centre for Development of Democracy Foundation, Cracow, March 10-12, 1995): 135-38 at 137. 60 Interview with Dr Vladimir G. Yaroslavtsev, Justice of the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 61 Id. 62 The rule was established in 1793, when the Court refused to provide an opinion, sought by President George Washington, concerning the obligations of the 1778 Franco-American Treaty. 63 The technical grounds for this distinction are that state courts (including state Supreme Courts) are not controlled by Art. III of the Constitution of the United States, which describes the jurisdiction of federal courts in terms of cases and controversies. Landes and Posner provide an economic explanation for the rejection of advisory opinions at the federal level, but not necessarily at the state level, see Landes & Posner, supra note 56 at 712. 64 Robert J. Pushaw, “Justiciability and Separation of Powers: A Neo-Federalist Approach”, Cornell Law Review 81 (1996): 393-512 at 443. See also Abner J. Mikva, “Why Judges Should Not Be Advicegivers: A Response to Professor Neal Katyal”, Stanford Law Review 50 (1998): 1825-32. 65 See Quill v. Vacco, 80 F.3d 716, 738-43 (2d Cir. 1996) (Calabresi, J., concurring), see also U.S. v. Then, 56 F.3d 464, 468-69 (2d Cir. 1995) (Calabresi, J., concurring). 66 80 F.3d at 738. 67 Alexander Bickel & Harry H. Wellington, “Legislative Purpose and the Judicial Process”, Harvard Law Review 71 (1957) 1, 34, quoted in 56 F.3d at 469 (Calabresi, J., concurring). For a recent
ENDNOTES
317
endorsement of this idea, see Michael C. Dorf, “The Supreme Court, 1997 Term – Foreword: The Limits of Socratic Deliberation”, Harvard Law Review 112 (1998): 4-83 at 69-70. 68 56 F.3d at 469. 69 See id. at 469. 70 See Dorf, supra note 67 at 70. 71 80 F.3d at 742, footnote omitted. 72 The possibility of an override did not apply when the preliminary abstract review was initiated by the President. This exemption of decisions of unconstitutionality taken as a result of Presidential initiative from the parliamentary override was the result of the Constitutional Tribunal’s own interpretation of its powers; see Decision W 1/95, discussed in Leszek Garlicki, “Orzecznictwo Trybunaáu Konstytucyjnego w 1995 roku”, Przegląd sądowy (1996:7-8): 110-38 at 116. 73 The discontinuation came into effect in November 1999; during the transitional period 1997-1999 the Parliament could, legally, override the laws adopted under the rule of the old Constitution. Of the eleven cases of a successful override, three took place in this transitional period. 74 See, e.g, Irena Grudzinska-Gross, “Interview with Professor Andrzej Zoll, Chief Justice of the Polish Constitutional Tribunal”, East Europ. Constit. Rev 6:1 (Winter 1997): 77-8 at 78: “The main problem is the ability of the Sejm to overrule Tribunal decisions….”. Elsewhere, Professor Zoll has argued that the retention of the parliament’s power to override the Tribunal’s decisions after the changes of 1989 was a remnant of a totalitarian system. His successor as Chief Justice, Professor Marek Safjan, declared that the end to the possibility of an override was “a victory of the Constitution over politics”, Marek Safjan, “Epitafium dla nieostatecznych orzeczeĔ”, Rzeczpospolita (Warsaw) 4 October 1999 at C-2. 75 Interview with Professor Lucian Mihai, President of the Constitutional Court, Bucharest, 9 March 2001. 76 Interview with Dr Vladimir G. Yaroslavtsev, Justice of the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 77 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (in 1991-94), Sofia, 10 May 2001. 78 Art. 79 (3). For an expression of mild scepticism as to the “absoluteness” of constitutional entrenchment of rights in German Basic Law, see Ackerman, supra note 7 at 144, n. 21: “It remains unclear how much the “absolute” character of German entrenchment is merely smoke and mirrors – and a good thing, too, for clarity comes only after a severe constitutional crisis”. 79 Francisco Rubio Llorente, quoted by Stone Sweet, supra note 20 at 59. 80 Art. 89. 81 Art. 57 (1). 82 See, similarly, Francis Hamon, Michel Troper, Georges Burdeau, Droit constitutionnel (L.G.D.J.: Paris, 2001, 27th ed.) at 41. 83 There is also an alternative, and more burdensome, form of constitutional amendment available in France, which requires an approval by a referendum. This is when the amendment procedure is not proposed by the Government with the President’s approval (Art. 89). 84 Jon Elster, Ulysses Unbound (Cambridge University Press: Cambridge 2000) at 100. 85 Interview with Dr Vladimir G. Yaroslavtsev, Justice of the Constitutional Court of the Russian Federation, Moscow, 19 November 2001 86 See, e.g., Sergio Bartole, “Concluding Remarks”, in Giuseppe de Vergottini, Giustizia costituzionale e sviluppo democratico nei paesi dell’Europa centro-orientale (Torino: G. Giappichelli Editore, 2000): 355-64 at 364; Hamon et al., supra note 82 at 743. 87 Stone Sweet, supra note 20 at 89. 88 Art. 70.1 of the Constitution of Hungary, see Ludwikowska, supra note 50 at 65 n. 32. 89 Herman Schwartz, “The New East European Constitutional Courts”, in A. E. Dick Howard (ed.), Constitution Making in Eastern Europe (Woodrow Wilson Center Press: Washington D.C., 1993): 163-208 at p. 176. 90 Art. 147 of the Romanian Constitution. 91 Interview with Professor Cristian Parvulescu, Bucharest, 8 March 2001. 92 Id. 93 Interview with Professor Mihai Constantinescu, Bucharest, 9 March 2001. 94 Id. 95 Interview with Professor Lucian Mihai, President of the Constitutional Court, Bucharest, 9 March 2001.
318
ENDNOTES
96
Id. Interview with Mr Horiatiu Dumitru, Bucharest, 10 March 2001 98 More specifically, it applies only to the provisions included in section 2 of the Charter (“Fundamental Freedoms”) and sections 7 to 15 (“Legal Rights” and “Equality Rights”). 99 Michael J. Perry, “The Constitution, the Courts, and the Question of Minimalism”, Northwestern University Law Review 88 (1993): 84-164 at 158. 100 Mark Tushnet, “Policy Distortion and Democratic Debilitation: Comparative Illumination of the Countermajoritarian Difficulty”, Michigan Law Review 94 (1995): 245-301 at 299. 101 Jeremy Webber, “Institutional Dialogue between Courts and Legislatures in the Definition of Fundamental Rights: Lessons from Canada (and elsewhere)”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 61-99 at 95; see also James Allan, “Bills of Rights and Judicial Power – a Liberal’s Quandary”, Oxford Journal of Legal Studies 16 (1996): 337-352 at 350 n. 59. 102 Martin H. Redish, “Congressional Power to Regulate Supreme Court Appellate Jurisdiction Under the Exceptions Clause: An Internal and External Examination”, Villanova Law Review 27 (1982): 900928. 103 See Tushnet, supra note 100 at 287. 104 Neal Devins, Shaping Constitutional Values (The Johns Hopkins University Press: Baltimore, 1996). 105 Id. at 7. 106 Id. at 55. 107 See generally, L.G. Ratner, “Majoritarian Constraints on Judicial Review: Congressional Control of Supreme Court Jurisdiction”, Villanova Law Review 27 (1981): 929-58 at 930-32. 108 In 1916 and 1919, Congress attempted to strike at child labour indirectly, using the interstate commerce and taxation powers; however, the Supreme Court invalidated both of these attempts, in Hammer v. Dagenhart, 247 U.S. 251 (1918), and in Child Labor Tax Case, 259 U.S. 20 (1922), respectively; in 1938 Congress returned to the original 1916 bill struck down in Hammer, and a unanimous Court approved the child labour legislation in 1941. 109 For an example of such behaviour in the American context, see J.R. Macey, “Thayer, Nagel, and the Founders’ Design: A Comment”, Northwestern University Law Review 88 (1993): 226-40 at 235. 97
Chapter 4 1
Alec Stone Sweet, Governing with Judges (Oxford University Press 2000) at 61. Rumyana Kolarova, “Bulgaria: A Self-Restricting Court”, East Europ. Constit. Rev. 2:2 (Spring 1983): 48-50 at 49. 3 Id. at 49. 4 US v. Manuel Then, 56 F.3d 464, 466 (2d Cir. 1995). 5 Id. at 469. 6 For a detailed description, see Gábor Halmai & Kim Lane Scheppele, “Living Well Is the Best Revenge: The Hungarian Approach to Judging the Past”, in A. James McAdams (ed), Transitional Justice and the Rule of Law in New Democracies (Notre Dame: University of Notre Dame Press, 1997): 155-84 at 174-7; see also Chapter 9 in this book, pp. 242-43. 7 Decision 48/1998 of 23 November 1998, described in Bull. Const. Case Law 3 (1998): 421-22; see also Chapter 6.1, pp. 133-34. 8 Decision U-I-13/94 of 21 January 1994; English translation on file with the author. 9 Martin Shapiro, “Some Conditions for the Success of Constitutional Courts: Lessons from the U.S. Experience”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 37-60 at 38; see also Stone Sweet, supra note 1 at 75. 10 Irena Grudzinska-Gross, “Interview with Professor Andrzej Zoll, Chief Justice of the Polish Constitutional Tribunal”, East Europ. Constit. Rev. 6:1 (Winter 1987): 77-78 at 78. 11 Interview with Prof. Lucian Mihai, President of the Constitutional Court, Bucharest 9 March 2001. 12 See Keith E. Whittington, “Legislative Sanctions and the Strategic Environment of Judicial Review”, I.CON 1 (2003): 446-474 at 462-3. 13 Stone Sweet, supra note 1 at 52-55. 14 Id. at 54. 2
ENDNOTES 15
319
Stephen M. Griffin, American Constitutionalism: From Theory to Politics (Princeton University Press: Princeton 1996) at 98. 16 Stone Sweet, supra note 1 at 53-54. 17 Renate Weber, “Constitutionalism as a Vehicle for Democratic Consolidation in Romania”, in Jan Zielonka, ed., Democratic Consolidation in Eastern Europe, vol. I: Institutional Engineering (Oxford University Press: Oxford 2001): 212-42 at 227. 18 Petr Kopecký, “The Czech Republic: From the Burden of the Old Federal Constitution to the Constitutional Horse Trading among Political Parties”, in Zielonka, ed., supra note 17: 319-46 at 343. 19 Interview with Mr Petr Pithart, President of the Senate of the Czech Republic, Prague 21 March 2002. 20 In this respect the Czech Senate is different from the Polish and Romanian Senates, where the term of office for Senators is the same as for members of the lower chamber: four years. 21 Decision of the Czech Constitutional Court no. 13/99 of 15 September 1999; see Jiri Priban, “Judicial Power vs. Democratic Representation: The Culture of Constitutionalism and Human Rights in the Czech Legal System”, in Sadurski, ed., supra note 9: 373-94 at 387 22 Such a suggestion was made by the President of the Senate; Interview with Mr Petr Pithart, President of the Senate of the Czech Republic, Prague 21 March 2002. It is also interesting to note that the President of the Constitutional Court, Zdenek Kessler, filed a dissenting opinion to this judgment. 23 Interview with Mr. Mark Gillis, Prague 21 March 2002. 24 Stone Sweet, supra note 1 at 55. 25 Id. at 55. 26 As is the case in e.g. Bulgaria, Latvia and Ukraine, see chapter 1, note 119. 27 As is the case if Hungary, see chapter 1, note 120. 28 “Constitution Watch: Bulgaria”, East Europ. Constit. Rev. 4:2 (Spring 1995): 5-7 at 7. 29 Venelin I. Ganev, “Interview with Constitutional Court Justices Todor Todorov and Tsanko Hadjistoichev”, East Europ. Constit. Rev. 6:1 (Winter 1997): 65-71 at 66. 30 See Anna M. Ludwikowska, Sądownictwo konstytucyjne w Europie ĝrodkowo-Wschodniej w okresie przeksztaáceĔ demokratycznych (TNOiK, Torun 1997) at 180. 31 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (in 1991-94), Sofia 10 May 2001. 32 Id. 33 Alina Mungiu-Pippidi, “Interview with President of the Romanian Constitutional Court, Ion Muraru”, East Europ. Constit. Rev. 6:1 (Winter 1997): 78-83 at 81. 34 Leigh Sprague, “The Russian Constitutional Court”, Parker Sch. J.E. Eur. L. 4 (1997): 339-56 at 349. 35 Id. 36 Interview with Professor Boris A. Strashun, of Center for Analysis of Constitutional Justice at Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 37 Interview with Professor Andrey Nikolaevich Medushevsky, constitutional expert of the Institute of Law and Public Policy, Moscow, 19 November 2001 38 Id. 39 Id. 40 Oral remarks by Dr Pedro Magalhães, Workshop on Constitutional Adjudication in Southern and Western Europe, Fondazione Adriano Olivetti, Rome, 25 March 2002. 41 For a recent plea against the use of the notion of judicial activism, see Aharon Barak, “The Supreme Court, 2001 Term – Foreword: A Judge on the Judging: The Role of a Supreme Court in a Democracy”, Harvard Law Review 116 (2002): 16-162 at 126-27. 42 These features are pointed to by Ganev in the context of his discussion of whether the Bulgarian Constitutional Court can be dubbed as “activist”, see Venelin I. Ganev, “The Bulgarian Constitutional Court, 1991-1997: A Success Story in Context”, Europe-Asia Studies 55 (2003): 597-611 at 606. 43 See East European Case Reporter of Constitutional Law 3 (1996), no. 1. 44 See Chapter 6.1. 45 Id. 46 See Chapter 9.2. 47 See Chapter 9.3. 48 For example, the Hungarian Constitutional Court struck down important aspects of a number of laws that were meant to constitute a package of austerity measures introduced by the Government in 1995; see e.g. decision 43/1995 of 30 June 1995 on social security benefits, reprinted in László Sólyom &
320
ENDNOTES
Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000) at 323-32. 49 See, e.g., the decision of the Polish Constitutional Tribunal; no. K 8/97 of 16 December 1997 striking down a number of provisions of the tax statute of 26 July 1991, reprinted in Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1997 (Warszawa: C.H. Beck 1998): 545-59. 50 In Slovenia, the Constitutional Court decided Case No. U-I-206/97, annulling on 17 June 1998 part of a law on the amendments to the Law on Foreigners. The amendments would have changed the required period before an immigrant could apply for permanent resident status from three to eight years. See “Constitution Watch: Slovenia”, East Europ. Constit. Rev. 7:3 (Summer 1998): 36-37. 51 On 13 April 1991, the Hungarian Constitutional Court declared the use of uniform personal identification numbers unconstitutional, decision 15/1991, reprinted in Sólyom & Brunner, supra note 48 at 139-50. 52 The Croatian Constitutional Court invalidated, in 1998, a provision of the 1993 Code on Equating Retirement Incomes on the basis that the code demanded that pensions increase relative to changes in the cost of living rather than relative to the increase of average incomes, see “Constitution Watch: Croatia”, East Europ. Constit. Rev. 7:3 (Summer 1998): 8-9 at 9. 53 Decision K. 19/96 of 24 February 1997, in Orzecznictwo, supra note 49: 65-77 at 72. 54 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia 11 May 2001. A similar presumption of constitutionality is in practice adopted by the Russian Constitutional Court; Interview with Dr Vladimir G. Yaroslavtsev, Justice of the Constitutional Court of the Russian Federation, Moscow, 19 November 2001. 55 E.g. in Poland, decision of the Constitutional Tribunal K 17/93 of 7 June 1994. 56 Pavel Holländer, “The Role of the Czech Constitutional Court: Application of the Constitution in Case Decisions of Ordinary Courts”, Parker Sch. J.E.Eur. L. 4 (1997): 445-65 at 450-52. 57 Id. at 452. 58 For a description of this case, see Mark Gillis, “Constitutionalism in the Czech Republic: An investigation of Two Major Aspects – The Infusion of Constitutional Principles into the Remainder of the Legal Order and the Generally Binding Nature of Constitutional Court Decisions”, Journal of Constitutional Law in Eastern and Central Europe 5 (1998): 105-245 at 139-42. 59 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (in 1991-94), Sofia 10 May 2001. 60 Decision no. 64/1991 (XII 17) AB of 17 December 1991, East European Case Reporter of Constitutional Law 1 (1994) 27. 61 E.g. Decision of Polish Constitutional Tribunal no. K 22/95 of 29 May 1996 in Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1996 vol. 1 (Warszawa: C.H. Beck, 1996): 106-21 at 120. 62 See Decision of Polish Constitutional Tribunal no. K 13/95 of 24 September 1996, in Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1996 [Case Law of the Constitutional Tribunal, 1996], vol. 2 (Warszawa: C.H. Beck, 1996): 79-105 at 104. 63 Judgment of the Constitutional Court of the Czech Republic no. 46/96 of 13 November 1997, quoted in Holländer, supra note 56 at 463 n. 23. 64 E.g. Decision of Polish Constitutional Tribunal no. K 19/96 of 24 February 1997, supra note 53 at 7273. 65 Id. at 72-73. 66 See, e.g., Decision of Polish Constitutional Tribunal no. K. 2/98 of 23 March 1999, in Orzecznictwo Trybunaáu Konstytucyjnego w 1999 r. vol. 1 (Warszawa: TK 1999): 176-81 at 178 emphasis added. 67 James Bradley Thayer, “The Origin and Scope of the American Doctrine of Constitutional Law”, Harv. L. Rev. 7 (1893): 17-156 at 144. 68 Andras Mink, “Interview with László Sólyom, President of the Hungarian Constitutional Court”, East Europ. Constit. Rev. 6:1 (Winter 1997): 71-76 at 72. 69 Quoted by Gábor Halmai, “Comment: The Constitutional Court of the Republic of Hungary”, East European Case Reporter of Constitutional Law 1 (1994) 116 at 116. 70 Id. 71 Decision 23/1990 of 31 October 1990, reprinted in Sólyom & Brunner, supra note 48 at 118-38. 72 Id. at 133 (Sólyom, P., concurring).
ENDNOTES
321
73
Decision no. K 26/96 of 28 May 1997, in Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1997 (Warszawa: C.H. Beck, 1998): 173-246. This decision was reprinted in East European Case Reporter of Constitutional Law 6 (1999): 38-129. 74 The new Constitution was adopted by the National Assembly on 2 April 1997, subjected to the constitutional referendum on 25 May 1997, promulgated by the President on 16 July 1997, and entered into force on 17 October 1997. 75 Jan WoleĔski, “Glosa do orzeczenia TK z 28 V 1997, K 26/96”, PaĔstwo i Prawo 53:1 (1998): 88-98 at 91. 76 Id at 91. 77 Id. 78 Id 79 See T. Alexander Aleinikoff, “Constitutional Law in the Age of Balancing”, Yale L. J. 96 (1987) 9431005 at 987. 80 Enzo Cheli & Filippo Donati, “Methods and Criteria of Judgment on the Questions of Rights to Freedom in Italy”, in David M. Beatty, ed., Human Rights and Judicial Review (Kluwer: Dordrecht 1994): 227-65 at 261. 81 Resolution U-I-121/97 of 23 May 1997, reprinted in East European Case Reporter of Constitutional Law 4 (1997): 279-303. 82 Id. at 288, italics added. 83 Id. at 286.
Chapter 5 1
See in particular Ronald Dworkin, Taking Rights Seriously (London: Duckworth, 1978, 2nd ed.) chapters 4 and 5; A Matter of Principle (Harvard University Press, 1985), chapters 1 and 2; Law’s Empire (London: Fontana, 1986), pp. 373-79; Freedom’s Law (Oxford University Press, 1996), pp. 1-38, 35272. 2 Freedom’s Law, supra note 1 at 12. 3 See e.g. Dworkin’s critique of City of Richmond v. Croson, 488 U.S. 469 (1988) (striking down an affirmative action plan) id. at 158; Webster v. Missouri Reproduction Services, 992 U.S. 490 (1989) (upholding restrictions on the availability of abortion), id. at 60-71; Buckley v. Valeo, 424 U.S. 1 (1976) (striking down limits on expenditure in an election campaign) in Sovereign Virtue (Harvard University Press, 2000) at 351-385; Bowers v. Hardwick, 478 U.S. 186 (1986) (upholding a state antisodomy statute), id. at 454-55, Washington v. Glucksberg, 521 U.S. 702 (1997) and Vacco v. Quill, 521 U.S. 793 (1987) (upholding state statutes prohibiting doctor-assisted suicide), id. at 453-73. 4 Freedom’s Law, supra note 1 at 352. 5 At times he admits to a negative record but he quickly announces that the positive decisions far outweigh the negative ones; see e.g. Law’s Empire, supra note 1 at 375-76. 6 See Neal Devins, Shaping Constitutional Values (Baltimore: The Johns Hopkins University Press, 1996), pp. 16-17 and 32; Neal Devins, “The Democracy-Forcing Constitution”, Michigan Law Review 97 (1999): 1971-93 at 1987-88; Stephen M. Griffin, American Constitutionalism (Princeton University Press, 1996) at 116-17. 7 Earl Warren was Chief Justice in 1953-1969, Warren Earl Burger in 1969-1986, and William H. Rehnquist since 1986. Roger B. Taney was the Court’s Chief Justice in 1836-1854, and he authored the Court’s opinion in Dred Scott v. Sanford, 60 U.S. (17 How.) 393 (1857). This decision affirmed the right to own a slave as a constitutional right, prohibited Congress from preventing the spread of slavery into the free states and territories, and denied Africans in America the status of citizenship. 8 In Texas v. Hopwood, 78 F.3d 932 (5th Cir. 1996) a federal Court of Appeals held a Texas university affirmative-action policy to be unconstitutional. The Supreme Court subsequently refused to hear the university’s appeal, Texas v. Hopwood, 518 U.S. 1033 (1996). Note, however, that more recently the Supreme Court upheld, on very narrow grounds and by a 5-4 majority, race-conscious preferences in university admissions (in Grutter v. Bollinger, 539 U.S. 306 (2003) while at the same time striking down a “quota-based” affirmative action admissions system (in Gratz v. Bollinger, 539 U.S. 244 (2003).
322 9
ENDNOTES
This seems to be Dworkin’s view: “we would have more to regret if the Court had accepted passivism wholeheartedly: southern schools might still be segregated, for example”, Law’s Empire, supra note 1 at 375. 10 Mark Tushnet, Taking the Constitution Away from the Courts (Princeton University Press, 1999) at 129-54. 11 Freedom’s Law, supra note 1 at 34. 12 Id. at 35. 13 Dworkin, supra note 3 at 189, see also id. at 357. 14 First put forward in Jeremy Waldron, “A Rights-Based Critique of Constitutional Rights”, Oxford Journal of Legal Studies 13 (993): 18-51; and then developed in Jeremy Waldron, Law and Disagreement (Oxford: Oxford University Press, 1999), especially chapters 10-13. 15 Waldron, “A Rights-Based Critique”, supra note 14 at 42. 16 See id. at pp. 44-5; James Allan, “Bills of Rights and Judicial Power – A Liberal’s Quandary”, Oxford Journal of Legal Studies 16 (1996): 337-52 at 349-50. 17 Griffin, supra note 6 at 123. 18 Waldron, “A Rights-Based Critique”, supra note 14 at 50, emphasis added. 19 See Neil K. Komesar, Imperfect Alternatives (The University of Chicago Press, 1994), pp. 56, 79-81; Bruce A. Ackerman, “Beyond Carolene Products”, Harvard Law Review 98 (1985): 713-46 at 71819. 20 John Hart Ely, Democracy and Distrust: A Theory of Judicial Review (Cambridge, Mass.: Harvard University Press, 1980). 21 Waldron, Law and Disagreement, supra note 14 at 295-96. 22 See Laurence Tribe, “The Puzzling Persistence of Process-Based Constitutional Theories”, Yale Law Journal 89 (1980): 1063-1080, pp. 164-65. 23 See, e.g., Articles 18.2 and 21.2.6 of the Broadcast Law in Poland of December 29, 1992, upheld as constitutional by the Constitutional Tribunal on 7 June 1994; see chapter 6 below. 24 Waldron, Law and Disagreement, supra note 14 at 222. 25 See also Cécile Fabre, “The Dignity of Rights”, Oxford Journal of Legal Studies 20 (2000): 271-282 at pp. 275-6. 26 Tushnet, supra note 10 at 57-70. 27 Id. at 137. 28 “Legislators may define their jobs as excluding considerations of the Constitution precisely because the courts are there. The judicial overhang might make the Constitution outside the courts worse than it might be”, id. at 58. 29 For the United States, see Devins, “The Democracy Forcing Constitution”, supra note 6 at 1985. 30 This basically applies only to those few cases of constitutional courts that can act on their own initiative, and that have the power of identifying unconstitutional legislative omissions (such as the Hungarian Constitutional Court). 31 Dworkin, Freedom’s Law, supra note 1 at 31. 32 See John Rawls, Political Liberalism (New York: Columbia University Press, 1993) at 144-68. 33 Dworkin, Taking Rights Seriously, supra note 1 at 85. See also Dworkin, A Matter of Principle, supra note 1 at 24 and 70; Sovereign Virtue, supra note 3 at 208. 34 Owen Fiss, “The Forms of Justice”, Harvard Law Review 93 (1979): 1-58 at 10 35 Michael S. Moore, “Law as a Functional Kind”, in Robert P. George, ed., Natural Law Theory: Contemporary Essays (Oxford: Oxford University Press, 1992): 188-242 at 231. 36 Freedom’s Law, supra note 1 at 31, emphasis added. For a powerful critique, see Waldron, Law and Disagreement, supra note 14 at 289-91; see also Cass R. Sunstein, The Partial Constitution (Harvard University Press: Cambridge Mass. 1993) at 145-46. 37 Christopher L. Eisgruber, “Is the Supreme Court an Educative Institution?”, NYU Law Review 67 (1992), 961-1032 at 1003, footnote omitted. Note that this observation by Eisgruber is not made in the context of a discussion on rights-related reasoning by courts. 38 Ronald Dworkin, Freedom’s Law: The Moral Reading of the American Constitution (Oxford University Press, 1996). 39 For a similar argument formulated as a criticism of Christopher Eisgruber’s theory of judicial review (developed in Christopher L. Eisgruber, Constitutional Self-Government (Harvard University Press:
ENDNOTES
323
Cambridge Mass., 2001)) see Jeremy Waldron, “Eisgruber’s House of Lords”, University of San Francisco Law Review 37 (2002) 89-114 at 90-99. 40 Eisgruber, supra note 39 at 58. In the text accompanying the quoted words, Eisgruber focuses on life tenure and guaranteed salary as factors guaranteeing the disinterestedness of US Supreme Court justices. 41 Abner J. Mikva, “Why Judges Should Not Be Advicegivers: A Response to Professor Neal Katyal”, Stanford Law Review 50 (1998): 1825-32 at 1829. 42 Waldron, Law and Disagreement, supra note 14 at 307.
Chapter 6 1
Art. 54 (1). 2 Art. 30. 3 Jerzy Oniszczuk, Konstytucja Rzeczypospolitej Polskiej w orzecznictwie Trybunaáu Konstytucyjnego (Zakamycze: Kraków, 2000), p. 236. According to Catherine Dupré, by “importing” the law developed by German Constitutional Court, the Hungarian CC used the rights and principles associated with human dignity as a “modern substitute for natural law”, see Catherine Dupré, “Importing Human Dignity from German Constitutional Case Law”, in Gabor Halmai, ed., The Constitution Found? The First Nine Years of Hungarian Constitutional Review on Fundamental Rights (Indok: Budapest 2000): 215-26 at 220-22. 4 Decision no. 19/97 of 21 November 1997, summarized in Bull. Const. Case L. 3 (1997): 357-58. 5 Art. 4 (2). 6 Decision 8/1993 of 27 February 1993, E. Europ. Case Rep. 1 (1994): 243-44. 7 Art. 2. 8 Art. 54 (1). 9 The distinction is less sharp in case-law practice than in theory. While in the Hungarian constitution “human dignity” is expressly stated as a “right”, one commentator notes that “Hungarian judges are coming round to the widely held view that human dignity, described as ‘source-law’ or ‘parent-law’ in Hungarian case-law, is the source of other constitutional rights, that it is the basis of other rights and, indeed, of the constitutional system itself”, Catherine Dupré, “The Right to Human Dignity in Hungarian Constitutional Case Law”, in The Principle of Respect for Human Dignity (Council of Europe Publishing: Strasbourg 1999): 68-79 at 76. 10 Decision SK 20/98 of 1 June 1999, Orzecznictwo Trybunaáu Konstytucyjnego TK w 1999, cz. 1 (Wydawnictwa Trybunaáu Konstytucyjnego, Warszawa 2000): 353-70. 11 Decision U. 5/97 of 19 May 1998, Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1998 (C.H.Beck, Warszawa 1999): 195-210. 12 Art. 47. 13 Art. 51. 14 Decision U. 5/97, supra note 11 at 204. 15 For details, see Agata Fijalkowski, “The Abolition of the Death Penalty in Central and Eastern Europe”, Tilburg Foreign Law Review 9 (2001): 62-83 at 72-73. 16 Not a member of Council of Europe. 17 See http://conventions.coe.int/Treaty/EN/cadreprincipal.htm. 18 See, e.g., Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago, 2000) at 89. 19 See Fijalkowski, supra note 15 at 63. 20 Belarus (article 24), Georgia (article 15), Moldova (article 24), Russia (article 20), Montenegro (article 21) and Serbia (article 14) [note that, hereinafter, the name of a country followed by the number of an article refers to the relevant provision of the constitution of that country]. 21 The words in quotation marks are taken from the Constitution of Belarus, art. 24. 22 Constitution of Georgia: “[C]apital punishment before its full abrogation may be envisaged by the organic law…” (art. 15 (2), emphasis added); Constitution of Moldova: “Until its final prohibition, capital punishment may be applied…” (art. 24 (3), emphasis added). 23 http://www.web.amnesty.org/rmp/dplibrary.nsf/index?openview. 24 Namely, the constitutions of Croatia (art. 21); Charter of Rights of the Czech Republic [hereinafter referred to as Czech Charter] (art. 6); Macedonia (art. 10); Romania (art. 22); Slovenia (art. 17); Slovakia (art. 15).
324 25
ENDNOTES
Decision No. 23/1990 of 31 October 1990; translated in East European Case Reporter of Const. Law 1 (1994): pp. 177-205, and in László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000) at 118-138. 26 Sólyom & Brunner, id. at 118. 27 See Anna M. Ludwikowska, Sądownictwo konstytucyjne w Europie ĝrodkowo-Wschodniej w okresie przeksztaáceĔ demokratycznych (TNOiK: ToruĔ 1997) at 129. 28 See Chapter 4.4, p. 101. 29 László Sólyom, “Introduction to the Decisions of the Constitutional Court of the Republic of Hungary”, in Sólyom & Brunner, supra note 25 at 6, emphasis added. 30 Sólyom, P. concurring, see the text of the opinion in Sólyom & Brunner, supra note 25 at 125. 31 Section V of the Opinion of the Court, id. at 122. 32 Section IV of the Opinion of the Court, id. at 122. 33 Dissent by Schmidt, J., id. at 123. 34 Id. at 136-8. 35 Id. at 133. 36 Art. 6 (1). Similarly Slovakia art. 15 (1). 37 Art. 55 (1). 38 Kim Lane Scheppele, “Women’s Rights in Eastern Europe”, East Europ. Constit. Rev. 4:1 (Winter 1995): 66-69 at 68. 39 Decision 64/1991 of 17 December 1991, translation reprinted (excerpts) in Sólyom & Brunner, supra note 25 at 178-199. 40 Comment to decision 64/1991 by Sólyom in Sólyom & Brunner, supra note 25 at 178. 41 Decision 64/1991 at 193-95. 42 Id. at 184. 43 Id. at 187. 44 Id. at 187. 45 Id. at 188. 46 Id. at 188. 47 Decision 48/1998 of 23 November 1998, described in Bull. Const. Case Law 3 (1998): 421-22. 48 “Constitution Watch: Hungary”, East Europ. Constit. Rev. 7:4 (Fall 1998): 15-18 at 17. 49 I am grateful to Renata Uitz of Central European University in Budapest for this information. 50 Decision U. 8/90 of 15 January 1991. 51 The two judges who formed the majority, Professors Andrzej Zoll and (the late) Tomasz Dybowski, never made a secret of their deep religious beliefs. 52 Typed text of the Decision U.8/90, on file with the author, p. 3. 53 Id. pp. 5-6. 54 In Polish: “izba lekarska” corresponds to the doctors’ self-regulatory body. 55 Decision of 7 October 1992, U.1/92. 56 Typed text of Decision of 7 October 1992, U.1/92, on file with the author, at 6. 57 Decision no. K 26/96, reprinted in East European Case Reporter of Constitutional Law 6 (1999) 38129, and in Polish Law Journal 6 (2001): 229-329. 58 Chapter 4.4, pp. 101-103. 59 Proposals to include the words “from the moment of conception” to a right to life provision had frequently been made in the constitution-drafting process, and ultimately rejected. As one academic commentator noted: “We therefore have had access to an authentic interpretation”, Jan WoleĔski, “Glosa do orzeczenia TK z 28 V 1997, K 26/96”, PaĔstwo i Prawo 53:1 (1998): 88-98 at 91. 60 The only cases that the Tribunal upheld as legislatively permitted abortion were those justified on strictly defined medical grounds (because of threat to mother’s health, or the genetic defects of the foetus) or resulting from rape. 61 WoleĔski, supra note 59 at 91. 62 Women’s Rights Centre, “Women’s Rights in the Constitution of Poland”, Polish Legal Journal 6 (2001): 87-104 at 101. 63 Id. at 101. 64 Russia art. 14. 65 Czech Republic art. 2 (1); Lithuania art. 43 (7).
ENDNOTES 66
325
Leszek Lech Garlicki, “Perspectives on Freedom of Conscience and Religion in the Jurisprudence of Constitutional Courts”, Brigham Young Univ. L. Rev. (2001): 467-510 at 479. 67 Hungary art. 60 (3); Slovenia art. 7 (1), Russia art. 14 (2); Bulgaria art. 13 (2). 68 Garlicki, supra note 66 at 476. 69 Art. 25 (2). 70 Art. 25 (3). 71 Wojciech Sadurski, Moral Pluralism and Legal Neutrality (Kluwer: Dordrecht, 1990), chap. 7. 72 Croatia art. 41 (1); Hungary art. 60 (3); Latvia art. 99; Russia art. 14; Slovenia art. 7; Ukraine art. 35 [note that, hereinafter, a name of a country followed by a number of an article refers to the provision of the constitution of that country]; Charter of Human and Minority Rights and Civil Liberties of the State Union of Serbia and Montenegro [hereinafter referred to as Serbia and Montenegro Charter] art. 27. 73 Czech Charter art. 2 (1); Estonia art. 40 (2); Lithuania art. 43 (7); Slovakia art. 1. One minor departure from the separation principle in the Lithuanian Constitution is the statement that the State shall recognize marriages registered in Church, art. 38 (4). 74 Bulgaria art. 13 (3); Georgia art. 9. 75 Moldova art. 31 (4); Romania art. 29 (5). 76 Montenegro art. 11. 77 Serbia art. 41. 78 Albania art. 10 (4); Poland art. 25 (3). 79 See http://www.lrkt.lt/2000/r000613.htm. 80 Art. 25 (2). 81 Decision K11/90 of 30 January 1991, discussed in Leszek Garlicki, “Przegląd orzecznictwa Trybunaáu Konstytucyjnego w 1991 roku”, Przegląd Sądowy (1991, no. 11-12): 43-63 at 59-60. 82 The separate opinion by Judge Dziaáocha concluded that the regulation would be constitutional but only if decided in a statutory form by the Parliament, K 11/90, typed decision on file with the author, at 62. 83 K 11/90 at 38. 84 Decision U. 12/92 of 20 April 1993. 85 But see Constitutions of Georgia (art. 19), Latvia (art. 99) and Romania (art. 29) where the right is formulated in a fairly basic form. 86 Lithuania art. 26 (5); Moldova art. 35 (9); Poland art. 48; Romania art. 29 (6); Slovenia art. 41. 87 Albania art. 24 (3); Estonia art. 42; Poland art. 53 (7); Slovenia art. 41; Montenegro art. 34; Serbia and Montenegro Charter art. 26. 88 Ten Constitutions containing such an article are constitutions of Albania art. 24 (3); Belarus art. (31); Bulgaria art. 37; Czech Republic art. 15 (1); Lithuania art. 26(3); Poland art. 53 (6); Romania (art. 29); Russia art. 28; Ukraine art. 35. 89 See, e.g., Czech Republic art. 16 (1); Hungary art. 33 (2); Poland art. 53 (2). 90 Albania art. 10 (6); Czech Republic art. 16 (2); Lithuania art. 43 (3); Moldova art. 31 (2); Romania art. 29 (3); Slovakia art. 24 (3); Montenegro art. 11; Serbia art. 41; Serbia and Montenegro Charter art. 27. 91 Croatia art. 41(2); Lithuania art. 43(3); Macedonia art. 19(3); Serbia art. 41. 92 Hungary, Macedonia and Montenegro. 93 With the exception of Georgia, Moldova, Romania and Serbia. 94 With the exception of Moldova and Romania. 95 Belarus, Bulgaria, Lithuania, Poland and Russia. 96 Moldova and Romania. 97 Decision of 16 February 1999, SK 11/98, in OTK 1999, part I, poz. 5, at 81-96. 98 Id. at 93. 99 Id. at 93. 100 Decision U.br.223/97, of 24 December 1998, Bull. Const. Case Law 1998 (3): 487-88, MKD-1998-3009. 101 Decision U.br.114/99, of 10 November 1999, Bull. Const. Case Law 1999 (3): 464-65, MKD-1999-3010. 102 Poland art. 85. 103 Slovakia art. 25. See also Croatia art. 47, Czech Charter art. 15 (3), Estonia Art. 47, Russia Art. 59, Slovenia Art. 123.
326 104
ENDNOTES
Hungary art. 70H, Lithuania art. 139. Belarus art. 57, Bulgaria art. 59. This is indirectly stated also in three constitutions that mention that service alternative to military service (thus noting such a possibility) shall not be considered to be forced labor: Romania art. 39, Moldova art. 44 (2), Ukraine art. 43. 106 Decision IV.US 81/95 of 18 September 1995, http://www.concourt.cz/angl_verze/doc/4-81-95.html. 107 Art. 31 (2): “No one may again be tried for an act for which he was already sentences and for which a final court judgment was passed”. 108 Czech Charter art. 15 (3). 109 Decision IV.US 81/95 of 18 September 1995. 110 Decision Pl.US 18/98, of 2 June 1999, see Bull. Const. Case Law 1999 (1): 44-45, CZE-1999-007. 111 Decision Pl.US 19/98, Bull. Const. Case Law 1999 (1): 39-40, CZE-1999-1-003. 112 Decision Pl.US 18/95 of 24 May 1995, the original text in http://www.concourt.sk/S/nal_uzn/ 1995/15_95s.htm. 113 Article 25(2) of the Slovak Constitution: "No one must be forced to perform military service if this runs counter to his conscience or religious belief. The details will be specified in a law." 114 Art. 24. 115 Decision no. U-I-20/1992, decided 18 February 1998, summarized in Bull. Const. Case Law (1) 1998, CRO-1998-1-004. 116 Decision U-I-48/94 of 25 May 1995, at http://www.us-rs.si/en/casefr.html. 117 Art. 123. 118 Section 7 of Decision U-I-48/94. 119 This is the case of 17 out of 22 constitutions of the region, namely: Bosnia and Herzegovina art 8 of the European Convention of Human Rights; Bulgaria art. 32; the Czech Charter art. 7; Estonia art. 26; Latvia art. 96; Macedonia art. 25; Moldova art. 28; Poland art. 47; Romania art. 26; Russia art. 23; Slovakia art. 19; Slovenia art. 35; the Ukraine art. 32; Serbia and Montenegro Charter art. 24; Montenegro (art. 20 and Serbia art. 18. 120 Czech Charter art. 7 (1). 121 Moldova (art. 28). 122 Albania art. 36; Bosnia and Herzegovina art. 8 of the ECHR; Belarus art. 28; Bulgaria art. 34; Croatia art. 36; the Czech Charter art. 13; Estonia art. 43; Georgia art. 20; Latvia art. 96; Macedonia art. 17; Moldova art. 30; Poland art. 49; Romania art. 28; Russia art. 23; Slovakia art. 22; Slovenia art. 37; Ukraine art. 31; Serbia and Montenegro Charter art. 24; Montenegro art. 30; Serbia art. 19. 123 Czech Charter art. 13. 124 Constitution of Moldova art. 30. 125 Croatia art. 37; Czech Charter art. 10; Macedonia art. 18; Slovenia art. 38 and Serbia art. 20. 126 Art. 37 (1). 127 Belarus art. 50 (1): “Everyone shall have the right to preserve his ethnic affiliation, and equally, no one may be compelled to define or indicate his ethnic affiliation”. 128 Article 48 of the Serbia and Montenegro Charter: “No one shall be obliged to reveal his/her national affiliation”. Also present in the constitutions of: Russia art. 26, as well as Montenegro art. 34 and Serbia art. 49. 129 This chapter, section 1, pp. 128-29. 130 Decision 19/97, of 21 November 1997, summarized in Bull. Const. Case Law 1997 (3): 357-358, BUL-1997-3-004. 131 Decision no. 4 of 4 November 1993, translated in East Europ. Case Rep. 2 (1995): 141-47. 132 Art. 33 of the Constitution. 133 Decision 27/96 of 27 June 1996, http://www.cecl.gr/RigasNetwork/databank/Jurisprudence /FYROM/ Jur_fyrom.htm, also summarized in Bull. Const. Case Law 1996 (2) at 286-87, MKD-96-2-005. 134 Decision U-I-25/95, of 27 November 1997, at http://www.us-rs.si/en/casefr.html. 135 Id., para. 57. 136 Id, paras. 53-54. 137 Id. para. 83. 138 Decisions U-I-158/95 and U-I-144/96 of 2 April 1998, at http://www.us-rs.si/en/casefr.html. 139 Decision 15/1991 of 13 April 1991, reprinted in Sólyom & Brunner, supra note 25 at 139-50 140 Id. at 140. 141 Id. at 149. 105
ENDNOTES 142
327
Sólyom, “Introduction”, supra note 25 at 9. Decision no. 46/1995, see “Constitution Watch: Hungary”, East Europ. Constit. Rev. 4:3 (Summer 1995): 10-13 at 12. 144 “Constitution Watch: Hungary”, East Europ. Constit. Rev. 2:2 (Spring 1993):7-8 at 8. 145 William J. Prosser, “Privacy [a legal analysis]”, in Ferdinand D. Schoeman (ed.) Philosophical Dimensions of Privacy (Cambridge University Press: Cambridge 1984): 104-55. 146 Decision PL.US43/95, of 10 September 1996, summarized in Bull. Const. Case Law 1996 (3): 404405, SVK-96-3-006. It should be mentioned that the right to privacy is one of three independent grounds of the Court's decision, the other two being the right to equality before the law and the principle of independence of the judiciary. 147 Art. 19(2). 148 Albania art. 38, Bosnia and Herzegovina para. 3, Bulgaria art. 35 (1), Poland art. 52 (1) and Slovenia art. 32. In addition, whilst not expressly stated in the provision enshrining the right to freedom of movement and residence, the Constitutions of the Czech Republic art. 42 (2) and Slovakia art. 52 (2) grant all rights to foreigners, unless this is expressly excluded (which was not done here). 149 Croatia art. 32 (1); Estonia art. 34; Georgia art. 22(1); Hungary art. 58 (1); Latvia art. 97; Russia art. 27 (1), and the Ukraine art. 33. 150 Belarus art. 30, Lithuania art. 32 (1), Macedonia art. 27 (1), Moldova art. 27 (1), Montenegro art. 28, Romania art. 25, Serbia art. 17. 151 The exception is the Constitution of Montenegro, and note that the Constitution of Albania only contains the right to leave the country. 152 The Constitutions of Belarus (art. 30), Croatia (art. 32), Lithuania (art. 32), Macedonia (art. 27), Moldova (art. 27), Romania (art. 25(2)), Serbia (art. 17). 153 The Constitutions of Bosnia and Herzegovina Art 2 and 3 of the 4th Protocol to ECHR, Bulgaria art. 35, Czech Rep art. 14, Estonia art. 35, Georgia art. 22, Hungary art. 58(1) and 69(2), Latvia art. 98, Poland art. 52, Russia art. 27(2), Slovakia art. 23 and Ukraine art. 33. The Constitution of Slovenia art. 32 is phrased somewhat differently in that it states that everyone may leave and everyone may return. However, it fits more into this category (everyone may leave but only citizens may enter) because it allows for entry by aliens to be limited by law without putting any restrictions on when or how the law may do this. Thus, it seems that the Constitutional right of entry for aliens is very weak. 154 Art. 14 of the Czech Charter. 155 Decision no. Pl.US 25/97 of 13 May 1998, summarized in Bull. Const. Case Law 1998 (2): 213-14, CZE-1998-2-007. 156 Act no.123/1992 on the Residence of Foreign Nationals. Apart from the prohibition of residence provision, the Court was also asked to examine, and eventually struck down, another provision of the same law, under which the filing of an appeal against the ban on residence did not have a suspensive effect on the decision that had been made. This was found to be contrary to various constitutional principles of fair trial. 157 Decision no. 3-4-1-9-11998 of 25 November 1998, http://www.nc.ee/english/const/98/4a9809i.html. The law in question was the Police Service Act of 1998, Articles 15 and 17. 158 Id. 159 Decision no. 139 of 14 December 1994, http://www.cecl.gr/RigasNetwork/databank/Jurisprudence/ romania/Jur_romania.htm. 160 No. 50 of 12 August 1994. 161 Article 25 (1) of the Constitution: “The right of free movement within the national territory and abroad is guaranteed. The law shall lay down the conditions for the exercise of this right”. 162 Article 49 of the Constitution states that law may restrict rights “only if absolutely unavoidable, as the case may be, for: the defence of national security, public order, health, or morals, or the citizens’ rights and freedoms. As required for conducting a criminal investigation; and for the prevention of the consequences of a natural calamity or extremely grave disaster”. 163 Decision No 47 of 1993. 164 Interview with Mr Horatiu Dumitru (a private lawyer at the time of the interview), Bucharest, 10 March 2001. 165 Art. 49 (2) of the Constitution. 166 Decision of 15 January 1998, see http://ks.rfnet.ru./english/codicese.htm. 143
328 167
ENDNOTES
Suren Avanesyan, “Constitutional Protection for Human Rights in the Russian Federation”, Journal of East European Law 6 (1999): 437-68 at 452. 168 Decisions of 25 April 1995, of 4 April 1996, see Jean-Piere Massias, Droit constitutionnel des États d'Europe de l'Est (Presses Universitaires de France: Paris, 1999) at 387, and of 2 February 1998, the Russian text of the decision on file with the author. 169 Interview with Ms Ekaterina Gezenkhadze, counsellor at the Constitutional Court of Russian Federation, 19 November 2001. 170 The only exception is Romania, which confines the general applicability of constitutional rights and freedoms to “all citizens” only (art. 15 (1)) and in a separate article promises to aliens and stateless persons living in Romania a “general protection of person and assets, as guaranteed by the Constitution and other laws”, it then goes on to spell out whether particular constitutional rights are granted to “all persons” or to “every citizen”. 171 Art. 37 (1). 172 Art. 37 (2). Similarly Moldova art. 19 (1), Russia art. 62 (3), Bulgaria art. 26 (2), Belarus art. 11, Macedonia art. 29 (1), Slovenia art. 13, Ukraine art. 26. 173 Art. 19. 174 Art. 32 (1). Similarly: Latvia, Czech Charter, Slovakia, Hungary, 175 Estonia art. 9 (1), Serbia and Montenegro Charter (article 3); Croatia art. 15 (1). 176 A representative example of such a provision can be found in the constitution of Lithuania, Article 33: “(3) A citizen may not be prohibited from returning to Lithuania. (4) Every Lithuanian person may settle in Lithuania”. The other examples are: Croatia art. 32 (2); Czech Charter art. 14; Estonia art. 36; Georgia art. 22; Hungary art. 69; Latvia art. 98; Lithuania art. 32; Romania art. 25; Slovakia art. 23 (4) and Serbia art. 17. 177 Lithuania art. 33 (2); Serbia art. 48. 178 Albania art. 59 (1) (but note that this “right” is formulated as a goal of state policy in the Albanian constitution) and Russia art. 40 (3). The Russian Constitution, to be sure, formulates a “right to a home” as a right of “everyone” (art. 40 (1)) but the goes on to specify the right of “low income citizens” to state provided free or low-rent accommodation, art. 40 (3). 179 Croatia art. 44, Czech Charter art. 21, Georgia art. 29, Latvia art. 101, Lithuania art. 33, Slovakia art. 30. 180 Albania, Hungary, Romania and Serbia. 181 The right to assembly is limited to citizens in the constitutions of Croatia art. 42; Lithuania art. 36; Montenegro art. 39 and Serbia art. 43. The right of association restricted to citizens: Croatia art. 43; Lithuania art. 35; Romania art. 37; Montenegro art. 40. 182 See the constitutions of Latvia art. 97; Lithuania art. 32; Montenegro art. 28 and Serbia art. 17. 183 Croatia art. 57; Czech Charter art. 30; Estonia art. 28; Hungary art. 70 E; Lithuania (article 52); Romania art. 43; Slovakia art. 35(3); Montenegro art. 55 and Serbia art. 39. Albania art. 52 and Latvia art. 109 extend this protection to all people. 184 Albania art. 55; Croatia art. 58; Czech Charter art. 31; Lithuania art. 53; Slovakia art. 40. 185 Czech Charter art. 33; Georgia art. 35; Hungary art. 70 F; Serbia art. 32; and, in relation to higher education, Lithuania art. 41. 186 For example, art. 35(1) of Georgia's Constitution: “Each citizen has the right to education. Freedom of choice in education is guaranteed”. 187 For example, art. 33 of the Czech Charter, states: “(1) Everyone has the right to education. School attendance shall be obligatory for the period specified by law. (2) Citizens have the right to free elementary and secondary school education …” [my italics]. 188 Bulgaria art. 22; Lithuania art. 47 (1); Romania art. 41; Russia art. 36; Slovenia art. 68 (2); Croatia art. 48 (3); Estonia art. 32 (3); Macedonia art. 31. 189 Art. 69 (2), emphasis added. 190 Art. 44, both emphases added. 191 There is no right to vote guaranteed in the sparse constitution of Bosnia and Herzegovina. 192 Bulgaria art. 42 (1). 193 This example is from art. 34 (3) Lithuanian constitution. 194 Decision 18/98 of 15 October 1998, summarised in Bull. Constit. Case Law 3 (1998): 460-62, SVK1998-3-010.
ENDNOTES 195
329
Decision Pl. US 25/96, translated in E. Europ. Case Reporter 5 (1998): 159-75, see also http://www.concourt.cz/angl_verze/doc/p-25-96.html. 196 Decision 2/1992 of 30 June 1992, translated in E. Europ. Case Reporter 2 (1995): 229-36. 197 Art. 8 (1): “Pluralism in Romanian society is a condition and a safeguard of constitutional democracy”. 198 Art. 8 (2): “Political parties may be constituted and pursue their activities in accordance with the law. They contribute to the definition and expression of the political will of the citizens, while observing national sovereignty, territorial integrity, the legal order and the principles of democracy”. 199 Decision U-I-106/95 of 25 January 1996, summarised in Bull. Constit. Case Law 1996 (1): 74-77, SLO-96-1-002. 200 Decision no. 16/1994 of 25 March 1994, East Europ. Case Rep. 1 (1994): 245-46. 201 Decision no. 2/97 of 12 March 1997, http://www.cecl.gr/RigasNetwork/databank/Jurisprudence/ jurisprudence_main.htm. 202 Decision numbers 03/3600-97, 03-3808-97, 1-12/98 of 26 January 1998, summarised in Bull. Constit. Case Law 1998 (1): 146-48. 203 Decision no. 2/1992 of 30 June 1992, East Europ. Case Reporter 2 (1995): 229-36. 204 Art. 49 (1). 205 The exceptions being the constitutions of Albania, Belarus, Bosnia and Herzegovina, and Georgia. 206 For example, art. 45 of the Bulgarian constitution states: “Citizens have the right to lodge complaints, proposals, and petitions with the state authorities”. Similar examples are also present in the constitutions of Estonia art. 46; Hungary art. 64; Lithuania art. 33; Poland art. 63; and Russia art. 33. 207 For example, the constitution of Croatia states in art. 46: “All citizens have the right to submit petitions and complaints, to make proposals to government and other public bodies, and to receive answers from them” [emphasis added]. The other constitution that follows this model is that of Latvia art. 104. 208 See, e.g., Serbia art. 48. 209 Art. 47 (2). Moldova’s Constitution has a similar provision, art. 52. 210 Art. 47 (3). 211 Slovakia art. 27 (2), Czech Charter art. 18 (3). 212 Slovakia art. 27 (3), Czech Charter art. 18 (2). 213 Croatia art. 42. The other examples of constitutions that contain only such a stark, basic, provision, are those of Bosnia and Herzegovina art. 11 of ECHR, Hungary art. 62, Moldova art. 40), Lithuania art. 36; Romania art. 36; Russia art. 31 and Slovenia art. 42. 214 Estonia art. 47, Bulgaria art. 43 (3) (this requirement applies in Bulgaria only to the meetings held indoors). 215 Latvia art. 103, Georgia art. 25, Ukraine art. 39, Montenegro art. 39 and Serbia art. 43. 216 Macedonia art. 21 (2). 217 Czech Charter art. 19 (2). 218 Thus, the constitution of Croatia states, in art. 43 (1), that: “Citizens are guaranteed the right to free association for the purposes of protection of their interests or promotion of social, economic, political, national, cultural, and other convictions and objectives”. See also Ukraine art. 36. 219 Lithuania art. 35 (2), Russia art. 30 and Ukraine art. 36. 220 The constitution of Montenegro, in article 40, states: “The state shall offer assistance to political, trade union and other associations whenever there is a public interest thereof”. 221 The constitutions of Albania art. 46; Bulgaria art. 44, Croatia art. 43, Georgia art. 26 and Macedonia art, 20. 222 Georgia art. 26 (3), Bulgaria art. 44 (2). 223 Bulgaria art. 44 (2). 224 Poland art. 58 (2). 225 Decision No. 20 of 10 May 1996, reprinted in East European Case Reporter of Constitutional Law 4 (1997) 57-63, see also http://www.nc.ee/english/const/96/4a9601i.html. The quotations in the main text are from the translation on that Website. 226 Similarly, the Slovenian Constitutional Court found unconstitutional a provision of the law on association that required a parent's permission for minors to join an association, Decision U-I-391/96 of 11 June 1998, translation at http://www.us-rs.si/en/casefr.html. In this decision, the Court did not annul the provision (because such an annulment would leave a legal gap as far as minors’ enrolment is association was concerned) but ordered the legislature to remove this unconstitutionality. In the same decision, the Court (acting on its own initiative) invalidated a provision, in the same statute,
330
ENDNOTES
establishing a minimum of ten members to form an association; the Court decided that three was enough. 227 Art. 48 (2). 228 Decision of 21 December 2000, see http://www.lrkt.lt/2000/r001221.htm. 229 Art. 35 (2). 230 Part II.2 of the Decision of 21 December 2000. 231 Part III of the Decision. The same argument was raised by the representative of the government, see Part IV of the Decision. 232 Part II.3.1 of the Decision. 233 Part II.3.3 and II.3.4 of the Decision. 234 See, e.g., the Constitution of the People’s Republic of Poland of 1952, art. 83. 235 E.g. in Bulgaria (art. 39: freedom of expression, art. 40: freedom of the press and other media) 236 E.g. in Poland, freedom of the press is included outside the constitutional charter of rights, in the first chapter of the Constitution (chapter entitled “The Republic”, art. 14). Freedom of expression is proclaimed in art. 54. 237 See Czech Charter art. 17 (4). 238 E.g. Bulgaria art. 40 (1), Czech Charter art. 17 (3), Poland art. 54 (2) 239 Art. 30 (3). 240 Art. 31 (5). 241 Case No. 19/94, 20 April 1995, reprinted in East Europ. Case Reporter of Constit. Law 3 (1996) 6179, http://www.lrkt.lt/angdoc.htm. 242 Decision 7/96 of 4 June 1996, Bull. Const. Case L. 1996 (2): 187-89, see also, for a discussion, Schwartz, supra note 18 at 182-83. 243 Art. 39. 244 Art. 40. 245 Art. 41. 246 See http://www.unhchr.ch/html/menu3/b/treaty2_asp.htm. 247 Art. 30 (7). 248 Art. 39 (2); see also Art. 40 (2) (the same restriction applied to freedom of the press). 249 Decision 5/92, reprinted in E.Europ. Const. Case Rep. 6 (1999): 1-17; see also http://www.concourt.cz/angl_verze/doc/p-5-92.html. 250 Art. 17. 251 Id. 252 Sólyom & Brunner, supra note 25 at 229. 253 Decision no. 30/1992 (V.18) AB, of 18 May 1992, reprinted in E.Europ. Const. Case Rep 2 (1995): 826. 254 Andras Sajo, “Hate Speech for Hostile Hungarians”, East Europ. Constit. Rev. 3:2 (Spring 1994): 8287 at 84. 255 Decision 30/1992, E.Europ. Const. Case Rep 2 (1995): 8-26 at 25. 256 Id. at 12. 257 Id. at 13. 258 Id. at 22. 259 Id. at 23. 260 Id. at 23 261 Sajo, supra note 254 at 84; Gábor Halmai, “Criminal Law as Means Against Racist Speech? The Hungarian Legal Approach”, Journal of Constitutional Law in Eastern and Central Europe 4 (1997): 41-52 at 45. 262 Decision 30/1992, at 16. 263 Decision 12/199. 264 I am grateful to Professor Renata Uitz for this insight. 265 Decision 14/2000, discussed in “Constitution Watch: Hungary”, East Europ. Constit. Rev. 9:3 (Summer 2000): 18-21 at 20-21. 266 Art. 61 (1) of the Constitution. 267 Art. 70 (A). 268 R.A.V. v. City of St. Paul, 505 U.S. 377 (1992).
ENDNOTES 269
331
Decision 13/2000, discussed in “Constitution Watch: Hungary”, East Europ. Constit. Rev. 9:3 (Summer 2000): 18-21 at 20-21. 270 Texas v. Johnson, 491 U.S. 397 (1989), United States v. Eichman, 496 U.S. 310 (1990). 271 Decision K. 17/93, reprinted in East European Case Reporter of Constitutional Law 5 (1998) 55-66. 272 For a critique of the law, see Irena Grudzinska Gross, “Broadcasting Values”, East Europ. Constit. Rev. 2:3 (Summer 1993): 51-53. 273 Decision K. 17/93 at 61. 274 See Grudzinska Gross, supra note 272 at 53. 275 The law was criticised by Helsinki Watch, an international human rights NGO, which said that the provisions would “chill legitimate speech as broadcasters are forced to censor themselves to fit within the undefined boundaries of the law”, quoted in Mark F. Brzezinski, “Constitutionalism and PostCommunist Polish Politics”, Loy. L.A. Int'l & Comp. L.J. 20 (1998): 433-53 at 445 n. 38. 276 Decision W. 3/93. 277 Decision Pl. US 43/93, of 12 April 1994, in East European Case Reporter of Constitutional Law 5 (1998): 33-46. 278 Interview with Professor Vojtech Cepl, Justice of the Constitutional Court of Czech Republic, Prague, 21 March 2002. 279 Decision Pl. US 43/93 at 41-42 280 Id. at 42-44. 281 Id. at 44. 282 “Constitution Watch: Slovakia”, East Europ. Constit. Rev. 11:1/2 (Winter/Spring 2002): 43-46 at 4546. 283 Id. at 46. 284 Decision 36/1994, of 24 June 1994, reprinted in East European Case Reporter of Constitutional Law 3 (1996): 148-62 285 376 U.S. 254 (1964). 286 Sólyom & Brunner, supra note 25 at 12. 287 Decision no. 37/1992 (VI.8) AB of 8 June 1992, reprinted in East European Case Reporter of Constitutional Law 2 (1995) at 27-35. 288 Sólyom & Brunner, supra note 25 at 239. 289 Art. 61 (1) and (2). 290 East European Case Reporter of Constitutional Law 2 (1995) at 31. 291 See Sólyom, “Introduction” in Sólyom & Brunner, supra note 25 at 14. 292 Decision no. 21/1996, of 14 November 1996. 293 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia 11 May 2001 294 Case No. 19/94, 20 April 1995, reprinted in East European Case Reporter of Constitutional Law 3 (1996) 61-79, http://www.lrkt.lt/angdoc.htm. 295 Decision 57/2001 (XII. 5) of December 2001, described in “Constitution Watch: Hungary”, East Europ. Constit. Rev. 11:1/2 (Winter/Spring 2002): 19-22 at 21-22. 296 Not the Court’s own words but a summary in “Constitution Watch”, id. at 22. 297 Id. at 22 298 Decision of 13 February 1997, http://www.lrkt.lt/angdoc.htm, reprinted in Rulings and Decisions of the Constitutional Court of the Republic of Lithuania, vol. 8 (Constitutional Court of Lithuania: Vilnius 1997): 14-51. The references below are to this edition. 299 Interview with Constitutional Court Justice Teodora Staugaitienơ, Vilnius, 22 June 1998. 300 Decision of 13 February 1997 (printed text on file with the author) at 39-40. 301 Id. at 19. 302 Id. at 19. 303 Id. at 32-4, 39. 304 Id. at 39. 305 Id. at 42.
332
ENDNOTES
Chapter 7 1
The Political Dimension of EU Enlargement: Looking Towards Post-Accession, Report of the Reflection Group chaired by Jean-Luc Dehaene (Robert Schuman Centre at the EUI, Florence 2001), http://www.iue.it/RSC/pdf/dehaenereport.pdf at18. 2 For a discussion of different “political orientations” influencing constitution-making in post-communist Poland, see Piotr Winczorek, “Axiological Foundations of the Constitution of Poland”, St. LouisWarsaw Transatlantic L.J. (1997): 59 at 61-62. 3 Jean-Marie Henckaerts & Stefaan Van der Jeught, “Human Rights Protection Under the New Constitutions of Central Europe”, 20 Loy. L.A. Int’l & Comp. L. Rev. 20 (1998): 475-506 at 491. 4 See, e.g. the recent public opinion survey by the reputable OBOP institute in Poland, in June 2002. 65% of respondents believe that the State should look after the welfare of its citizens; 53 % believe that budget expenditure for social-welfare purposes is more important than spending to stimulate economic growth; 62 % believe that the State should subsidise employment if this is necessary to fight unemployment, even if it is not economically profitable. See “Opinia spoáeczna: Miáo juĪ byáo”, Rzeczpospolita (Warsaw) 13 August 2002 at A-1, also at http://www.rzeczpospolita.pl/ wydanie_020813/ publicystyka/publicystyka_a_3.html. 5 Tadeusz ZieliĔski, Panel discussion, in Konstytucja w sáuĪbie demokracji; Constitution in Service of Democracy, conference papers: The International Centre for Development of Democracy Foundation, 10-12 March 1995, Cracow, at 211-212. 6 Id. at 212-213. 7 Herman Schwartz, “In Defense of Aiming High”, East Europ. Constit. Rev. 1:3 (Fall 1992): 25-28 at 2627. 8 Id. at 27. 9 Id. at 28. 10 See Cécile Fabre, “Constitutionalising Social Rights”, Journal of Political Philosophy 6 (1998): 263-84 at 268-70. 11 See, in particular, Henry Shue, Basic Rights (Princeton University Press: Princeton N.J. 1980) at 39-40, 55-56. 12 This understanding is not equivalent to the notion of a “programmatic” right, because the latter requires the state to have a program. A “minimal” use of the right merely requires that, if there is a program, it must not be arbitrarily denied to some beneficiaries. 13 Schwartz, supra note 7 at 27. It is important to note that this was not the only function of socioeconomic rights prescribed by Professor Schwartz in his article. 14 With the partial exception of Scandinavian states, see footnote 32 below. 15 E.g. German Constitution, art. 20. 16 See, e.g., Constitutions of Belgium, Ireland, Italy, Luxembourg, Netherlands, Greece, Spain and Portugal. 17 See, e.g., Spain and Italy. 18 Art. 14 (1) and (2) of the EU Charter of Fundamental Rights. 19 Art. 24 (1) and (3). 20 Art. 15. 21 Art. 33 (2). 22 Art. 34. 23 Art. 35. 24 Art. 36. 25 Art. 35, 2nd sentence. 26 Art. 33 (1). 27 Art. 26. 28 Art. 25. These two latter provisions (Articles 26 and 25), which draw on the equivalent provisions of the European Social Charter and the Community Charter of the Fundamental Social Rights of Workers, use the language of “rights” ("The Union recognises and respects the rights. . .”). However, I believe that it is more accurate to view them as descriptions of policy directives; similarly Agustín José Menéndez, “The Sinews of Peace: Rights to Solidarity in the Charter of Fundamental Rights of the Union”, in Erik Oddvar Eriksen, John Erik Fossum & Agustín José Menéndez, The Chartering of
ENDNOTES
333
Europe: The Charter of Fundamental Rights in Context (ARENA Report 8/2001, Oslo 2001): 201-26 at 215. 29 See, e.g., San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 37 (1973) (finding no fundamental right to public education); Lindsey v. Normet, 405 U.S. 56, 74 (1972) (finding no fundamental right to adequate housing). 30 Jackson v. City of Joliet, 715 F.2d 1200, 1203 (7th Cir. 1983). 31 For a general discussion, see Wiktor Osiatynski, “Rights in New Constitutions of East Central Europe”, Colum. Hum. Rts. L. Rev. 26 (1994): 111-66 at 138-45. 32 This is not entirely accurate; the Nordic constitutions contain some socio-economic rights (in particular, the right to work and a commitment to full employment) but, by and large, they are not exhaustive and are not accompanied by a “social state” clause. This may be partly explained by the fact that, with the exception of the Swedish Instrument of Government (1974), they originate from the first half of the 19th Century, although they have been amended many times. As one commentator has noted, they “retain the liberal character of the time of their adoption”, see George S. Katrougalos, “The Implementation of Social Rights in Europe”, Columbia Journal of European Law 2 (1996): 277-312 at 294. 33 Ulrich K. Preuss, “Patterns of Constitutional Evolution and Change in Eastern Europe”, in Constitutional Policy and Change in Europe (J.J. Hesse & N. Johnson eds., 1995): 95-126 at 103; see also Jon Elster, “The Impact of Rights on Economic Performance”, in Andras Sajo, ed., Western Rights? PostCommunist Application (Kluwer: The Hague 1996): 347-59. 34 Preuss, supra note 33 at 101. 35 Id.; see also Andrzej Rapaczynski, “Constitutional Politics in Poland: A Report on the Constitutional Committee of the Polish Parliament”, in A.E. Dick Howard, ed., Constitution Making in Eastern Europe (Woodrow Wilson Center Press: Washington 1993): 93-132 at 107-8. 36 Andras Sajo, “How the Rule of Law Killed Hungarian Welfare Reform”, East Europ. Constit. Rev 5:1 (Winter 1996): 31-41. 37 For excerpts of one of the central decisions in this series, Decision No. 43/1995 of 30 June 1995 concerning social security benefits, see László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 322-32. For a discussion, see Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago, 2000) at 92-93; Bojan Bugaric, “Courts as Policy-Makers: Lessons from Transition”, Harv. Int’l L.J. 42 (2001): 247-88 at 251. 38 Jerzy Ciemniewski, “Sejm i Senat w projekcie Konstytucji RP”, in Józef Krukowski (ed.), Ocena projektu Konstytucji RP (Towarzystwo Naukowe Katolickiego Uniwersytetu Lubelskiego: Lublin, 1996): 37-47 at 41. 39 Leszek Garlicki, “Orzecznictwo Trybunaáu Konstytucyjnego w 1993 roku”, Przegląd Sądowy (1994, no. 10): 31-52 at 37. 40 Leszek Garlicki, “Orzecznictwo Trybunaáu Konstytucyjnego w 1997 roku”, Przegląd Sądowy (1998, no. 6): 35-58 at 55. 41 See, e.g., Jon Elster, “Constitution-Making in Eastern Europe: Rebuilding the Boat in the Open Sea”, Public Admin. 71 (Spring/Summer 1993): 169-217 at 198. 42 Herman Schwartz, in Konstytucja, supra note 5: 215-23 at 221. 43 Cass R. Sunstein, “Against Positive Rights”, East Europ. Constit. Rev. 2:1 (Winter 1993): 35-39 at 36. 44 Id. at 37. For a more recent exposition by Sunstein of his views on constitutional socio-economic rights, see Cass Sunstein, Designing Democracy: What Constitutions Do (Oxford University Press: Oxford 2001) at 221-38. 45 Albania art. 52; Belarus art. 41 (1); Croatia art. 57; Czech. Charter art. 26; Estonia art. 28; Hungary art. 70E; Latvia art. 109; Lithuania art. 52; Moldova art. 47; Poland art. 67; Romania art. 43; Slovakia art. 35 and 39; Ukraine art. 46; Serbia and Montenegro Charter art. 42, Montenegro art. 55; Serbia art. 39. 46 E.g., Poland art. 67; Montenegro art. 55; Serbia art. 39. 47 Albania art. 52; Belarus art. 47; Estonia art. 28; art. 70E; Latvia art. 109; Lithuania art. 52; Moldova art. 47(2); Poland art. 67; Slovakia art. 39 (1); Ukraine art. 46. 48 Belarus art. 47; Hungary art. 70E; Lithuania art. 52; Moldova art. 47 (2); Poland art. 67. 49 Belarus art. 47; Croatia art. 57; Hungary art. 70E; Latvia art. 109; Lithuania art. 52; Moldova art. 47(2); Poland art. 67; Romania art. 46; Slovakia art. 38 (1); Ukraine art. 46. 50 Belarus art. 47; Estonia art. 28; Lithuania art. 52; Slovakia art. 39 (1); Ukraine art. 46.
334 51
ENDNOTES
Hungary art. 70E (widows and orphans); Lithuania art. 52 (widows); Moldova arts. 47(2) (widows), 49(3) (orphans). 52 Russia, Bulgaria, Macedonia, and Slovenia all include provisions for social security. The Russian Constitution (art. 39) provides for social security in cases of old age, illness, disability, and loss of breadwinner. The Bulgarian Constitution (art. 51) provides for social security only in cases of old age, disability or temporary unemployment. The Macedonian constitution only provides for social security in cases of temporary unemployment (art. 32) and then states that all other social security rights will be determined by law (art. 34). The Slovenian constitution (art. 50) provides that all those who fulfil the conditions laid down by law will receive social security benefits. 53 Georgia, and Bosnia and Herzegovina. 54 See, generally, Viktor Mavi, “The Right to Health and the New East European Constitutions”, Journal of Constitutional Law in Eastern and Central Europe 3 (1996): 213-24. 55 Belarus art. 45; Croatia art. 58; Czech Charter art. 30; Estonia art. 28; Latvia art. 111; Lithuania art. 53 (although note that this constitutional provision uses the language of the State's duty to take care of people's health rather than of an individual right to health care); Macedonia art. 39; Moldova art. 36; Romania art. 43; Russia art. 41; Slovakia art. 40; Ukraine art. 49. 56 Poland art. 68; Serbia art. 30. 57 Bulgaria art. 52; Slovenia art. 51; Montenegro art. 55; Serbia art. 30. 58 Albania art. 55; Bulgaria art. 52; Georgia art. 37. 59 Hungary art. 70D. 60 Bulgaria art. 55, Macedonia art. 43, Slovenia art. 72. 61 Croatia art. 69. 62 Belarus art. 49; Lithuania art. 41; Moldova art. 35; Poland art. 70; Romania art. 32; Russia art. 43; Slovenia art. 57; Ukraine art. 53. 63 Albania art. 57; Bulgaria art. 53; Czech Charter art. 33; Latvia art. 112; Slovakia art. 42. 64 Croatia art. 65; Georgia art. 35; Hungary art. 70F (although it also guarantees financial support to all students); Macedonia art. 44; Serbia and Montenegro Charter art. 43, Montenegro art. 62. Two constitutions are unclear about the specific level at which free education is guaranteed: The Serbian Constitution mentions all “regular education” (art. 32), and the Estonian provides for such a right to “school-age children” (art. 36). 65 Belarus, Hungary, Poland, Romania, Russia, Slovakia, Ukraine, Montenegro and Serbia. 66 In addition, five other constitutions establish a good environment as an aim for the state, though not enforceable as a right. 67 Belarus art. 48; Russia art. 40; Slovenia art. 78; Ukraine art. 47. 68 Albania art. 59; Poland art. 75. 69 Belarus, Croatia, Czech Republic, Moldova, Poland, Romania, Russia, Slovakia and Ukraine. 70 Bulgaria, Hungary, Macedonia, Slovenia, Montenegro, and Serbia. 71 Three Baltic states: Estonia, Latvia, and Lithuania. In fact, Lithuania falls in between this and the first category, with a middling number of work-protection and other rights. 72 Bosnia and Herzegovina, and Georgia. 73 Rett R. Ludwikowski, Constitution-Making in the Region of Former Soviet Dominance (Duke University Press: Durham 1996) at 230. 74 Belarus, Bulgaria, Croatia, Estonia, Georgia, Hungary, Latvia, Lithuania, Macedonia, Romania, Russia, Ukraine, Montenegro, and Serbia. 75 For example, in Hungary. 76 In January 1991, the federal parliament of Czechoslovakia adopted the Charter of Fundamental Rights and Freedoms as a constitutional act. Even before the formal dissolution of the federation, the two republics adopted slightly different legal strategies towards the Charter, in their respective constitutions: The Slovak Republic incorporated the Charter into its constitution (of September 1992) while the Czech Constitution (of December 1992) stated that the Charter formed a part of the constitutional order of the Republic without incorporating it directly, see Czech Rep. Const. art. 112 (1). 77 Czech Charter art. 41; Slovakia art. 51. 78 Russia art. 39. 79 E.g., Czech Charter arts. 32(5)-(6) (providing assistance to parents raising their children); Slovakia art. 43(2) (providing the right of access to the cultural heritage).
ENDNOTES 80
335
Albania, Moldova, Poland, and Slovenia. Poland art. 81. 82 Id. at art. 67 (1). 83 Article 81 states that certain rights, listed earlier in the Constitution, can be claimed only within the limits defined by a particular statute. These rights include: minimum income (Art. 65 para 4), full employment and state bodies to combat unemployment (Art. 65 para 5), safety and hygiene at work (Art. 66), days free from work and annual paid holiday (Art. 66 para. 2), assistance to handicapped persons (Art. 69), protection of families and special protection of mothers (Art. 71), protection of the environment (Art. 74), satisfaction of needs of accommodation, combating homelessness and protection of tenants’ rights (Art. 75), and protection of consumers’ rights (Art. 76). 84 Zdzisáaw Czeszejko-Sochacki, Leszek Garlicki, Janusz TrzciĔski, Komentarz do ustawy o Trybunale Konstytucyjnym (Wydawnictwo Sejmowe: Warszawa, 1999) at 35. 85 The Polish text was published in Rzeczpospolita (Warsaw), 15 November 1992. For an English translation, see “Draft of the Charter of Rights and Freedom”, St. Louis-Warsaw Transatlantic L.J. (1996): 73-84. For useful discussion on this issue, see generally Stanisáaw Frankowski, “Lech WaáĊsa’s Draft of the Charter of Rights and Freedoms: An Overview”, St. Louis-Warsaw Transatlantic L.J. (1996): 65-72. This constitutional draft was eventually aborted, and disowned by President WaáĊsa himself. 86 “Draft of the Charter”, supra note 85, ch. V. 87 Id. art. 48. 88 Decision 31/1990, quoted in László Sólyom, “Introduction to the Decisions of the Constitutional Court of the Republic of Hungary”, in Sólyom & Brunner, supra note 37 at 36. 89 Péter Paczolay, “Human Rights and Minorities in Hungary”, J. of Const. L. in E. and Cent. Eur. 3 (1996): 111-26 at 121. 90 Sólyom, supra note 88 at 35. 91 Id. at 37. 92 Summarized in “Constitution Watch: Hungary”, East Europ. Constit. Rev. 9:1/2 (Winter/Spring 2000): 18-21 at 20-21. 93 As a general proposition (not in the context of CEE), the thesis of non-justiciability of the socioeconomic rights of individuals, and of the socio-economic duties of the State, was strongly refuted by K. D. Ewing, “Social Rights and Constitutional Law”, Public Law (1999): 104-23 at 119-21. 94 Czeszejko-Sochacki et al., supra note 84 at 163: a breach of a “programmatic norm” (including those that proclaim socio-economic rights, the details of which are to be spelled out by legislators) happens when “the legislator incorrectly interprets a provision of the Constitution that defines a particular goal or task of public authorities, and, in particular, enacts a statute that provides for such measures that cannot lead to that goal and thus breaches constitutional liberties or rights”. 95 Jon Elster, Claus Offe & Ulrich K. Preuss, Institutional Design in Post-communist Societies (Cambridge University Press: Cambridge, 1998) at 87. 96 Decision K. 1/88; the full text of the Decision on file with the author. 97 Id. at 7 (section III.1 of the Decision). 98 Id. at 8, section IV of the Decision. The use of the notion of “ratchet” requirement is mine, not the Tribunal's. 99 Decision K. 21/95 of 25 February 1997, discussed in Jerzy Oniszczuk, Konstytucja Rzeczypospolitej Polskiej w orzecznictwie Trybunaáu Konstytucyjnego (Zakamycze: Kraków, 2000) at457. 100 See e.g. Decision K. 7/89 of 8 November 1989. According to Professor Garlicki, there is a sharp difference between the early stage of Constitutional Tribunal jurisprudence, when this programmatic nature of socio-economic rights was very pronounced, and the later stage when the Tribunal had no doubts about treating these rights as a basis for evaluating laws, see Leszek Garlicki, “Orzecznictwo Trybunaáu Konstytucyjnego w 1993 roku”, Przegląd Sądowy (1996, no. 7-8): 110-38 at 119. 101 See Decision K 8/96, of 17 July 1996, invalidating a law that prohibited the indexation of pensions. The Tribunal announced that the non-indexation of pensions (that is, a failure to adjust them to the rise of costs of living) amounts to a violation of the pensioners’ constitutional rights to their pensions. 102 Czeszejko-Sochacki et al., supra note 84 at 163. 103 Tadeusz ZieliĔski, “Prawo do chleba, mieszkania i pracy”, Gazeta Wyborcza (Warsaw), 27August 1996 at 12. 81
336 104
ENDNOTES
Decision of the Bulgarian CC no 12/1997 of 25 Sept. 1997, described in the Bulletin on Constit. CaseLaw 3 (1997) at 357, originally published in Bulgarian in Darzhaven Vestnik no. 89 of 7.10.1997. 105 For similar reasons, the Romanian Constitutional Court invalidated, in 1998, a provision of the law on the social protection of unemployed persons. This law stated that those who had completed secondary education and were in vocational training were not eligible for unemployment benefits; Decision no. 81/1998 of 19 May 1998, summarised in Bull. Constit. Case-Law 1998 (2): 288-89, ROM-1998-2004. This was meant to remove students benefiting from student grants from the group of those eligible for unemployment benefits, but one of the consequences was that those who had been working prior to (or during) their studies, and then lost their employment while studying, were denied unemployment benefits. The Court, in invalidating the provision, argued (among other things) that the exercise of one right (education) cannot be used as the ground for curtailing another (unemployment benefit). 106 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia 11 May 2001. 107 Id. 108 Id. 109 Id. 110 Approximately US$ 32 at 2001 exchange rates. 111 Interview with Professor Todor Todorov, Justice of the Constitutional Court of Bulgaria, Sofia 11 May 2001. 112 See Venelin I. Ganev, “Bulgaria: The (Ir)Relevance of Post-communist Constitutionalism”, in Jan Zielonka (ed.), Democratic Consolidation in Eastern Europe, vol. 1: Institutional Engineering (Oxford University Press, Oxford 2001): 186-211 at 186, 198. 113 Decision U-I-86/96 of 12 December 1996, English translation available at http://www.usrs.si/en/casefr.html. 114 Section 10 of the Decision. 115 See Mark Tushnet, Taking the Constitution Away from the Courts (Princeton University Press: Princeton, 1999) at 169-72. 116 The Political Dimension of EU Enlargement, supra note 1 at 38. 117 E.g. in Poland in 2001, around 18% of the labour force was unemployed; in the Czech Republic, 8.2%; in Hungary, 5.8%, OECD Quarterly Labour Force Statistics, No. 1 (2002), Paris. 118 Bob Hepple, “A Right to Work?”, Industrial Law Journal 10 (1981): 65-83 at 73. 119 They are those of Albania (art. 49), Belarus (art. 41), Bulgaria (art. 48), Croatia (art. 54), the Czech Republic (art. 26), Hungary (art. 70B), Macedonia (art. 32), Slovenia (art. 49), Romania (art. 38), Slovakia (art. 35), Ukraine (art. 43), Serbia and Montenegro Charter (art. 40), Montenegro (art. 52), and Serbia (art. 35). 120 See Sólyom, “Introduction”, supra note 88 at 35. 121 Estonia (art. 29), Georgia (art. 30), Latvia (art. 106), Lithuania (art. 48), Moldova (art. 43), Poland (art. 65), Russia (art. 37). The only Constitution which contains no work-related rights is that of Bosnia and Herzegovina. 122 Five constitutions contain a general provision granting the right to be trained for work, namely, the Constitutions of Belarus (art. 41) (although this does only apply to those who are unemployed through no fault of their own), the Czech Republic (Czech Charter art. 26), Estonia (art. 29(3)), Slovakia (art. 35), and Ukraine (art. 43). In addition, some constitutions grant the right to be trained to specific groups in society. Thus, the Czech Charter (art.29) and Slovak Const. (art. 38) provide for training for the young. Those of Moldova (art. 51), Poland (art. 69), Romania (art. 46), and Slovenia (art. 52) provide for training for the disabled. The Slovak Const. (art. 38) also provides for training for the unhealthy. 123 The quotation is from the Const. of Bulgaria, art. 48; see also similarly Slovenia (art. 66) and Ukraine (art. 43). The Const. of Georgia has a slightly different type of provision, stating at art. 32, inter alia, that “The state must help the unemployed to find work”. 124 Albania art. 59, Belarus art. 41, Poland (art. 65). 125 The old Constitution (of 1952) contained article 68, which was kept in force until 1997: “Citizens . . . shall have the right to work, that is, the right to employment paid in accordance with the quantity and quality of the work done”. 126 Decision K. 14/91 of 11 February 1992.
ENDNOTES
337
127
Decision K. 7/96 of 7 January 1997. Art. 65. 129 Decision K. 33/98 of 26 April 1999. 130 See discussion of this case in Oniszczuk, supra note 99 at 466. 131 Decision 8/1993 of 27 February 1993, transl. in East Europ. Case Reporter of Const. Law 1 (1994): 247-48. 132 Decision U-I-344/94, of 1 June 1995, transl. in http://www.us-rs.si/en/casefr.html. 133 Section 6 of the Decision. As the text of the Decision further reveals, this was in response to one of the petitioner’s arguments, that the conditions for appointment to the position of notary mean, in the words of the Court, “a reintroduction of the former criteria on socio-political (un)suitability”, Section 9 of the Decision. 134 Section 14 of the Decision. 135 Section 12 of the Decision. 136 Decision of 4 March 1999, http://www.lrkt.lt/1999/n9a0304a.htm. 137 See Chapter 10.3, pp. 279. 138 Decision of 12 February 2002, http://www.lrkt.lt/angdoc.htm. 139 Decision no. U-I-222/1995 of 9 November 1998, summarised in Bull. Constit. Case-Law 1998 (3): 403, CRO-1998-3-018. Another, much more marginal, example was the decision of the Polish Constitutional Tribunal that upheld a provision of a statute on anti-alcohol measures, against a challenge of inconsistency with the state’s constitutional duty to protect public health (under an old Constitution, superseded in 1997), Decision no. K. 3/97 of 23 June 1997. The challengers (a group of MPs) claimed that a provision on temporary, one-off licenses to sell alcohol during public open-air events was too lax and, as a result, did not guarantee sufficient protection against alcohol abuse, thus unconstitutionally endangering public health. The Tribunal rejected this claim on the basis that the provision under challenge did not constitute a “drastic” breach of the legislature’s duties stemming from the constitutional provisions on health protection. However, note that this decision, strictly speaking, was not made under a “right to health” heading. 140 Decision SK 18/99 of 8 November 2000, see Wybór tez i sentencji OrzeczeĔ Trybunaáu Konstytucyjnego, II Póárocze 2000 (Wydawnictwo TK, Warszawa 2001), at 30-33. 141 Note that, in Polish, the word “schools” includes also tertiary education institutions such as universities. 142 See, e.g., Decision K. 8/96 of 17 July 1996. 143 The Croatian Constitutional Court invalidated, in 1998, a provision of the 1993 Code on Equating Retirement Incomes, on the basis that the code demanded that pensions increase relative to changes in the cost of living, rather than relative to the increase of average incomes, see “Constitution Watch: Croatia”, East Europ. Constit. Rev. 7: 3 (Summer 1998): 8-9 at 9. 144 See The Political Dimension of EU Enlargement, supra note 1 at 19. 145 See text accompanying notes 140-41 above. 146 Andras Sajo, “Welfare Rights in the Post-Communist Constitutional Experience”, in Mihaela Serban Rose, ed., Constitutionalism in Transition: Africa and Eastern Europe (The Helsinki Foundation for Human Rights: Warsaw, 2003): 41-73. 147 See Ronald Dworkin, Sovereign Virtue (Harvard University Press: Cambridge Mass., 2000), chapters 8 and 9. 148 See Sajo, supra note 146 at 54. 149 See pp. 183-84. 150 Sajo, supra note 146 at 61 and 64. 151 Id. at 55. 152 See János Mátyas Kovács, “Approaching the EU and Reaching the US? Rival Narratives on Transforming Welfare Regimes in East-Central Europe”, in Peter Mair & Jan Zielonka, eds, The Enlarged European Union: Diversity and Adaptation (Frank Cass: London, 2002): 175-204 at 197. 128
Chapter 8 1
See, e.g., with reference to Bulgarian Constitutional Court, Venelin Ganev, “Bulgaria: The (Ir)Relevance of Post-comunist Constitutionalism”, in Jan Zielonka, ed., Democratic Consolidation in
338
ENDNOTES
Eastern Europe, vol. I: Institutional Engineering (Oxford University Press: Oxford, 2001): 186-211 at 198. 2 Art. 32. 3 Art. 89. 4 For statistics on anti-homosexual views in Poland, see e.g. Helsinki Committee for Human Rights, “Gender Equality: Legal and Institutional Framework On Women’s Rights and Equal Opportunities; De Jure And De Facto Discrimination In Poland”, Polish Law Journal 6 (2001): 149-228 at 216 (for instance, according to a 1996 survey, only 25 percent of respondents said that homosexuals should be allowed to hold high public offices, while as many as 63 percent would not accept homosexuals in such positions. 71 percent of respondents would not permit homosexuals to be teachers; 71 percent excluded the possibility of a homosexual marriage, and 88 percent would not permit adoption by same-sex couples, id. at 216-17. 5 Decision no. 81 of 15 July 1994, striking down the Criminal Code's prohibition of homosexual intercourse. 6 Decision 14/1995 of 15 March 1995, striking down a rule of the civil code that defined "domestic partnership" as a woman and a man living together in a common household outside marriage. 7 This is the case of eight constitutions in the region: Bosnia Herzegovina (in the ECHR), Croatia, the Czech Republic, Estonia, Hungary, Russia, Slovakia and Serbia and Montenegro Charter. In addition, two other constitutions (those of Slovenia and Serbia) state that “any other personal reasons” are also impermissible grounds for discrimination. 8 See, generally, Wojciech Sadurski, “The Concept of Legal Equality and an Underlying Theory of Discrimination”, Saint Louis-Warsaw Transatlantic L. J. (1998): 63-104. 9 Decision K. 6/89 of 24 October 1989 (the text on file with the author) at 6. I should add that, in this context, the Constitutional Tribunal is referring to my own book, Wojciech Sadurski, Teoria sprawiedliwoĞci (PWN: Warszawa 1988) at 94. 10 Decision U. 1/96 of 16 December 1996, discussed in Aldona DomaĔska, “Analiza treĞci konstytucyjnej zasady równoĞci w oparciu o wybrane orzeczenia Trybunaáu Konstytucyjnego”, Studia PrawnoEkonomiczne 62 (2000): 47-58 at 51. 11 See, e.g., Decision K. 7/90 of 22 August 1990, discussed id. at 53. 12 Decision K. 14/91 of 11 February 1992, discussed id. at 53. 13 Decision no. P. 2/87 of 3 March 1987 (the text of the decision on file with the author), pp. 12-13 (italics added). 14 DomaĔska, supra note 10 at 52. 15 Decision No 47 of 17 May, 1994. 16 Id. 17 Id. 18 Decision U-I-107/96, of 5 December 1996. 19 Id., para. 16. 20 Id., para 28. 21 Ken Jowitt, New World Disorder. The Leninist Extinction (University of California Press: Berkeley, 1992) at 322. 22 On Poland, see Helsinki Committee, supra note 4 at 156. According to Eleonora ZieliĔska of the University of Warsaw, the difference in average wages between men and women in Poland varies between 30 and 40 percent, see “Praw kobiet nie wprowadzimy czarodziejską róĪdĪką” (Interview with Professor ZieliĔska), Rzeczpospolita (Warsaw) 17 April 2002, at A9; on Czech Republic, see 2002 Regular Report on Czech Republic’s Progress Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC (2002) 1402, available at http://europa.eu.in/comm/ enlargement/report2002/cz_en.pdf, at 30. Unequal treatment of women as regards working conditions, remuneration, professional training and career opportunities has been noted by the EU Commission regarding almost every CEE candidate state in the annual reports on these countries’ progress towards accession; see, e.g., for Slovakia, 2002 Regular Report on Slovakia’s Progress Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC (2002) 1410, available at http://europa.eu.in/comm/ enlargement/report2002/sk_en.pdf, at 29. 23 Helsinki Committee, supra note 4 at 163. 24 For statistics in Poland, see id. at 190-91. For some other examples: In Romania only 11 percent of deputies and 9 percent of senators are women; see 2002 Regular Report on Romania’s Progress
ENDNOTES
339
Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC (2002) 1409, available at http://europa.eu.in/comm/enlargement/report2002/ro_en.pdf, at 34. 25 Decision Kw. 5/91 of 24 September 1991. 26 Decision K. 15/97 of 29 September 1997; Decision K 35/99 of 5 December 2000. 27 Decision K 15/99 of 13 June 2000. 28 Decision K 27/99 of 28 March 2000. 29 Decision K. 15/97 of 29 September 1997, Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1997 (C.H. Beck: Warszawa 1998): 367-86. 30 The decision of the Supreme Court of 14 May 1996, discussed and cited in the decision of Constitutional Tribunal 15/97, id. at 373-74. The Supreme Court had considered the matter not from the point of view of the constitutionality of the relevant rule, but as a top judicial appellate body, in the process of so-called “extraordinary appeal” from a decision of the Supreme Administrative Court (NSA), which had considered the matter in 1993. 31 Id., cited at 373-74. 32 Id. at 376, quoting its decision P. 2/87 of 3 March 1987. 33 Id. at 378 34 Id. at 378. 35 Id. at 381. 36 Id. at 381. 37 Id. at 381. 38 Id. at 382. 39 Decision 9/1990, discussed in Kim Lane Scheppele, “Women's Rights in Eastern Europe”, East Europ. Constit. Rev. 4:1 (Winter 1995): 66-69 at 69. 40 Id. at 69. 41 For examples of such publicly stated views in Poland, see Helsinki Committee, supra note 4 at 218-19. 42 Decision 14/1995 of 13 March 1995, translated in East European Case Reporter of Constitutional Law 2 (1995) at 194-200, and in László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000) at 316-22. The references below are to the latter edition. 43 Id. at 318. 44 Id. at 320. 45 Id. at 320. 46 Decision of 25 June 1999, no. 20/1999, summarized in Bull. Const. Case-Law 1999 (3): 389-90, no. HUN-1999-3-005. 47 Decision no. 81 of 15 July 1994, Curtea ConstituĠională - Decizii de Constatare a NeconstituĠionalităĠii, 1992-1998 (Editura Militarăa: Bucureúti, 1999): 335-39. The quotations that follow are from the English translation of the Decision, on file with the author. 48 Articles 11 (treaties ratified by the Parliament become part of the domestic law), 20 (the precedence of ratified international covenants on human rights over domestic law, in cases of conflict) and 26 (the right to privacy). 49 In the words of the Court, art. 200 was unconstitutional in so far as it applied “to same-sex relations between adult consenting persons, that are not committed in public and do not produce public scandal”. 50 E.g. the Charter of Fundamental Rights and Freedoms [hereinafter: Czech Charter] Art. 29. 51 E.g. the Czech Charter art. 29 52 E.g. Ukraine art. 24. 53 E.g. Hungary art. 66. 54 E.g. Bulgaria art. 47. 55 E.g. Romania art. 46. 56 Art. 29 (2). 57 Art. 6. 58 Art. 34 (3). 59 Art. 70 A. 60 Decision No. 9/1990 (IV.25), discussed in Peter Paczolay, “Human Rights and Minorities in Hungary”, Journal of Constitutional Law in Eastern and Central Europe 3 (1996): 111-26 at 114-15. 61 Quoted id. at 115.
340 62
ENDNOTES
Quoted id. at 115. Id. at 116. 64 See Ronald Dworkin, Taking Rights Seriously (Duckworth: London, 1977) at 223-39. 65 See Andras Sajo, “Protecting Nation States and National Minorities: A Modest Case for Nationalism in Eastern Europe”, U. Chi. L. Sch. Roundtable (1993): 53-74. 66 See Peter S. Green, “Roma Seeking Sense of Unity to Combat Racial Bias”, New York Times 10 May 2002. The marginalisation, victimisation, economic and political discrimination and extremely poor living conditions of the Roma minority have been noted by the EU Commission regarding almost every CEE candidate state in the annual reports on these countries’ progress towards accession; see, e.g., for Slovakia, 2002 Regular Report on Slovakia’s Progress Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC(2002) 1410, available at http://europa.eu.in/comm/enlargement/report2002/sk_en.pdf, at 30; for Lithuania, 2002 Regular Report on Lithuania’s Progress Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC(2002) 1406, available at http://europa.eu.in/comm/ enlargement/report2002/lt_en.pdf, at 30; for Hungary, see 2002 Regular Report on Hungary’s Progress Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC(2002) 1404, available at http://europa.eu.in/comm/enlargement/report2002/hu_en.pdf, at 31; on Czech Republic, see 2002 Regular Report on Czech Republic’s Progress Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC(2002) 1402, available at http://europa.eu.in/comm/enlargement/report2002/cz_en.pdf, at 3-32. 67 There are some minor exceptions to this rule; for example, in the 2002 parliamentary elections, four Roma members of Parliament of Hungary were elected. 68 Sajo, supra note 65 at 54. 69 See Wiktor Osiatynski, “Rights in New Constitutions of East Central Europe”, Columbia Human Rights Law Review 26 (1994): 111-166 at 134-35. 70 As one observer has noted: “Hungarian politics fosters the interests of Hungarian minorities living abroad rather than those of non-Hungarian minorities who live on the territory of the Hungarian state”, Osiatynski id. 137. These words written in 1994 became even more valid recently, with the controversial “status law” adopted by the Hungarian parliament in June 2001: the law provides for rights and certain preferences for ethnic Hungarians who live beyond Hungary’s borders, such as the right to work in Hungary for a three-month period each year, financial support for publictransportation costs as well as assistance for ethnic-Hungarian students from neighbouring states to study in universities in Hungary, and also assistance to ethnic Hungarians who live in their home countries who have more than two children in Hungarian-language schools, see “Constitution Watch: Hungary”, East Europ. Constit. Rev. 10:4 (Fall 2001): 18-21 at 18-19. 71 They were not all ethnic Russians but also other nationalities for whom Russian was the mother tongue (such as Belorussians, Ukrainians etc). 72 Vello Pettai, “Democratic Norm Building and Constitutional Discourse Formation”, paper presented at the workshop “Rethinking the Rule of Law in Post-Communist Europe”, European University Institute, Florence 22-23 February 2002 (text on file with the author), at 23. 73 For details, see Antonina Zhelyazkova, “The Bulgarian Ethnic Model”, East Europ. Constit. Rev. 10 (Fall 2001): 62-66 at 62-63. 74 In 1992, the population census showed Turks constituting 9.4% of population at that time, and Roma, 3.7%, see id. at 66. 75 See Nikolai Petrov, “Political Institutions and the Regulation of Ethnic Conflicts: Russia's Experience”, paper presented at Conference on “Legal Framework to Facilitate the Settlement of Ethno-Political Conflicts in Europe”, Baku 11-12 January 2002 (text on file with the author), at pp. 2-3. 76 Id. at 7. 77 Id. at 3. 78 Will Kymlicka, "Western Political Theory and Ethnic Relations in Eastern Europe", in Will Kymlicka & Magda Opalski, eds., Can Liberal Pluralism be Exported? Western Political Theory and Ethnic Relations in Eastern Europe (Oxford University Press 2001): 13-106 at 61. 79 In Part 6 of this chapter. 80 Albania art. 20; Belarus art. 15 and 50, Croatia art. 15, Czech Charter art. 25, Estonia art. 52 and 37, Georgia art. 38, Hungary art. 68, Latvia art. 114, Lithuania art. 37, Macedonia art. 48, Moldova art. 35, Poland art. 35, Romania art. 6 and 32, Russia art. 26, Slovakia art. 34, Slovenia art. 11 and 61, the 63
ENDNOTES
341
Ukraine art. 10 and 53, Serbia and Montenegro Charter art. 52, Montenegro art. 34 and 68-73, and Serbia art. 49 and 32. 81 Art. 6. 82 Art. 68 (2). See also the constitutions of Albania art. 20, Belarus art. 50, Czech Charter art. 25, Hungary art. 68, Macedonia art. 48 (4), Slovakia art. 34, Romania art. 32 (3), the Ukraine art. 53, Serbia and Montenegro Charter art. 52, Montenegro art. 68 and Serbia art. 32. 83 Art. 48 (2). 84 Art. 51 (2). 85 Art. 68 (2) and (3). 86 Art. 73. 87 Art. 7. 88 Art. 48. 89 Heading at art. 64. 90 Art. 64. 91 See András László Pap, “Representation or Ethnic Balance: Ethnic Minorities in Parliaments”, Journal of East European Law 7 (2000): 261-339 at 289. 92 This particular quote is taken from article 6 of the Romanian constitution. The following constitutions have similar provisions: Albania art. 20; Croatia art. 15; Czech Charter art. 25; Georgia art. 38; Latvia art. 114; Lithuania art. 37; Macedonia art. 48; Poland art. 35 (1), although section 2 of the same article uses the language of group rights; Romania art. 6; Slovakia art. 34; Slovenia art. 61, although note the exception relating to Hungarian and Italian minorities); Ukraine art. 53 and Serbia art 32. 93 Art. 68 (4). But note that the statute on the rights of national and ethnic minorities adopted on 7 July 1993 uses both the language of collective and individual rights, see Paczolay, supra note 60 at 123. 94 Thus, art. 64 states that: “The autochthonous Italian and Hungarian ethnic communities and their members shall be granted the right to…”. 95 Art. 35 (2). 96 Art. 35 (1). For other examples of the mixed use of both group- and individual rights language, see Estonia art. 49-51. 97 As an account of the actual, authoritative legal situation of the United States this is certainly an oversimplification: the rejection of group rights is not absolute in the United States law. For example, when the U.S. Supreme Court allowed Amish families to keep their children out of school up to a certain age (see Wisconsin v. Yoder, 406 U.S. 205 (1994)), or when it upheld Native American tribal law that imposed patrilineal kinship rules that limited women's marital choices (see Santa Clara Pueblo v. Martinez, 436 U.S. 49 (1978)), it clearly recognised the legal weight of group-based claims for treatment different to that accorded by universally binding legal rules. Similar group-based thinking is visible in the enhanced legal protection of those who are victims of crimes motivated by hatred of a group (in the form of enhanced punishment for hate crimes, see Wisconsin v. Mitchell, 508 U.S. 476 (1993)). On the qualified nature of the group/individual rights distinction in U.S. law, see Jack Greenberg, “Affirmative Action in Higher Education: Confronting the Condition and Theory”, Boston College Law Review 43 (2002): 521-621 at 580-81. 98 See Dimitrina Petrova, “Racial Discrimination and the Rights of Minority Cultures”, in Sandra Fredman, ed., Discrimination and Human Rights: The Case of Racism (Oxford University Press: Oxford 2001) at 65; Miriam J. Aukerman, “Definitions and Justifications: Minority and Indigenous Rights in a Central/East European Context”, Human Rights Quarterly 22 (2000): 1011-1050 at 102930. 99 The terminology is of J.A. Laponce, discussed by Aukerman, id. at 1029. 100 See Sajo, supra note 65 at 70-71. 101 Michel Rosenfeld, “Can Human Rights Bridge the Gap between Universalism and Cultural Relativism? A Pluralist Assessment Based on the Rights of Minorities”, Columbia Human Rights Law Review 30 (1999): 249-84 at 254. 102 Brian Barry, Culture and Equality (Polity: Cambridge, 2001) at 134. 103 “All post-communist states of the region claim adherence to liberal constitutionalism, and no national minority . . . would question main liberal tenets”, Nenad Dimitrijeviü, “Ethno-Nationalized States of Eastern Europe: Is There a Constitutional Alternative?”, Studies in East European Thought 54 (2002): 246-69 at 247, emphasis added.
342 104
ENDNOTES
Tibor Várady, “On the Chances of Ethnocultural Justice in East Central Europe”, in Kymlicka & Opalski, supra note 78: 135-49 at 147-48. 105 Quoted in Stephen Deets, “Reconsidering East European Minority Policy: Liberal Theory and European Norms”, East European Politics & Societies 16 (2002): 30-53 at 49. 106 Alexander Ossipov, “Some Doubts about ‘Ethnocultural Justice’”, in Kymlicka & Opalski, supra note 78: 171-85 at 175. 107 Petrova, supra note 98 at 66. See similarly Rosenfeld, supra note 101 at 257. 108 Aukerman, supra note 98 at 1032. 109 Boris Tsilevich, “New Democracies in the Old World”, in Kymlicka & Opalski, supra note 78: 154-70 at 159. 110 Kymlicka, supra note 78 at 89 n. 44. 111 The Constitutions containing no provision relating to an official language are those of Bosnia and Herzegovina, the Czech Republic, Hungary, and Russia. 112 The exception, containing no such right, is the Constitution of Bosnia and Herzegovina. 113 Belarus art. 50, Bulgaria art. 36 (2), Croatia art. 15 (2), Czech Charter art. 25 (1), Georgia art. 38 (1), Hungary art. 68 (2), Russia art. 26, Slovakia art. 34 (1), Slovenia art. 61, Ukraine art. 10, Serbia art. 49, Montenegro art. 68. 114 Of this group, only the Constitutions of Croatia, Hungary, Ukraine and Montenegro do state in express terms that the right is for minority groups. 115 Albania art. 20 (1), Latvia art. 114, Lithuania art. 37, Macedonia art. 48 (2), Moldova art. 10 (2), Poland art. 35 (1), Romania art. 6 (1). 116 Macedonia art. 7 (2) and (3), Serbia art. 8, Montenegro art. 9, and Serbia and Montenegro Charter art. 52. 117 Czech Charter art. 25 (2), Estonia art. 51 and 52, Slovakia art. 34 (2), Slovenia art. 62, and Montenegro art. 72. The Constitution of Montenegro uses both these techniques for official uses of minority languages. The Estonian Constitution states that the right to an official use (not the exclusive official use!) of a minority language exists when the majority of the residents of any given locality belong to the minority in question. 118 Albania art. 20 (2), Belarus art. 50, Bulgaria art. 36 (2), Czech Republic art. 25 (2), Estonia art. 37 (4), Hungary art. 68 (2), Macedonia art. 48 (4), Moldova art. 35 (2), Romania art. 32 (3), Russia art. 26, Slovakia art. 34 (2), Ukraine art. 53, Montenegro art. 68, Serbia art. 32, and Serbia and Montenegro Charter art. 52. The Estonian Constitution only allows for this right to be exercised in schools specially established for minorities. 119 Belarus, Czech Republic, Macedonia, Romania, Slovakia, Slovenia, Ukraine, Montenegro, Serbia. 120 Macedonia art. 48 (4) and Moldova art. 35 (3). 121 Art. 36 (1). 122 Albania art. 28 (1), Bosnia and Herzegovina (ECHR) art. 6 (3) (a), Croatia art. 24 (2), Estonia art. 21 (1), Poland art. 41 (3), Romania art. 23 (5), Montenegro art. 22, and Serbia and Montenegro Charter art. 16. 123 Albania art. 31, Bosnia and Herzegovina (ECHR) art. 6 (3) (e), Czech Charter art. 37 (4), Moldova art. 118, Romania art. 127 (2), Serbia art. 123 and Serbia and Montenegro Charter art.16. 124 Decision 8/96 of 26 August 1997, summarized in Bull. Constit. Case Law (1997, no. 2): 252-53, SVK1997-2-007. 125 Decision U.br.49/98 of 20 May 1998, described in Bull. Constit. Case Law (1998 no. 2) at 326, MKD1998-2-004. 126 Decision U.br.36/98 of 25 November 1998. 127 Decision U.br. 32/99, of 9 June 1999, described in Bull. Const. Case-Law 1999 (2) at 286-87, MKD1999-2-007. 128 Art. 27. 129 Janusz TrzciĔski, Remarks about Article 27, in L. Garlicki, ed., Konstytucja Rzeczypospolitej Polskiej: Komentarz (Wydawnictwo Sejmowe: Warszawa 1999) (loose leaf). 130 Id. at 4 (quoting, approvingly, J. Boü). 131 Id. at 4. 132 Decision W. 7/96 of 14 May 1997, Orzecznictwo Trybunaáu Konstytucyjnego: Rok 1997 (C.H. Beck: Warszawa 1998) at 770-96.
ENDNOTES 133
343
TrzciĔski, supra note 129 at 3; Jerzy Oniszczuk, Konstytucja Rzeczypospolitej Polskiej w orzecznictwie Trybunaáu Konstytucyjnego (Zakamycze: Kraków, 2000) at 230. 134 Decision W. 7/96, at 773. 135 Id. at 796. 136 Professor TrzciĔski’s authoritative status on this particular issue is beyond any doubt because, before having written the constitutional commentary from which this excerpt is quoted, he had authored the Constitutional Tribunal's decision of 14 May 1997. 137 TrzciĔski, supra note 129 at 4-5. 138 Petai, supra note 72 at 22. 139 Vello Pettai, “Definitions and Discourse: Applying Kymlicka's Models to Estonia and Latvia”, in Kymlicka & Opalski, supra note 78 at 267, footnote omitted. 140 See Kymlicka, supra note 78 at 76-9. 141 Pettai, supra note 72 at 26. 142 President Meri initially vetoed the law, and after the parliament (the Riigikogu) adopted the law without any of the amendments postulated by the President, the President exercised his constitutional right to challenge the law before the Constitutional Review Chamber. I am indebted for this description of the decision and its background to Pettai, supra note 72 at 26-29. 143 Decision 3-4-1-1-98 of 5 February 1998, summarised in Bull. Const. Case Law 1998 (1): 37-8, EST1998-1-001. 144 Pettai, supra note 72 at 28. 145 In November 1998 the Constitutional Review Chamber considered a challenge, which reached it via a lower court, to the original Language Act (not the 1997 amendments) requirements for local deputies; see id. at 28-29. 146 Id. at 29. 147 Pap, supra note 91 at 262 and 267. 148 Id. at 263. 149 For a more detailed description, see id. 284-85. 150 Id. at 285-86. 151 Art. 59 (2). 152 See Pap, supra note 91 at 286-88. 153 See Stephen Deets, “Reconsidering East European Minority Policy: Liberal Theory and European Norms”, East European Politics & Societies 16 (2002): 30-53 at 45-48. 154 Art. 80 (3). 155 For a detailed description of the complicated system, see Pap, supra note 91 at 288-89. 156 Art. 68 (3). 157 Art. 68 (4). 158 Paczolay, supra note 60 at 125. 159 Deets, supra note 153 at 49-51. 160 See Pap, supra note 91 at 320-24. 161 See 2002 Regular Report on Hungary’s Progress Towards Accession, Commission of the European Communities, Brussels 9 October 2002, SEC(2002) 1404, available at http://europa.eu.in/ comm/enlargement/report2002/hu_en.pdf, at 20. 162 The Law on Parties of July 1991, see Pap, supra note 91 at 279-80. 163 Art. 11 (4). 164 Pap, supra note 91 at 280, footnote omitted. 165 See id. at 282-83. 166 Decision 19/98 of 15 October 1998, summarised in Bull. Constitution. Case Law 3 (1998) 460-62, SVK-1998-3-010. 167 Decision U-I-283/94 of 12 February 1998. 168 For descriptions and analysis of the decision, see Emil Konstantinov, “Turkish Party in Bulgaria Allowed to Continue”, East Europ. Constit. Rev. 1:2 (Summer 1992): 11-12; Jean-Piere Massias, Droit constitutionnel des États d'Europe de l'Est (Presses Universitaires de France: Paris, 1999) at 161-62; Anna M. Ludwikowska, Sądownictwo konstytucyjne w Europie ĝrodkowo-Wschodniej w okresie przeksztaáceĔ demokratycznych (TNOiK: ToruĔ 1997) at 137-39; Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago, 2000) at 172-73.
344
ENDNOTES
169
There was a second constitutional ground cited by the petition: the ban on organisations that call into question the sovereignty and territorial integrity of the country, or that foment ethnic or religious enmity and the violation of the rights and freedoms of citizens (Art. 44 (2)). The petitioners claimed that, by favouring a policy of ethnic assimilation of Bulgarian Moslems into the Turkish minority, the MRF promotes ethnic and religious confrontation. Both for our purposes, and in the argument of the Court, the claim based on Art. 11 (4) was dominant. 170 See Kymlicka, supra note 78 at 55. 171 Antonina Zhelyazkova, “The Bulgarian Ethnic Model”, East Europ. Constit. Rev. 10:4 (Fall 2001): 6266 at 65. 172 See Konstantinov, supra note 168 at 11. 173 Quoted in Venelin Ganev, “Foxes, Hedgehogs and Learning: Notes on the Past and Future Dilemmas of Postcommunist Constitutionalism”, paper presented at the workshop “Rethinking the Rule of Law in Post-Communist Europe”, European University Institute, Florence 22-23 February 2002, at 12. 174 There are 12 judges on the Constitutional Court but only eleven took part in the consideration of the MRF case (one judge was ill). The rule that the constitutional requirement of “more than half of the votes of all Justices” (art. 151 (1)) means a requirement of at least seven (regardless of the number of Justices participating in the vote) for the decision of unconstitutionality does not have a clear textual mooring but, as Ganev explained, evolved as an established practice that can now can be viewed as a constitutional convention, see Venelin Ganev, “The Rise of Constitutional Adjudication in Bulgaria”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 247-64 at 253. 175 Id. 176 Zhelyazkova, supra note 171 at 65. 177 See Cindy Skach, “Rethinking Judicial Review: Shaping the Toleration of Difference?”, paper presented at the workshop “Rethinking the Rule of Law in Post-Communist Europe”, European University Institute, Florence 22-23 February 2002.
Chapter 9 1
On Spanish model of transition to democracy, see Michel Rosenfeld, “Constitution-Making, Identity Building, and Peaceful Transition to Democracy: Theoretical Reflections Inspired by the Spanish Example”, Cardozo Law Review 19 (1998): 1891-1920. 2 See Jon Elster, “Coming to Terms with the Past. A Framework for the Study of Justice in the Transition to Democracy”, Arch. europ. sociol. 39 (1998): 7-48 at 14, where he defines retroactive justice as “political decisions made in the immediate aftermath of the transition and directed towards individuals on the basis of what they did or what was done to them under the earlier regime” (footnote omitted); he also makes it clear that this concept is coextensive with “transitional justice”, id. at 7. 3 Id. at 46. 4 Adam Michnik, “The Rebirth of Civil Society”, Public lecture at the London School of Economics, 20 October 1999, http://www.lse.ac.uk/Depts/global/Michnik89.htm. 5 Vojtech Cepl, “Ritual Sacrifices”, East Europ. Constit. Rev. 1:1 (Spring 1992): 24-26 at 25. 6 JiĜina Šiklová, “Lustration or the Czech Way of Screening”, in Martin Krygier & Adam Czarnota, The Rule of Law after Communism (Ashgate: Dartmouth 1999): 248-58 at 254-55. 7 Eric A. Posner & Adrian Vermeule, “Transitional Justice as Ordinary Justice”, Harvard Law Review 117 (2004): 761-825 at 806. 8 Aviezer Tucker, “Paranoids May Be Persecuted: Post-Totalitarian Retroactive Justice”, Arch. europ. sociol., 40 (1999): 56-100 at 64. 9 Ruti G. Teitel, Transitional Justice (Oxford University Press: Oxford 2000) at 98. 10 Šiklová, supra note 6 at 255. 11 Jon Elster, “Retribution in the Transition to Democracy”, in Arend Soeteman (ed.), Pluralism and Law (Kluwer Academic Publishers: Dordrecht, 2001): 19-36 at 23. 12 An analogy may be drawn to a central argument by Lee Bollinger, who warns against restraints upon racist speech in the US on the basis that such restraints may originate from wrongful motives, such as sheer psychological intolerance, or the sense of guilt for not having helped European Jews during the Holocaust, see Lee Bollinger, The Tolerant Society (Oxford University Press: Oxford 1986) at 12930, 274-75 n. 17.
ENDNOTES 13
345
Cepl, supra note 5 at 25. I cannot resist the temptation of quoting this remarkable passage, which illustrates nicely the uncertain boundaries between irrational vengeance and a more respectable sense of retribution as well as consequentialist argument: “Most supporters of lustration in post-communist universe are not very good at articulating their anxieties or grounding them in reality. This created the impression that they were looking for vengeance in western non-paranoid eyes. But their paranoia aside, persecuted they were, and though they could not articulate their fears of the nomenklatura, they had good reason to fear a class of people that survive by stealing anything that can be moved and corrupting any being with a soul”, Tucker supra note 8 at 97-98. 15 From a statement of the Chairman of the Senate of Czech Republic on the amendment to the lustration law, quoted in the decision of the Czech Constitutional Court Pl. US 9/01 of 5 December 2001, see http://www.concourt.cz/angl–verze/doc/p-9-01.html at 5. 16 From the Hungarian Constitutional Court Decision 60/1994 of 24 December 1994 (discussed below) in László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 306-15 at 312. 17 This is a summary of the petitioners’ objections provided by the Czech Constitutional Court in its Decision 19/93 of 21 December 1993, reprinted in East European Case Reporter 4 (1997): 149-174. 18 As summarized by the Constitutional Court, id. at 152. 19 Id. at 152. 20 See text accompanying footnote 17 above. 21 Id. at 155. Article 15 (2) of the Czech Charter (“freedom of scientific research”) was one of the grounds for a challenge to the Law. 22 Id. at 155. 23 JiĜí PĜibáĖ, “Moral and Political Legislation in Constitutional Justice: A Case Study of the Czech Constitutional Court”, J.E. Eur. L. 8 (2001): 15-34 at 20-21. 24 Perhaps better translated as the Institute of National Remembrance. The Polish word “pamiĊü”, which features in the name of the Institute, has both of these meanings. 25 Teitel supra note 9 at 81. 26 Id. at 83. 27 Bruce Ackerman, The Future of Liberal Revolution (Yale University Press: New Haven, 1992) at 71. 28 Natalia Letki, “Lustration and Decommunisation in East-Central Europe”, Europe-Asia Studies 54 (2002): 529-52 at 539, footnote omitted. 29 In addition, it has been claimed (not inaccurately) that Western ideas about decommunisation have been partly shaped by the selectiveness and bias of Westerners’ contacts with the new élites in postcommunist states: “the Western visitors come for a short time, stay only in Prague or Budapest and talk with a few English speaking intellectuals with dissident background. The kind of locals that visitors are likely to interview are likely to be eager to contribute to the integration of their country in Europe”, Tucker supra note 8 at 75. There is therefore an element of self-perpetuation of the views hostile to decommunisation: opinions in CEE are partly shaped by Western attitudes, which, in turn, are partly shaped by selectively sampled informers in CEE. 30 Id. at 73. 31 Resolution of the Parliamentary Assembly of Council of Europe no. 1096 of 27 June 1996 on measures to dismantle the heritage of former communist totalitarian systems, http://stars.coe.fr/ Main.asp?link=http%3A%2F%2Fstars.coe.fr%2FDocuments%2FAdoptedText%2Fta96%2FERES10 96.htm. For a working document, referred to in the Resolution, entitled “Measures to dismantle the heritage of former communist totalitarian systems”, Doc. 7568 of 3 June 1996 see http://stars.coe.fr/ Main.asp?link=http%3A%2F%2Fstars.coe.fr%2FDocuments%2FWorkingDocs%2FDoc96%2FEDO C7568.htm. 32 Paragraph 11 of the Resolution 1096, supra note 31. 33 Id., paragraph 12. 34 Id. paragraph 12, see also Doc. 7568, supra note 31, paragraph 16. 35 Teitel, supra note 9 at 164. 36 Elster, supra note 2 at 14. 37 The German legal solution was occasionally referred to, with approval, by CEE Constitutional Courts when scrutinising lustration and decommunisation measures in their own countries; see, for example, Constitutional Court of Lithuania, Decision of 4 March 1999, http://www.lrkt.lt/1999/n9a0304a.htm, 14
346
ENDNOTES
Part 2 of Court’s decision; Constitutional Court of Czech Republic, Decision 19/93 of 21 December 1993, in East European Case Reporter 4 (1997): 149-174 at 174 n. 1. 38 Glaring exceptions to this general rule are provided by many former member states of the USSR (though not the Baltic ones), including Russia itself, in which neither “lustration” or “decommunisation” ever occurred – less, however, by virtue of a principled, consensual decision to adopt this form of reconciliation, and more “by default”, see Elster supra note 2 at 18. 39 Noel Calhoun, “The Ideological Dilemma of Lustration in Poland”, East European Politics and Societies 16 (2002): 494-520 at 495. 40 Aleks Szczerbiak, “Dealing with the Communist Past or the Politics of the Present? Lustration in PostCommunist Poland”, Europe-Asia Studies 54 (2002): 553-72 at 570, emphasis added. 41 Cepl supra note 5 at 24. 42 Act No. 451/1991 Sb. of 4th October 1991 43 Helga A. Welsh, “Dealing with the Communist Past: Central and East European Experiences after 1990”, Europe-Asia Studies 48 (1996): 413-428 at 415. 44 “Constitution Watch: Czech Republic”, East Europ. Constit. Rev. 4:4 (Fall 1995): 8-10 at 9. 45 Act no. 422/2000. 46 Act no. 279/1992 provided for lustration of candidates for work in the police. 47 Act no. 147/2001. 48 Constitutional Court of the Czech and Slovak Federal Republic Decision No. 1/92 of 26 November 1992, see http://www.concourt.cz/angl_ver/decisions/doc/p-1-92.html. 49 Id. at 7. 50 Id. at 8. 51 Id. at 8. 52 See, similarly, Teitel supra note 9 at 165-66. 53 Decision 1/92, supra note 48 at 9. 54 Id. at 10. 55 Id. at 9. 56 Id. at 9. 57 Decision 9/01 of 5 December 2001, see http://www.concourt.cz/angl–verze/doc/p-9-01.html. 58 Id. at 3. 59 Id. at 15-16. 60 Id. at 18. 61 Id. at 18. 62 Id. at 19. 63 “Constitution Watch: Albania”, East Europ. Constit. Rev. 5:1 (Winter 1996): 2-3. 64 For a detailed discussion of this decision see Kathleen Imholz, “A Landmark Constitutional Court decision in Albania”, East Europ. Constit. Rev. 2:3 (Summer 1993): 23-5. 65 The statute “On the Assessment of the USSR Committee of State Security (NKVD, NKGB, MGB, KGB) and Present Activities of the Regular Employees of This Organisation” of 16 July 1998. 66 Decision of 4 March 1999, http://www.lrkt.lt/1999/n9a0304a.htm. I discuss this decision also in Chapter 10, pp. 279. 67 Id., Section 5 of the Court’s decision. 68 Id., Section 5.2 of the Court’s decision. 69 In the first elections after the fall of the Zhivkov rule, held in June 1990, the post-Communist Bulgarian Socialist Party (BSP) won a solid majority in the Grand National Assembly; after the collapse of the BSP government in September 1990, a coalition led by both the BSP and the liberal-democratic Union of Democratic Forces (UDF) was formed. 70 Decision 60/1994 (X11.24) of 22 December 1994, reprinted in East European Case Reporter of Constitutional Law 2 (1995): 159-193; excerpts in Sólyom & Brunner, supra note 16 at 306-15. The references below are to the excerpts in the latter volume. 71 “[T]he Constitutional Court must consider the transition as a historical fact”, id. at 312. 72 Law LXII/1996. For a political background see Gábor Halmai & Kim Lane Scheppele, “Living Well Is the Best Revenge: The Hungarian Approach to Judging the Past”, in A. James McAdams (ed), Transitional Justice and the Rule of Law in New Democracies (Notre Dame: University of Notre Dame Press, 1997): 155-84 at 176-77. 73 “Constitution Watch: Hungary”, East Europ. Constit. Rev. 11:3 (Summer 2002): 22-26 at 25.
ENDNOTES 74
347
Sejm [The lower chamber of Polish parliament] Resolution of 28 May 1992, described in Andrzej S. Walicki “Transitional Justice and the Political Struggles of Post-communist Poland” in McAdams, supra note 72: 185-238 at 197-98; see also Wiktor Osiatynski, “Agent Walesa?”, East Europ. Constit. Rev. 1:2 (Summer 1992): 28-30; Calhoun, supra note 39 at 503-505. 75 Decision U. 6/92 of 19 June 1992; Polish text on file with the author, p. 6. 76 Decision W. 5/94 of 14 July 1993, http://www.trybunal.gov.pl/OTK/teksty/otkpdf/1993/W_05_93.pdf. 77 Judgment K. 3/98 of 24 June 1998, translation in East European Case Reporter of Constitutional Law 6 (1999) 130-211. The ex-ante review was initiated by the President. 78 Id. at 201-2 (Rymarz, J., dissenting), and at 203-11 (Zdyb, J., dissenting). 79 Id. at 190. 80 Decision K. 39/97 of 10 November 1998, O.T.K. (1998), Item No. 26 at 491-557. 81 Id. at 633-41. 82 Decision K. 39/97, OTK at 525. 83 Id. at 518-19. 84 Decision P. 3/00 of 14 June 2000, http://www.trybunal.gov.pl/OTK/teksty/otkpdf/2000/p_03_00.pdf. 85 Id. at 11. 86 For example, a draft law proposed by a centrist party Freedom Union in April 2001, see Jolanta Kroner, “Precyzyjniej, z domniemaniem niewinnoĞci, bez donosu poselskiego”, Rzeczpospolita (Warsaw) 9 April 2001 at C1. 87 Decision K 11/02 of 19 June 2002, http://www.trybunal.gov.pl/OTK/teksty/otkpdf/2002/K_11_02.pdf. 88 Id. at 3. 89 See id. at 30-34 (Mazurkiewicz, J., dissenting). 90 Id. at 34-40 (Safjan, C.J., dissenting); similarly id. at 40-42 (StĊpieĔ, J., dissenting). 91 Decision 8/92 of 27 July 1992, discussed in Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago, 2000) at 188-89. 92 Decision 11/92 of 29 July 1992. 93 Interview with Professor Neno Nenovsky, former Justice of the Constitutional Court of Bulgaria (in 1991-94), Sofia 10 May 2001. He was one of the dissenting judges in this case. 94 Which is not to say that it was found constitutional. The Bulgarian Constitutional Court takes decisions by majority; in this particular case, six judges voted to uphold the law, five to strike it down, and one abstained. There was therefore not a sufficient majority (of seven) to strike down the law but neither was the law explicitly confirmed as constitutional. 95 Decision 1/93 of 11 February 1993, see “Constitution Watch: Bulgaria”, East Europ. Constit. Rev. 2:1 (Winter 1993): 2-3 at 3. 96 Rumyana Kolarova, “Bulgaria: A Self-Restricting Court”, East Europ. Constit. Rev. 2:2 (Spring 1993): 48-50 at 50. 97 “National Assembly Repasses Act Reversing ‘Panev Law’”, BBC Summary of World Broadcasts, 25 March 1995 (available in LEXIS-NEXIS). 98 “Constitution Watch: Bulgaria”, East Europ. Constit. Rev. 6:4 (Fall 1997): 6-10 at 8. 99 Decision no 14/1997 of 22 September 1997. 100 Annual Report 1999, International Helsinki Federation for Human Rights, at http://www.ihfhr.org/reports/ar99/ar99bul.htm. 101 Decision no. 02/99 of 21 January 1999, summarised in Bull. Const. Case-Law 1999 (1): 28-29, BUL1999-1-002. 102 “Constitution Watch: Bulgaria”, East Europ. Constit. Rev. 8:1/2 (Winter/Spring 1999): 5-7 at 7. 103 Decision 19/93 of 21 December 1993, reprinted as “Judgment of the Constitutional Court of the Czech Republic”, Parker Sch. J.E. Eur. L. 1 (1994): 363-91; see also comment by Herman Schwartz, “The Czech Constitutional Court Decision on the Illegitimacy of the Communist Regime”, Parker Sch. J.E. Eur. L. 1 (1994): 392-8. See also the same translation in East European Case Reporter 4 (1997): 149174. The references below are to the latter source. 104 Id. at 157. 105 Id. at 158. 106 Id. at 157. 107 Id. at 160. 108 For a similar criticism of this aspect of the Court’s decision by a scholar otherwise supportive of the Court’s jurisprudence, see PĜibáĖ, supra note 23 at 28.
348
ENDNOTES
109
Decision 19/93, supra note 103 at 163. Id. at 163. 111 Id. at 166-69. 112 Law of 4 November 1991 on the “Right to Prosecute Serious Criminal Offences Committed Between December 21 1944 and May2 1990 That Had Not Been Prosecuted For Political Reasons” published in Journal of Constitutional Law in Eastern and Central Europe 1(1994): 129-157. 113 Constitutional Court decision no. 11/1992 of 5 March 1992, reproduced in Sólyom & Brunner, supra note 16 at 214-18. 114 Id. at 219. 115 Id. at 221. 116 Halmai and Scheppele, supra note 72 at 164-165. 117 Decisions 41/1993 (VI.30) Abh and 42/1993 (VI.30)Abh. 118 Decision 53/1993 (X.13) Abh. 119 Stephen J. Schulhofer, comments in the symposium “Dilemmas of Justice”, East Europ. Constit. Rev. 1:2 (Summer 1992): 17-22 at 18. 120 Id. at 18. 121 Decision 11/1992 of 5 March 1992 reproduced in Sólyom & Brunner, supra note 16 at 223. 122 See id, especially at 221-22. 123 Posner & Vermeule, supra note 7 at 797. 124 See, e.g., Spencer Zifcak, “Hungary’s Remarkable, Radical, Constitutional Court”, Journal of Const. Law in Eastern and Central Europe 3 (1996): 1-56 at 6-9. 125 Ruti Teitel, “Paradoxes in the Revolution of the Rule of Law”, Yale Journal of International Law 19 (1994): 239-47 at 244-45. 126 Id. at 246. 127 Michel Rosenfeld, comments in the symposium “Dilemmas of Justice”, supra note 119 at 20. 128 It is estimated that only a few hundred Czechs lost their jobs due to lustration, and that around five percent of the members of parliament were forced to resign, see Tucker supra note 8 at 84. 129 As Adam Czarnota and Piotr Hofmanski observe, “The attempt to make a legal break with the past … taken together with the simultaneous recognition of the continuity of the state, and with that [sic] the continuity of the system of law, amounts to an attempt to square a circle”, Adam Czarnota & Piotr Hofmanski, “Can we do Justice to the Past?”, in Krygier & Czarnota, supra note 6: 197-211 at 199. 130 Carlos Santiago Nino, Radical Evil on Trial (Yale University Press: New Haven, 1996) at 120. 131 Posner & Vermeule, supra note 7 at 795. 132 For an account, see Carlos H. Acuña & Catalina Smulovitz, “Guarding the Guardians in Argentina: Some Lessons about the Risks and Benefits of Empowering the Courts”, in McAdams, supra note 7: 93-122 at 102. 133 Teitel, supra note 9 at 14. 134 Constitutional Court decision of 5 March 1992, no. 11/1992, see supra notes 113-115 and the accompanying text. 135 See, e.g., the decision of Constitutional Tribunal K.7/90 of 22 August 1990 (upholding the constitutionality of the law that deprived ex-high officials of the Communist Party of their high pensions, based on the argument that their early retirement law was unjust in the first place); see similarly decision U.6/93 of 12 April 1994 (a decision by the Minister of Industry discontinuing the right to free electricity of employees of the energy industry, held constitutional because the right to free energy violated equality before the law). 110
Chapter 10 1 2
Bosnia and Herzegovina, Georgia, Lithuania and Montenegro. Compare, e.g., Basic Law of F.R.G. article 5 (freedom of expression can be subject to limitations by statutes ‘for the protection of youth” and for the protection of “personal honour”) with article 10 (privacy of letters, posts and telecommunications may be restricted by statutes in order “to protect the free democratic order or the existence or security of the Federation) and with article 11 (freedom of movement may be restricted by statutes when necessary to protect the free democratic order, to
ENDNOTES
349
combat the danger of epidemics, to deal with natural disasters or grave accidents, to protect young people or to prevent crime). 3 Art. 24 (1). 4 Art. 24 (4). 5 As examples of such rights provisions in the first category of constitutions, consider the right to strike in the Constitution of Lithuania (Art. 51: “(1) Employees shall have the right to strike in order to protect their economic and social interests. (2) The restrictions of this right, and the conditions and procedures for the implementation thereof shall be established by law”); consider also the right to vote in Lithuania (art. 20) and Georgia (art. 28). 6 Serbia and Montenegro Charter (art. 5), Macedonia (art. 54), Slovenia (art. 15), the Ukraine (art. 64) and Serbia (art. 11) [note that, hereinafter, a name of a country followed by a number of an article refers to the provision of the constitution of that country]. The same construction was envisaged by the abortive project of the charter of rights in Poland in 1992 which explicitly stated: “The rights and liberties guaranteed by this Charter may be restricted only by statute and only when such a restriction is envisaged by this Charter”, id. art. 5.2 7 Art. 55 (3). 8 Art. 56 (3). 9 These include the right to life (art. 20), to dignity and against torture (art. 21), to privacy (art. 23 (1)), etc. 10 Decision of 27 March 1996, summarised at http://ks.rfnet.ru/english/codicese.htm, and in the Bull. Constit. Case Law 1996 (2): 253-55. 11 Belarus, art. 23 (1). Other general clauses of this type are: Albania (art. 17); Belarus (art. 23); Croatia (art. 16), Czech Charter (Art. 4 (2, 3, 4)); Estonia (art. 11), Hungary (art. 8(2)); Latvia (art. 116), Moldova (art. 54), Poland (art. 31), Romania (art. 49); Slovakia (art. 13). 12 For instance, the Estonian constitution mentions only “necessity in a democratic society” and the requirement that the limits must not “distort the nature of rights and liberties”, Art. 11. 13 For instance, the Albanian constitution provides, inter alia, that the rights limitations may not exceed the limitations provided for in the ECHR, Art. 17 (2). 14 Moldova Art. 54 (2); Romania art. 49 (2). 15 E.g., Polish Constitutional Tribunal, Decision K.23/98 of 25 February 1999. 16 Consider these two examples. Article 29 of the Moldovan constitution allows for the following grounds for restricting the right to the inviolability of the domicile: execution of an arrest warrant; preventing the threat to life, physical integrity or the property of a person, and preventing the spread of a disease. Art. 26 of the Albanian Constitution envisages the following grounds for limiting the right against forced labour: execution of a judicial decision, the performance of military service, or of service resulting from a state of war, a state of emergency, or a natural disaster. This is the way in which the Constitution fleshes out the vague general clause (art. 17) to specific fact situations in which the limits on a particular right may be relevant. 17 See Poland Art. 31 (3), Estonia Art. 11, Russia Art. 55 (3), Slovenia Art. 15 (2). 18 See Albania Art. 17(1); Moldova Art. 54 (2); Romania Art. 49 (2). 19 See Romania, Art. 49. 20 The Slovenian Constitution has a “necessity” but not a “proportionality” requirement, Art. 15 (2). 21 Decision U-I-25/95, of 27 November 1997, at http://www.us-rs.si/en/casefr.html, para. 47. 22 More precisely, the requirement of necessity is present in: Articles 8-11 of the Convention rights to respect for privacy, to freedom of thought, conscience and religion, to freedom of expression, and to freedom of association and assembly, respectively; Article 2 of Protocol No. 4 (liberty of movement within a state); and Article 1 of the Protocol No. 7 (right of an alien not to be expelled before certain conditions are met). 23 See e.g. Goodwin v. United Kingdom 22 E.C.H.R. 123, 143-4 (1996); for discussion see Alastair Mowbray, Cases and Materials on the European Convention of Human Rights (Butterworths, London 2001) at 411-2, 448. 24 Marc-Andre Eissen, quoted by Mowbray, id. at 413. 25 See P. van Dijk and G.J.H. van Hoof, Theory and Practice of the European Convention on Human Rights, 3rd ed. (Kluwer Law International: The Hague, 1998) at 81. 26 The three tiers are the principles of suitability (relevance), necessity, and proportionality sensu stricto.
350 27
ENDNOTES
Robert Alexy, “Constitutional Rights, Balancing, and Rationality”, Ratio Juris 16 (2003): 131-40 at 135. 28 Id. at 135, emphasis added. 29 Id. at 135, emphasis in original. 30 Gerald Gunther, “The Supreme Court, 1971 Term – Foreword: In Search of Evolving Doctrine on a Changing Court: A Model for a Newer Equal Protection”, Harvard Law Review 86 (1972): 1-49 at 8. These words applied to the so-called “strict scrutiny” of “suspect classifications” under the 14th Amendment to the US Constitution. 31 Art. 31 (3) of Constitution. Before the Constitution of 1997, under the so called “Little Constitution”, which provided no textual basis for “necessity” or “proportionality” tests, the Constitutional Tribunal read the requirement of proportionality into the general principle of the rule of law, see Jerzy Oniszczuk, Konstytucja Rzeczypospolitej Polskiej w orzecznictwie Trybunaáu Konstytucyjnego (Zakamycze: Kraków, 2000), p. 252. 32 Decision no. P.2/98 of 12 January 1999. 33 The word used in the original is “samowola budowlana”, which means building without permission, but in addition has a strong negative connotation in Polish. 34 Decision no. P.2/98 of 12 January 1999, Orzecznictwo Trybunaáu Konstytucyjnego w 1999 r. - czĊĞü I (Wydawnictwo Trybunaáu Konstytucyjnego: Warszawa 2000): 13-29 at 15. 35 This article lists the grounds upon which the constitutional rights can be legitimately restricted by statutes. 36 Decision no. P.2/98, at 26. 37 Id. at 20. 38 Id. at 23. 39 See Dieter Grimm, “Human Rights and Judicial Review in Germany”, in David M. Beatty, ed., Human Rights and Judicial Review: A Comparative Perspective (Martinus Nijhoff: Dordrecht, 1994): 267-96. 40 Decision of 28 April 2000, no. 3-4-1-6-2000, at http://www.nc.ee/english/const/2000/3-4-1-6-00i.html. 41 Id., section 13 of the Decision. 42 Id, section 16. 43 Id. section 17. 44 Decision no. 26/1999 of 8 September 1999, summarised in Bull. Constit. Case-Law 1999 (3): 392, HUN-1999-3-008. 45 Decision no. 793/B/1997 of 24 February 1998, summarised in Bull. Constit. Case-Law 1998 (1): 57-58, HUN-1998-1-002. 46 Decision No 139 of 14 December 1994, http://www.cecl.gr/RigasNetwork/databank/Jurisprudence/ jurisprudence_main.htm. 47 Decision No 71 of 1993. 48 Decision No 139 of 1994. 49 Interview with Mr Horatiu Dumitru in Bucharest, 10 March 2001. Mr Dumitru, a private lawyer at the time of the interview, had worked as lawyer for the Government at the time of the Constitutional Court's decision discussed in the main text. 50 This is reinforced by a contrast with section (2), which is worded in characteristically right-oriented terms: “Citizens have the right to pensions, paid maternity leave…”, etc. 51 Namely, the defence of national security, public order, health or morals, the prevention of a natural calamity or extremely grave disaster. 52 Decision of 17 May 1996, no. 21/1996, reprinted in László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 333-45. 53 Puzzlingly, the Court justifies its decision by recalling the principle that “[r]estricting the right of association in the interest of protecting third persons from the infringement of their rights and liberties is constitutional if the restriction is made necessary by the other right and the extent of restriction is proportionate to the desired aim”, id. at 337, emphasis added. 54 Id. at 339 55 In Joel Feinberg’s language, it is a “single-party paternalism”, see The Moral Limits of the Criminal Law, vol. 3: Harm to Self (Oxford University Press: New York, 1986) at 9. 56 T. Alexander Aleinikoff, “Constitutional Law in the Age of Balancing”, Yale Law Journal 16 (1987): 943-1005.
ENDNOTES 57
351
The Charter of Fundamental Rights of the European Union, “solemnly proclaimed” by the European Council in Nice in December 2000, spells out expressly the principle of proportionality and the requirement of necessity (Art. 52.2), in contrast to the European Convention on Human Rights, which “acquired” the principle of proportionality through the doctrine of the Strasbourg Court. 58 Aleinikoff, supra note 56 at 962. 59 Poland Art. 31 (3), Albania art. 17 (2). 60 Art 4 (4) of Czech Charter, Slovakia art 13 (4); Hungary art 8 (2). 61 Moldova Art. 54 (2); Romania Art. 49 (2). 62 Estonia Art. 11. 63 E.g., Albania, which has an “essence” proviso, art. 17 (1), but a relatively weak constitutional court. 64 E.g., one of the most important decisions of the Hungarian Constitutional Court, in which death penalty was declared unconstitutional, was based, inter alia, on the argument that such a penalty necessarily intrudes upon the essential content of the right to life; see Decision 23/1990 of 31 October 1990, reprinted in Sólyom & Brunner, supra note 52 at 118-38. 65 Decision no 18/98, of 2 June 1999, Bull. Constit. Case-Law 1999 (1): 44-45, CZE-1999-007. 66 Decision P.2/98, of 12 January 1999. 67 Id. at 27. 68 E.g. Albania Art. 17; Czech Charter Art. 4 (1); Moldova art. 54 (1); Poland Art. 31 (3); Romania Art. 49 (1); Slovakia Art. 13 (2). 69 Decision no. U-I-920/1995 and U-I-950/1996 of 15 July 1998, Bull. Constit. Case-Law 1998 (3): 401402, CRO-1988-3-015. 70 Decision no. U-I-262/1998 and U-I-322/1998 of 15 July 1998, published (in Croatian) in Narodne novine, 98/1998, 2434-2437, see Bull. Constit. Case-Law 1998 (3): 401-402. 71 Decision no. U-I-58/1995 and U-I-772/1998, of 5 October 1999, Bull. Constit. Case-Law 1999 (3): 366-367, CRO-1999-3-014. 72 Decision no. U-I-241/1998 of 31 March 1999, Bull. Constit. Case-Law 1999 (1): 33, CRO-1999-1-004. 73 Decision of 26 October 1995, http://www.lrkt.lt/1995/n5a1026a.htm. 74 Decision of 19 December 1996, http://www.lrkt.lt/1996/n6a1219a.htm. 75 Decision of 13 February 1997, http://www.lrkt.lt/1997/n7a0213a.htm. 76 Decision of 4 March 1999, http://www.lrkt.lt/1999/n9a0304a.htm. 77 Section 4 of the Court’s decision. 78 Section 6.1 of the Court’s decision. 79 Decision of 10 March 1998, no. 14/97, http://www.lrkt.lt/1998/n8a0310a.htm. 80 Section 6 (5) of the judgment. 81 Czech Charter art. 4 (3); Macedonia art. 54 (3); Slovakia art. 13 (3). 82 See Chapter 8.1 above. 83 Decision Pl. US 1/92 of 26 November 1992, http://www.concourt.cz/angl_ver/decisions/doc/list.html. 84 Id., at 8. 85 E.g. Czech Charter art. 3 (3); Slovak Const. Art. 12 (4). 86 Decision no. PL.ÚS 18/97 of 28 May 1998, summarised in Bull. Constit. Case-Law 1998(2): 299-300; SVK-1998-2-005. 87 Article 25(2): “No person may be forced to perform military duties if it is contrary to his or her conscience or religious faith or conviction”. 88 Art. 12 (4). 89 Decision no. 12/97 of 25 September 1997, Bull. Constit. Case-Law 1997 (3): 357; BUL-1997-3-003. 90 See Laurence H. Tribe, American Constitutional Law (Foundation Press: Mineola, 1988, 2nd ed.) at 681-82. 91 Kathleen M. Sullivan, “Categorization, Balancing, and Government Interests”, in Stephen E. Gottlieb, ed., Public Values in Constitutional Law (Ann Arbor: University of Michigan Press, 1993): 241-68 at 256. 92 For such an argument see Wiktor Osiatynski, “Rights in New Constitutions of East Central Europe”, Columbia Human Rights Law Review 26 (1994): 111-166 at 152-55. 93 Tadeusz Jasudowicz, “Granice korzystania z praw czáowieka - rozwiązania Konstytucji RP na tle standartów europejskich”, in Cezary Mik, ed., Konstytucja Rzeczypospolitej Polskiej z 1997 roku a czáonkostwo Polski w Unii Europejskiej (TNOiK: ToruĔ, 1999): 33-73 at 50. 94 Id at 54.
352 95
ENDNOTES
Art. 56 (3). Art. 21 (2). 97 Art. 48. 98 Art. 21 (1). 99 Art. 23 (1). 100 Art. 28. 101 On a distinction between the strength and the scope of rights, see James W. Nickel, Making Sense of Human Rights (University of California Press: Berkeley, 1987) at 48-50. 102 Albania, Bosnia and Herzegovina and Latvia. 103 See generally Rett R. Ludwikowski, Constitution-Making in the Region of Former Soviet Dominance (Duke University Press: Durham, 1996) at 230. For an example of duties in a contemporary European constitution, see Spanish Const., art. 30(1) (declaring that citizens have “the right and the duty to defend Spain”) (emphasis added); id. at art. 31(1) (declaring that citizens have the duty to pay taxes); id. at art. 35(1) (declaring that citizens have the duty to work). 104 Belr. Const. art. 55; Bulg. Const. art. 55; Croat. Const. art. 69; Czech Rep. Charter art. 35; Est. Const. art. 53; Geor. Const. art. 37; Lith. Const. art. 53; Maced. Const. art. 43; Mold. Const. art. 59; Pol. Const. art. 86; Rom. Const. art. 41; Russ. Const. art. 58; Slovk. Const. art. 44; Slovn. Const. art. 72; Ukr. Const. art. 66; Mont. Const. art. 19. 105 Belr. Const. art. 32; Bulg. Const. art. 47; Croat. Const. art. 63; Czech Rep. Charter art. 32; Est. Const. art. 27; Hung. Const. art. 67; Maced. Const. art. 40; Mold. Const. art. 48; Rom. Const. art. 44; Russ. Const. art. 38; Slovn. Const. art. 54; Ukr. Const. art. 51; Mont. Const. art. 59; Serb. Const. art. 29. 106 Belr. Const. art. 57, Bulg. Const. art. 59; Croat. Const. art. 47; Est. Const. art. 24; Hung. Const. art. 70/H; Lith. Const. art. 139; Maced. Const. art. 28; Mold. Const. art. 57; Pol. Const. art. 85; Rom. Const. art. 52; Russ. Const. art. 59; Ukr. Const. art. 65; Serb. Const. art. 51. In addition, the Slovak Constitution proclaims cryptically that the defense of the Slovak Republic is “the duty and matter of honor of citizens.” art. 25(1). 107 Croat. Const. art. 47; Est. Const. art. 124; Pol. Const. art. 85; Russ. Const. art. 59. Articles 70/H and 139 of the Hungarian and Lithuanian constitutions, respectively, contain this as an implicit right, suggesting that the duty of protection of the state can be fulfilled by non-military service. 108 Belr. Const. art. 56; Bulg. Const. art. 60; Croat. Const. art. 51; Hung. Const. art. 70/I; Maced. Const. art. 33; Mold. Const. art. 58; Pol. Const. art. 84; Rom. Const. art. 53; Russ. Const. art. 57; Ukr. Const. art. 67; Mont. Const. art. 49; Serb. Const. art. 52. 109 Belr. Const. art. 32; Croat. Const. art. 63; Maced. Const. art. 40; Mold. Const. art. 48; Russ. Const. art. 38; Ukr. Const. art. 51; Mont. Const. art. 59; Serb. Const. art. 29. 110 Bulg. Const. art. 61; Mont. Const. art. 51; Serb. Const. art. 54. 111 Est. Const. art. 54(1). 112 Article 82 of the Polish constitution is typical: “Loyalty to the Republic of Poland, as well as concern for the common good, are the duty of every Polish citizen.” For similar articulations of generalized, state-based duties, see Geor. Const. art. 44; Mold. Const. art. 56; Rom. Const. art. 50(1). 113 “Bulgarian citizens have the right and obligation to study and use the Bulgarian language.” Bulg. Const. art. 36(1). 114 Roman Graczyk, Konstytucja dla Polski (Kraków: Znak, 1997) at 158; see also Maágorzata KoĪuch, “Konstytucyjny obowiązek wiernoĞci Rzeczypospolitej (Art. 82 Konstytucji RP) a czáonkostwo Polski w Unii Europejskiej”, in Cezary Mik, ed., Konstytucja Rzeczypospolitej Polskiej z 1997 roku a czáonkostwo Polski w Unii Europejskiej (ToruĔ: TNOiK, 1999): 353-73 at 357-59. 115 Decision K. 39/97 of 10 November 1998, Orzecznictwo Trybunaáu Konstytucyjnego, Rok 1998 (C.H. Beck: Warszawa, 1999), poz. 26, at 491-557, discussed in Chapter 9 above. 116 Articles 82 and 85 of the Constitution, respectively. 117 Id. at 501. 118 Id. at 520-21. 119 Similarly Graczyk, supra note 114 at 158. 120 Soviet Constitution of 1977, art. 59. 121 Moldova, Poland and Romania. 122 Consider, as an example, article 63(1) of the Croatian Constitution: “Parents shall have the duty to bring up, support, and educate their children, and have the right and freedom to independently decide on the upbringing of their children.” 96
ENDNOTES
353
123
Consider, as an example, article 37(3) of the Georgian Constitution: “All have the right to live in a healthy environment and use natural and cultural surroundings. All are obliged to protect the natural and cultural surroundings”. 124 Bulg. Const. art. 61.
Conclusions 1
Patricia M. Wald, Foreword, in Herman Schwartz, The Struggle for Constitutional Justice in PostCommunist Europe (University of Chicago Press: Chicago, 2000): ix-xvii at x. Wald, a former Chief Judge of the US Court of Appeals for the District of Columbia, is member of the International Criminal Tribunal for the Former Yugoslavia. 2 Herman Schwartz, “The New East European Constitutional Courts”, in A. E. Dick Howard (ed.), Constitution Making in Eastern Europe (Washington, D.C.: Woodrow Wilson Center Press, 1993), pp. 163-208 at p. 194. Another commentator, Spencer Zifcak, expressed his enthusiasm immediately in the title of his article “Hungary’s Remarkable, Radical, Constitutional Court”, Journal of Constit. Law in Eastern and Central Europe 3 (1996): 1-56, and concluded his discussion of the Hungarian Court by saying that “[t]here have been few, more distinctive or valuable judicial contributions to emergent democracy anywhere”, id. at 56. 3 Mark Tushnet, “Policy Distortion and Democratic Debilitation: Comparative Illumination of the Countermajoritarian Difficulty”, Michigan Law Review 94 (1995): 245-301 at 275. 4 J.M. Balkin, “Agreements with Hell and Other Objects of our Faith”, Fordham Law Review 65 (1997): 1703-38 at 1728. 5 László Sólyom, “The Hungarian Constitutional Court and Social Change”, Yale J. Int’l L. 19 (1994): 223-38, 223. 6 Ruti Teitel, “Paradoxes in the Revolution of the Rule of Law”, Yale J. Int’l L. 19 (1994): 239-47 at 246, footnote omitted. 7 Kim Lane Scheppele, “Democracy by Judiciary”, paper presented to the workshop “Rethinking the Rule of Law in Post-Communist Europe”, European University Institute, Florence 28-23 February 2002 (unpublished manuscript on file with the author). 8 This danger nowhere figures in the best book-length description of post-communist judicial review so far, namely Herman Schwartz, The Struggle for Constitutional Justice in Post-Communist Europe (The University of Chicago Press: Chicago, 2000). 9 See, similarly, Stephen Holmes & Cass R. Sunstein, “The Politics of Constitutional Revision in Eastern Europe”, in Sanford Levinson, ed., Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press 1995): 275-306 at 303-4. 10 Keith E. Whittington, “The Political Foundations of Judicial Supremacy”, in Sotirios A. Barber & Robert P. George, Constitutional Politics (Princeton University Press: Princeton 2001): 261-97 at 279. 11 Id. at 280, footnote omitted. 12 Klaus von Beyme, “Constitutional Engineering in Central and Eastern Europe”, in Stephen White, Judy Batt & Paul G. Lewis, Developments in Central and East European Politics, vol. 3 (Palgrave Macmillan: Basingstoke 2003): 190-210 at 208. 13 Marcin Król, “Autorytet parlamentu”, Tygodnik Powszechny (Cracow), 3 November 2002, p. 16. 14 Robert A. Dahl, Democracy and Its Critics (New Haven: Yale University Press, 1989) 189. 15 Interview with Mr Petr Pithart, President of the Senate of the Czech Republic, Prague 21 March 2002. 16 E.g. in Poland, in November 2002, 67 % of respondents had a negative opinion of the work of the Parliament, and only 17 %, a positive opinion. 80 % thought that the Sejm did not have sufficient prestige; two-thirds thought that it worked too slowly, and 50 % believed that it often took wrong decisions; see the results of the CBOS survey as reported in “Polacy bardzo Ĩle o Sejmie”, Rzeczpospolita 14 January 2003 at A2. 17 Teitel, supra note 6 at 246. 18 Stephen Holmes & Cass R. Sunstein, “The Politics of Constitutional Revision in Eastern Europe”, in Sanford Levinson, ed., Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press 1995): 275-306 at 300. 19 Alec Stone Sweet, “Constitutional Dialogues: Protecting Rights in France, Germany, Italy and Spain”, in Sally J. Kenney, William M. Reisinger & John C. Reitz, Constitutional Dialogues in Comparative Perspective (Macmillan: Houndmills, 1999): 8-41 at 26.
354 20
ENDNOTES
Whittington, supra note 10 at 269. Id. at 271, footnote omitted. 22 Stephen Gardbaum, “The New Commonwealth Model of Constitutionalism”, American Journal of Comparative Law 49 (2001): 707-60. 23 Id. at 760. 24 Michael J. Perry, “Protecting Human Rights in a Democracy: What Role for the Courts?”, Wake Forest Law Review 38 (2003): 635-94 at 674-75. 25 Aharon Barak, “The Supreme Court, 2001 Term – Foreword: A Judge on the Judging: The Role of a Supreme Court in a Democracy”, Harvard Law Review 116 (2002): 16-162 at 63. 26 Ruti Teitel, “Transitional Jurisprudence: The Role of Law in Political Transformation”, Yale Law Journal 106 (1997): 2009-2080 at 2034. 27 Decision no. 60/1994 of 24 December 1994, reprinted in László Sólyom & Georg Brunner, Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 306-15 at 312. 28 Decision K 121/93, cited by Leszek Garlicki, “Orzecznictwo w 1993 roku”, Przegląd sądowy (1994:10): 31-53 at 37 29 I should emphasise that I am not attributing such an attitude to either to Aharon Barak, or to Ruti Teitel, or to any other defender of “exceptionalism” in transitional constitutional theory. 30 John Gray, “From Post-Communism to Civil Society: The Reemergence of History and the Decline of the Western Model”, Social Philosophy and Policy (1993): 26-50 at 46. 31 Id. at p. 27. 32 Robert A. Dahl, “Decision-Making in a Democracy: The Supreme Court as a National Policy-Maker”, Journal of Public Law 6 (1957): 279-95 at 283, footnote omitted. 33 See Scheppele, supra note 7. 34 The words in quotation marks are from Christopher L. Eisgruber, Constitutional Self-Government (Harvard University Press: Cambridge Mass., 2001) at 1. They do not represent Eisgruber’s own conception of the grounds of judicial review but the one which he opposes. 21
BIBLIOGRAPHY Ackerman, Bruce, “Beyond Carolene Products”, Harvard Law Review 98 (1985): 713-46. . The Future of Liberal Revolution (Yale University Press: New Haven, 1992). . “The Rise of World Constitutionalism”, Virginia Law Review 83 (1997): 771-97. Acuña, Carlos H., & Catalina Smulovitz, “Guarding the Guardians in Argentina: Some Lessons about the Risks and Benefits of Empowering the Courts”, in A. James McAdams, ed., Transitional Justice in New Democracies (University of Notre Dame Press: Notre Dame, 1996): 93-122. Agh, Attila, “The Permanent ‘Constitutional Crisis’ in the Democratic Transition: The Case of Hungary”, in Joachim Jens Hesse & Nevil Johnson, eds, Constitutional Policy and Change in Europe (Oxford University Press: Oxford, 1995): 296-326. Aleinikoff, T. Alexander, “Constitutional Law in the Age of Balancing”, Yale Law Journal 96 (1987): 943-1005. Alexander, Larry, ed., Constitutionalism: Philosophical Foundations (Cambridge University Press: Cambridge, 1998). Alexy, Robert, “Constitutional Rights, Balancing, and Rationality”, Ratio Juris 16 (2003): 131-40. Allan, James, “Bills of Rights and Judicial Power – a Liberal’s Quandary”, Oxford Journal of Legal Studies 16 (1996): 337-52. Arato, Andrew, “Constitution and Continuity in the Eastern European Transitions: The Hungarian Case (part two)”, in Irena Grudzinska Gross, ed., Constitutionalism & Politics (Slovak Committee of the European Cultural Foundation: Bratislava, 1993): 277-87. Aukerman, Miriam J., “Definitions and Justifications: Minority and Indigenous Rights in a Central/East European Context”, Human Rights Quarterly 22 (2000): 1011-50. Avanesyan, Suren, “Constitutional Protection for Human Rights in the Russian Federation”, Journal of East European Law 6 (1999): 437-68. Balkin, J.M., “Agreements with Hell and Other Objects of our Faith”, Fordham Law Review 65 (1997): 1703-38. Barak, Aharon, “The Supreme Court, 2001 Term – Foreword: A Judge on the Judging: The Role of a Supreme Court in a Democracy”, Harvard Law Review 116 (2002): 16-162. Barber, Sotirios A., & Robert P. George, eds., Constitutional Politics (Princeton University Press: Princeton, 2001). Baric, Sanja, “The Constitutional Court of the Republic of Croatia: Its Institutional Role Within the System of Government”, in Giuseppe de Vergottini, ed., Giustizia costituzionale e sviluppo democratico nei paesi dell’Europa Centro-Orientale (G. Giappichelli Editore: Torino, 2000): 115-25. Barry, Brian, Culture and Equality (Polity Press: Cambridge, 2001). Bartole, Sergio, “Concluding Remarks”, in Giuseppe de Vergottini, ed., Giustizia costituzionale e sviluppo democratico nei paesi dell’Europa Centro-Orientale (G. Giappichelli Editore: Torino, 2000): 355-64. Beatty, David M., ed., Human Rights and Judicial Review: A Comparative Perspective (Kluwer: Dordrecht, 1994). Bell, John, French Constitutional Law (Oxford University Press: Oxford, 1992). Bickel, Alexander M., The Least Dangerous Branch: The Supreme Court at the Bar of Politics (BobbsMerrill: Indianapolis, 1962). Bickel, Alexander M., & Harry H. Wellington, “Legislative Purpose and the Judicial Process”, Harvard Law Review 71 (1957): 1-39. Bollinger, Lee, The Tolerant Society (Oxford University Press: Oxford, 1986). Brewer-Carías, Allan Randolph, Judicial Review in Comparative Law (Cambridge University Press: Cambridge, 1989). Brzezinski, Mark F., “Constitutionalism and Post-Communist Polish Politics”, Loyola of Los Angeles International & Comparative Law Journal 20 (1998): 433-53. Brilmayer, Lea, “The Jurisprudence of Article III: Perspectives on the ‘Case or Controversy’ Requirement”, Harvard Law Review 93 (1979): 297-321. . “A Reply”, Harvard Law Review 93 (1980): 1727-33. Brunner, Georg, “Structure and Proceedings of the Hungarian Constitutional Judiciary”, in László Sólyom & Georg Brunner, eds., Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 65-102. Bugaric, Bojan, “Courts as Policy-Makers: Lessons from Transition”, Harvard International Law Journal 42 (2001): 247-88.
356
BIBLIOGRAPHY
Calhoun, Noel, “The Ideological Dilemma of Lustration in Poland”, East European Politics and Societies 16 (2002): 494-520. Campbell, Tom, K. D. Ewing & Adam Tomkins, eds., Sceptical Essays on Human Rights (Oxford University Press: Oxford, 2001). Cepl, Vojtech, “Ritual Sacrifices”, East European Constitutional Review 1:2 (Spring 1992): 24-6. Cerar, Miroslav, “Slovenia’s Constitutional Court within the Separation of Powers”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 213-46. Chantebout, Bernard, Droit constitutionnel et science politique (A Colin: Paris, 1997) Chapman, John W. & Alan Wertheimer, eds., Majorities and Minorities: Nomos XXXII (New York University Press, New York, 1990). Cheli, Enzo & Filippo Donati, “Methods and Criteria of Judgment on the Questions of Rights to Freedom in Italy”, in David M. Beatty, ed., Human Rights and Judicial Review: A Comparative Perspective (Kluwer: Dordrecht 1994): 227-65. Ciemniewski, Jerzy, “Sejm i Senat w projekcie Konstytucji RP”, in Józef Krukowski, ed., Ocena projektu Konstytucji RP (Towarzystwo Naukowe Katolickiego Uniwersytetu Lubelskiego: Lublin, 1996): 3747. Clark, David S., ed., Comparative and Private International Law: Essays in Honor of John Henry Merryman on His Seventieth Birthday (Duncker u. Humblot: Berlin, 1990). Czarnota, Adam & Piotr Hofmanski, “Can we do Justice to the Past?”, in Martin Krygier & Adam Czarnota, eds., The Rule of Law after Communism (Ashgate: Dartmouth, 1999): 197-211. Czeszejko-Sochacki, Zdzisáaw, Leszek Garlicki & Janusz TrzciĔski, Komentarz do ustawy o Trybunale Konstytucyjnym (Wydawnictwo Sejmowe, Warszawa 1999). Dahl, Robert A., “Decision-Making in a Democracy: The Supreme Court as a National Policy-Maker”, Journal of Public Law 6 (1957): 279-95. . Democracy and Its Critics (Yale University Press: New Haven, 1989). de Vergottini, Guiseppe, ed., Giustizia costituzionale e sviluppo democratico nei paesi dell’Europa Centro-Orientale (G. Giappichelli Editore: Torino, 2000). Deets, Stephen, “Reconsidering East European Minority Policy: Liberal Theory and European Norms”, East European Politics & Societies 16 (2002): 30-53. Devins, Neal, Shaping Constitutional Values (The Johns Hopkins University Press: Baltimore, 1996). . “The Democracy-Forcing Constitution”, Michigan Law Review 97 (1999): 1971-93. Di Gregorio, Angela, “The Evolution of Constitutional Justice in Russia: Normative Imprecision and the Conflicting Positions of Legal Doctrine and Case-Law in Light of the Constitutional Court Decision of 16 June 1998”, Review of Central and East European Law 24 (1998): 387-419. Dimitrijeviü, Nenad, “Ethno-Nationalized States of Eastern Europe: Is there a Constitutional Alternative?”, Studies in East European Thought 54 (2002): 246-69. DomaĔska, Aldona, “Analiza treĞci konstytucyjnej zasady równoĞci w oparciu o wybrane orzeczenia Trybunaáu Konstytucyjnego”, Studia Prawno-Ekonomiczne 62 (2000): 47-58. Dorf, Michael C., “The Supreme Court, 1997 Term – Foreword: The Limits of Socratic Deliberation”, Harvard Law Review 112 (1998): 4-83. Dupré, Catherine, “The Right to Human Dignity in Hungarian Constitutional Case Law”, in The Principle of Respect for Human Dignity (Council of Europe Publishing: Strasbourg 1999): 68-79. .“Importing Human Dignity from German Constitutional Case Law”, in Gabor Halmai, ed., The Constitution Found? The First Nine Years of Hungarian Constitutional Review on Fundamental Rights (Indok: Budapest 2000): 215-26. Dworkin, Ronald, Taking Rights Seriously (Duckworth: London, 1977). . A Matter of Principle (Harvard University Press: Cambridge, Massachusetts, 1985). . Law’s Empire (Fontana: London, 1986). . Freedom’s Law: The Moral Reading of the American Constitution (Oxford University Press: Oxford, 1996). . Sovereign Virtue: The Theory and Practice of Equality (Harvard University Press: Cambridge, Massachusetts, 2000). Ebzeev, Boris, “Interview”, East European Constitutional Review 6:1 (Winter 1997): 83-8. Eisgruber, Christopher L., “Is the Supreme Court an Educative Institution?”, New York University Law Review 67 (1992), 961-1032. . Constitutional Self-Government (Harvard University Press: Cambridge, Massachusetts, 2001).
BIBLIOGRAPHY
357
Elster, Jon, “Constitution-Making in Eastern Europe: Rebuilding the Boat in the Open Sea”, Public Administration 71 (Spring/Summer 1993): 169-217. .“Majority Rule and Individual Rights”, in Stephen Shute and Susan Hurley, eds., On Human Rights (Basic Books: New York, 1994): 175-216. . “The Impact of Rights on Economic Performance”, in Andras Sajó, ed., Western Rights? PostCommunist Application (Kluwer: Dordrecht, 1996): 347-60. . “Coming to Terms with the Past. A Framework for the Study of Justice in the Transition to Democracy”, Arch. European Sociology 39 (1998): 7-48. . Ulysses Unbound: Studies in Rationality, Precommitment and Constraints (Cambridge University Press: Cambridge, 2000). . “Retribution in the Transition to Democracy”, in Arend Soeteman, ed., Pluralism and Law (Kluwer: Dordrecht, 2001): 19-36. Elster, Jon, Claus Offe & Ulrich K. Preuss, Institutional Design in Post-communist Societies (Cambridge University Press: Cambridge, 1998). Ely, John Hart, Democracy and Distrust: A Theory of Judicial Review (Harvard University Press: Cambridge, Massachusetts, 1980). Epp, Charles R., The Rights Revolution (University of Chicago Press: Chicago, 1998). Epstein, Lee, Jack Knight & Olga Shvetsova, “The Role of Constitutional Courts in the Establishment and Maintenance of Democratic Systems of Government”, Law & Society Review 35 (2001): 117-63. Eriksen, Erik Oddvar, John Erik Fossum & Agustín José Menéndez, The Chartering of Europe: The Charter of Fundamental Rights in Context (ARENA Report 8/2001: Oslo, 2001). Ewing, K. D., “Social Rights and Constitutional Law”, Public Law (1999): 104-23. Fabre, Cécile, “Constitutionalising Social Rights”, Journal of Political Philosophy 6 (1998): 263-84. . “The Dignity of Rights”, Oxford Journal of Legal Studies 20 (2000): 271-282. Fallon, Richard H., “The Supreme Court, 1996 Term – Foreword: Implementing the Constitution”, Harvard Law Review 111 (1997): 54-152. Favoreu, Louis, “American and European Models of Constitutional Justice”, in David S. Clark, ed., Comparative and Private International Law: Essays in Honor of John Henry Merryman on His Seventieth Birthday (Duncker u. Humblot: Berlin, 1990): 105-20. Feinberg, Joel, The Moral Limits of the Criminal Law, vol. 3: Harm to Self (Oxford University Press: New York, 1986). Ferejohn, John & Pasquale Pasquino, “Constitutional Courts as Deliberative Institutions: Towards an Institutional Theory of Constitutional Justice,” in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 21-36. Fijalkowski, Agata, “The Abolition of the Death Penalty in Central and Eastern Europe”, Tilburg Foreign Law Review 9 (2001): 62-83. Fiss, Owen, “The Forms of Justice”, Harvard Law Review 93 (1979): 1-58. . “Judiciary Panel: Introductory Remarks”, Yale Journal of International Law 19 (1994): 219-22. Fletcher, George, Basic Concepts of Legal Thought (Oxford University Press: New York, 1995). Frankowski, Stanisáaw, “Lech WaáĊsa’s Draft of the Charter of Rights and Freedoms: An Overview”, Saint Louis-Warsaw Transatlantic Law Journal (1996): 65-72. Fredman, Sandra, ed., Discrimination and Human Rights: The Case of Racism (Oxford University Press: Oxford, 2001). Fudge, Judy, “The Canadian Charter of Rights: Recognition, Redistribution, and the Imperialism of the Courts”, in Tom Campbell, K. D. Ewing & Adam Tomkins, eds., Sceptical Essays on Human Rights (Oxford University Press: Oxford, 2001): 335-58. Ganev, Venelin I., “Interview with Constitutional Court Justices Todor Todorov and Tsanko Hadjistoichev”, East European Constitutional Review 6:1 (Winter 1997): 65-71. . “Bulgaria: The (Ir)Relevance of Post-communist Constitutionalism”, in Jan Zielonka, ed., Democratic Consolidation in Eastern Europe, vol. 1: Institutional Engineering (Oxford University Press, Oxford 2001): 186-211. . “The Rise of Constitutional Adjudication in Bulgaria”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 247-64. . “Foxes, Hedgehogs and Learning: Notes on the Past and Future Dilemmas of Postcommunist Constitutionalism”, paper presented at the workshop “Rethinking the Rule of Law in Post-
358
BIBLIOGRAPHY
Communist Europe”, European University Institute, Florence, 22-23 February 2002 (unpublished manuscript on file with author). .“The Bulgarian Constitutional Court, 1991-1997: A Success Story in Context”, Europe-Asia Studies 55 (2003): 597-611. Gardbaum, Stephen, “The New Commonwealth Model of Constitutionalism”, American Journal of Comparative Law 49 (2001): 707-60. Garlicki, Leszek, “Przegląd orzecznictwa Trybunaáu Konstytucyjnego w 1991 roku”, Przegląd Sądowy (1991:11-12): 43-63. . “Orzecznictwo Trybunaáu Konstytucyjnego w 1993 roku”, Przegląd Sądowy (1994:10): 31-52. . “Orzecznictwo Trybunaáu Konstytucyjnego w 1995 roku”, Przegląd Sądowy (1996:7-8): 110-38. . “Trybunaá Konstytucyjny w projekcie Komisji Konstytucyjnej Zgromadzenia Narodowego”, PaĔstwo i Prawo 51:2 (1996): 3-19. . “Orzecznictwo Trybunaáu Konstytucyjnego w 1997 roku”, Przegląd Sądowy (1998:6): 35-58. . ed., Konstytucja Rzeczypospolitej Polskiej: Komentarz (Wydawnictwo Sejmowe, Warszawa, 1999). . “Orzecznictwo Trybunaáu Konstytucyjnego w 1998 roku”, Przegląd Sądowy (1999:6): 104-28. . “Orzecznictwo Trybunaáu Konstytucyjnego w 2000 roku”, Przegląd Sądowy (2001:9): 77-105. . “Perspectives on Freedom of Conscience and Religion in the Jurisprudence of Constitutional Courts”, Brigham Young University Law Review (2001): 467-510. . “The Experience of the Polish Constitutional Court”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 265-82. George, Robert P., ed., Natural Law Theory: Contemporary Essays (Oxford University Press: Oxford, 1992). Gibson, James L., Gregory A. Caldeira & Vanessa A. Baird, “On the Legitimacy of National High Courts”, American Political Science Review 92 (1998): 343-58. Gicquel, Jean, Droit constitutionnel et institutions politiques, 14th edition (Montchrestien: Paris, 1995). Gillis, Mark, “Constitutionalism in the Czech Republic: An Investigation of Two Major Aspects – The Infusion of Constitutional Principles into the Remainder of the Legal Order and the Generally Binding Nature of Constitutional Court Decisions”, Journal of Constitutional Law in Eastern and Central Europe 5 (1998): 105-245. Ginsburg, Tom, “Economic Analysis and the Design of Constitutional Courts”, Theoretical Inquiries in Law 3 (2002): 49-85. . “Constitutional Courts in New Democracies: Understanding Variation in East Asia”, Global Jurist Advances vol. 2:1 (2002). http://www.bepress.com/gj/advances/vol2/iss1/art4/. Gottlieb, Stephen E., ed., Public Values in Constitutional Law (University of Michigan Press: Ann Arbor, 1993). Graczyk, Roman, Konstytucja dla Polski (Znak: Kraków, 1997). Granat, Mirosáaw, “Droga do sądownictwa konstytucyjnego w paĔstwach Europy ĝrodkowej i Wschodniej”, PaĔstwo i Prawo no. 12/2001 (vol. 56): 15-24. Gray, John, “From Post-Communism to Civil Society: The Reemergence of History and the Decline of the Western Model”, Social Philosophy and Policy (1993): 26-50. Greenberg, Jack, “Affirmative Action in Higher Education: Confronting the Condition and Theory”, Boston College Law Review 43 (2002): 521-621. Griffin, Stephen M., American Constitutionalism: From Theory to Politics (Princeton University Press: Princeton, 1996). Grimm, Dieter, “Human Rights and Judicial Review in Germany”, in David M. Beatty, ed., Human Rights and Judicial Review: A Comparative Perspective (Kluwer: Dordrecht, 1994): 267-96. Grudzinska-Gross, Irena, ed., Constitutionalism & Politics (Slovak Committee of the European Cultural Foundation: Bratislava, 1993). . “Broadcasting Values”, East European Constitutional Review 2:3 (Summer 1993): 51-53. . “Interview with Professor Andrzej Zoll, Chief Justice of the Polish Constitutional Tribunal”, East European Constitutional Review 6:1 (Winter 1997): 77-8. Gunther, Gerald, “The Supreme Court, 1971 Term – Foreword: In Search of Evolving Doctrine on a Changing Court: A Model for a Newer Equal Protection”, Harvard Law Review 86 (1972): 1-48. Habermas, Jürgen, Between Facts and Norms, trans. William Rehg (Polity Press: Cambridge, 1996). . The Postnational Constellation: Political Essays, trans. Max Pensky (Polity Press: Cambridge, 2001).
BIBLIOGRAPHY
359
Halmai, Gábor, “Comment: The Constitutional Court of the Republic of Hungary”, East European Case Reporter of Constitutional Law 1 (1994): 116-118. . “Criminal Law as Means Against Racist Speech? The Hungarian Legal Approach”, Journal of Constitutional Law in Eastern and Central Europe 4 (1997): 41-52. . ed., The Constitution Found? The First Nine Years of Hungarian Constitutional Review on Fundamental Rights (Indok: Budapest 2000). .“The Hungarian Approach to Constitutional Review: The End of Activism? The First Decade of the Hungarian Constitutional Court”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 189-212. . “Who is the Main Protector of Fundamental Rights in Hungary? The Role of the Constitutional Court and the Ordinary Courts”, in JiĜí PĜibáĖ, Pauline Roberts & James Young, eds., Systems of Justice in Transition: Central European Experiences since 1989 (Ashgate: Aldershot, 2003): 50-73. Halmai, Gábor, & Kim Lane Schepple, “Living Well is the Best Revenge: The Hungarian Approach to Judging the Past” in A. James McAdams, ed., Transitional Justice in New Democracies (University of Notre Dame Press: Notre Dame, 1996): 155-84. Hamon, Francis, Michel Troper & Georges Burdeau, Droit constitutionnel, 27th edition (L.G.D.J.: Paris, 2001). Helsinki Committee for Human Rights, “Gender Equality: Legal and Institutional Framework On Women’s Rights and Equal Opportunities; De Jure And De Facto Discrimination In Poland”, Polish Law Journal 6 (2001): 149-228. Henckaerts, Jean-Marie, & Stefaan Van der Jeught, "Human Rights Protection Under the New Constitutions of Central Europe", Loyola of Los Angeles International & Comparative Law Review (1998): 475-506. Hepple, Bob, “A Right to Work?”, Industrial Law Journal 10 (1981): 65-83. Hesse, Joachim Jens, & Nevil Johnson, eds., Constitutional Policy and Change in Europe (Oxford University Press: Oxford, 1995). Holländer, Pavel, “The Role of the Czech Constitutional Court: Application of the Constitution in Case Decisions of Ordinary Courts”, Parker School Journal of East European Law 4 (1997): 445-65. Holmes, Stephen, & Cass Sunstein, “The Politics of Constitutional Revision in Eastern Europe”, in Sanford Levinson, ed., Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press: Princeton, 1995): 275-306. Howard, A.E. Dick, ed., Constitution Making in Eastern Europe (Woodrow Wilson Center Press: Washington D.C., 1993). . “The Indeterminacy of Constitutions”, Wake Forest Law Review 31 (1996): 383-410. Imholz, Kathleen, “A Landmark Constitutional Court Decision in Albania”, East European Constitutional Review 2:3 (Summer 1993) at 23-5. Jackson, Donald W., & C. Neal Tate, Comparative Judicial Review and Public Policy (Greenwood Press: Westport, Connecticut, 1992). Jasudowicz, Tadeusz, “Granice korzystania z praw czáowieka - rozwiązania Konstytucji RP na tle standartów europejskich”, in Cezary Mik, ed., Konstytucja Rzeczypospolitej Polskiej z 1997 roku a czáonkostwo Polski w Unii Europejskiej (TNOiK: ToruĔ, 1999): 33-73. Jowitt, Ken, New World Disorder. The Leninist Extinction (University of California Press: Berkeley, 1992). Katrougalos, George S., “The Implementation of Social Rights in Europe”, Columbia Journal of European Law 2 (1996): 277-312. Kenney, Sally J., William M. Reisinger & John C. Reitz, eds., Constitutional Dialogues in Comparative Perspective (Macmillan: London, 1999). Kolarova, Rumyana, “Bulgaria: A Self-Restricting Court”, East European Constitutional Review 2:2 (Spring 1993): 48-50. Komesar, Neil K., “Taking Institutions Seriously: Introduction to a Strategy for Constitutional Analysis”, University of Chicago Law Review 51 (1984): 366-446. . Imperfect Alternatives (The University of Chicago Press: Chicago, 1994). Konstantinov, Emil, “Turkish Party in Bulgaria Allowed to Continue”, East European Constitutional Review 1:3 (Summer 1992): 11-12. Kopecky, Petr, “The Czech Republic: From the Burden of the Old Federal Constitution to the Constitutional Horse Trading Among Political Parties” , in Jan Zielonka, ed., Democratic
360
BIBLIOGRAPHY
Consolidation in Eastern Europe, vol. 1: Institutional Engineering (Oxford University Press: Oxford, 2001): 319-346. Kovács, János Mátyas, “Approaching the EU and Reaching the US? Rival Narratives on Transforming Welfare Regimes in East-Central Europe”, in Peter Mair & Jan Zielonka, eds., The Enlarged European Union: Diversity and Adaptation (Frank Cass: London, 2002): 175-204. KoĪuch, Maágorzata, “Konstytucyjny obowiązek wiernoĞci Rzeczypospolitej (Art. 82 Konstytucji RP) a czáonkostwo Polski w Unii Europejskiej”, in Cezary Mik, ed., Konstytucja Rzeczypospolitej Polskiej z 1997 roku a czáonkostwo Polski w Unii Europejskiej (TNOiK: ToruĔ, 1999): 353-73. Krukowski, Józef, ed., Ocena projektu Konstytucji RP (Towarzystwo Naukowe Katolickiego Uniwersytetu Lubelskiego: Lublin, 1996). Krygier, Martin, “Poland: Life in an Abnormal Country”, The National Interest 18 (Winter 1989/90): 5564. Krygier, Martin, & Adam Czarnota, eds., The Rule of Law after Communism (Ashgate: Dartmouth, 1999). Kymlicka, Will, “Western Political Theory and Ethnic Relations in Eastern Europe”, in Will Kymlicka & Magda Opalski, eds., Can Liberal Pluralism be Exported? Western Political Theory and Ethnic Relations in Eastern Europe (Oxford University Press: Oxford, 2001): 13-106. Kymlicka, Will, & Magda Opalski, eds., Can Liberal Pluralism be Exported? Western Political Theory and Ethnic Relations in Eastern Europe (Oxford University Press: Oxford, 2001). Lambert, Edouard, Le gouvernement des juges et la lutte contre la législation sociale aux Etats-Unis; l'expérience américaine du contrôle judiciaire de la constitutionalité des lois (Giard: Paris, 1921). Landes, William L., & Richard A. Posner, “The Economics of Anticipatory Adjudication”, Journal of Legal Studies 223 (1994): 683-719. Landfried, Christine, ed., Constitutional Review and Legislation (Nomos: Baden-Baden, 1988). Letki, Natalia, “Lustration and Decommunisation in East-Central Europe”, Europe-Asia Studies 54 (2002): 529-52. Levinson, Sanford, ed., Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press: Princeton, 1995). Ludwikowska, Anna M., Sądownictwo konstytucyjne w Europie ĝrodkowo-Wschodniej w okresie przeksztaáceĔ demokratycznych (TNOiK: Torun, 1997). Ludwikowski, Rett R., Constitution-Making in the Region of Former Soviet Dominance (Duke University Press: Durham, 1996). MacCormick, D. Neil & Robert S. Summers, Interpreting Precedents: A Comparative Study (Dartmouth: Aldershot, 1997). Macey, J.R., “Transaction Costs and the Normative Elements of the Public Choice Model: An Application to Constitutional Theory”, Virginia Law Review 74 (1988): 471-518. . “Thayer, Nagel, and the Founders’ Design: A Comment”, Northwestern University Law Review 88 (1993): 226-40. Mair, Peter, & Jan Zielonka, eds., The Enlarged European Union: Diversity and Adaptation (Frank Cass: London, 2002). Malová, Darina, “The Role and Experience of the Slovakian Constitutional Court”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 349-72. Massias, Jean-Piere, Droit constitutionnel des États d'Europe de l'Est (Presses Universitaires de France: Paris, 1999). Masternak-Kubiak, Maágorzata, Ustawa o Trybunale Konstytucyjnym (Wydawnictwa Prawnicze PWN, Warszawa, 1998). Mavi, Viktor, “The Right to Health and the New East European Constitutions”, Journal of Constitutional Law in Eastern and Central Europe 3 (1996): 213-24. McAdams, A. James, ed., Transitional Justice in New Democracies (University of Notre Dame Press: Notre Dame, 1996). Menéndez, Agustín José, “The Sinews of Peace: Rights to Solidarity in the Charter of Fundamental Rights of the Union”, in Erik Oddvar Eriksen, John Erik Fossum & Agustín José Menéndez, The Chartering of Europe: The Charter of Fundamental Rights in Context (ARENA Report 8/2001: Oslo, 2001): 201-26. Michnik, Adam, “The Rebirth of Civil Society”, Public lecture at the London School of Economics, 20 October 199, http://www.lse.ac.uk/Depts/global/Michnik89.htm. Mik, Cezary, ed., Konstytucja Rzeczypospolitej Polskiej z 1997 roku a czáonkostwo Polski w Unii Europejskiej (TNOiK: ToruĔ, 1999).
BIBLIOGRAPHY
361
Mikva, Abner J., “Why Judges Should Not Be Advicegivers: A Response to Professor Neal Katyal”, Stanford Law Review 50 (1998): 1825-32. Mink, Andras, “Interview with László Sólyom, President of the Hungarian Constitutional Court”, East European Constitutional Review 6:1 (Winter 1997): 71-77. Moore, Michael S., “Law as a Functional Kind”, in Robert P. George, ed., Natural Law Theory: Contemporary Essays (Oxford University Press: Oxford, 1992): 188-242. Morawski, Lech & Marek Zirk-Sadowski, “Precedent in Poland”, in D. Neil MacCormick & Robert S. Summers, Interpreting Precedents: A Comparative Study (Dartmouth: Aldershot, 1997): 219-58. Morshchakova, Tamara, “Panel discussion”, in Konstytucja w sáuĪbie demokracji; Constitution in Service of Democracy, conference papers (The International Centre for Development of Democracy Foundation: Cracow, March 10-12, 1995). Mowbray, Alastair, Cases and Materials on the European Convention of Human Rights (Butterworths: London, 2001). Mungiu-Pippidi, Alina, “Interview with President of the Romanian Constitutional Court, Ion Muraru”, East European Constitutional Review 6:1 (Winter 1997): 78-83. Neuborne, Burt, “Judicial Review and Separation of Powers in France and the United States”, New York University Law Review 57 (1982): 363-442. Nickel, James W., Making Sense of Human Rights (University of California Press: Berkeley, 1987). Nino, Carlos Santiago, Radical Evil on Trial (Yale University Press: New Haven, 1996). Oniszczuk, Jerzy, Konstytucja Rzeczypospolitej Polskiej w orzecznictwie Trybunaáu Konstytucyjnego (Zakamycze: Kraków, 2000). OsiatyĔski, Wiktor, “A Bill of Rights for Poland”, East European Constitutional Review 1:4 (Fall 1992): 29-32. . “Agent Waáesa?”, East European Constitutional Review 1:3 (Summer 1992): 28-30. . “Rights in New Constitutions of East Central Europe”, Columbia Human Rights Law Review 26 (1994): 111-66. . “Social and Economic Rights in a New Constitution for Poland”, in Andras Sajó, ed., Western Rights? Post-Communist Application (Kluwer: Dordrecht, 1996): 233-69. . “Paradoxes of Constitutional Borrowing”, I.CON International Journal of Constitutional Law 1 (2003): 244-68. Ossipov, Alexander, “Some Doubts about ‘Ethnocultural Justice’”, in Will Kymlicka & Magda Opalski, eds., Can Liberal Pluralism be Exported? Western Political Theory and Ethnic Relations in Eastern Europe (Oxford University Press: Oxford, 2001): 171-85. Paczolay, Péter, “Human Rights and Minorities in Hungary”, Journal of Constitutional Law in Eastern and Central Europe 3 (1996): 111-26. Pap, András László, “Representation or Ethnic Balance: Ethnic Minorities in Parliaments”, Journal of East European Law 7 (2000): 261-339. Pasquino, Pasquale, “Lenient Legislation: The Italian Constitutional Court” (1999, unpublished manuscript on file with author). Perry, Michael J., “The Constitution, the Courts, and the Question of Minimalism”, Northwestern University Law Review 88 (1993): 84-164. . “Protecting Human Rights in a Democracy: What Role for the Courts?”, Wake Forest Law Review 38 (2003): 635-94. Petrov, Nikolai, “Political Institutions and the Regulation of Ethnic Conflicts: Russia's Experience”, paper presented at Conference on “Legal Framework to Facilitate the Settlement of Ethno-Political Conflicts in Europe”, Baku, 11-12 January 2002 (unpublished manuscript on file with author). Petrova, Dimitrina, “Racial Discrimination and the Rights of Minority Cultures”, in Sandra Fredman, ed., Discrimination and Human Rights: The Case of Racism (Oxford University Press: Oxford, 2001). Pettai, Vello, “Definitions and Discourse: Applying Kymlicka's Models to Estonia and Latvia”, in Will Kymlicka & Magda Opalski, eds., Can Liberal Pluralism be Exported? Western Political Theory and Ethnic Relations in Eastern Europe (Oxford University Press: Oxford, 2001): 259-69. . “Democratic Norm Building and Constitutional Discourse Formation”, paper presented at the workshop “Rethinking the Rule of Law in Post-Communist Europe”, European University Institute, Florence, 22-23 February 2002 (unpublished manuscript on file with author). Pettit, Philip, Republicanism (Oxford University Press: Oxford, 1997). Pizzorusso, Alessandro, Il patrimonio costituzionale europeo (Il Mulino: Bologna, 2002).
362
BIBLIOGRAPHY
Posner, Eric A. & Adrian Vermeule, “Transitional Justice as Ordinary Justice”, Harvard Law Review 117 (2004): 761-825. Preuss, Ulrich K., “Patterns of Constitutional Evolution and Change in Eastern Europe”, in Joachim Jens Hesse & Nevil Johnson, eds., Constitutional Policy and Change in Europe (Oxford University Press: Oxford, 1995): 95-126. PĜibáĖ, JiĜí, “Moral and Political Legislation in Constitutional Justice: A Case Study of the Czech Constitutional Court”, Journal of East European Law 8 (2001): 15-34. .“Judicial Power vs. Democratic Representation: The Culture of Constitutionalism and Human Rights in the Czech Legal System”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 373-94. PĜibáĖ, JiĜí, Pauline Roberts & James Young, eds., Systems of Justice in Transition: Central European Experiences since 1989 (Ashgate: Aldershot, 2003). Prochàzka, Radoslav, Mission Accomplished: On Founding Constitutional Adjudication in Central Europe (CEU Press: Budapest, 2002). Prosser, William J., “Privacy [a legal analysis,]” in Ferdinand D. Schoeman, ed., Philosophical Dimensions of Privacy (Cambridge University Press: Cambridge, 1984): 104-55. Pushaw, Robert J., “Justiciability and Separation of Powers: A Neo-Federalist Approach”, Cornell Law Review 81 (1996): 393-512. Rapaczynski, Andrzej, “Constitutional Politics in Poland: A Report on the Constitutional Committee of the Polish Parliament”, in A.E. Dick Howard, ed., Constitution Making in Eastern Europe (Woodrow Wilson Center Press: Washington D.C., 1993): 93-131. Raskin, Jamin B., Overruling Democracy: the Supreme Court vs. the American People (Routledge: New York, 2003). Ratner, L.G., “Majoritarian Constraints on Judicial Review: Congressional Control of Supreme Court Jurisdiction”, Villanova Law Review 27 (1981): 929-58. Rawls, John, Political Liberalism (Columbia University Press: New York, 1993). Redish, Martin H., “Congressional Power to Regulate Supreme Court Appellate Jurisdiction Under the Exceptions Clause: An Internal and External Examination”, Villanova Law Review 27 (1982): 90028. Reitz, John C., “Political Economy and Abstract Review in Germany, France and the United States”, in Sally J. Kenney, William M. Reisinger & John C. Reitz, eds., Constitutional Dialogues in Comparative Perspective (Macmillan: Houndmills, 1999): 74-84. Rose, Mihaela Serban, ed., Constitutionalism in Transition: Africa and Eastern Europe (The Helsinki Foundation for Human Rights: Warsaw, 2003). Rosenfeld, Michel, comments in the symposium “Dilemmas of Justice”, East European Constitutional Review 1:2 (Summer 1992): 17-22. . ed., Constitutionalism, Identity, Difference, and Legitimacy: Theoretical Perspectives (Duke University Press: Durham, 1994). . “Constitution-Making, Identity Building, and Peaceful Transition to Democracy: Theoretical Reflections Inspired by the Spanish Example”, Cardozo Law Review 19 (1998): 1891-920. . “Can Human Rights Bridge the Gap between Universalism and Cultural Relativism? A Pluralist Assessment Based on the Rights of Minorities”, Columbia Human Rights Law Review 30 (1999): 249-84. Russell, Peter H., “The Growth of Canadian Judicial Review and the Commonwealth and American Experiences”, in Donald W. Jackson & C. Neal Tate, Comparative Judicial Review and Public Policy (Greenwood Press: Westport, Connecticut, 1992): 29-39. Rzeplinski, Andrzej, “The Polish Bill of Rights and Freedoms: A Case Study of Constitution-Making in Poland”, East European Constitutional Review 2:3 (Summer 1993): 26-29. Sadurski, Wojciech, “Conventional Morality and Judicial Standards”, Virginia Law Review 73 (1987): 339-97. . Teoria sprawiedliwoĞci (PWN: Warszawa 1988). . Moral Pluralism and Legal Neutrality (Kluwer: Dordrecht, 1990). . “The Concept of Legal Equality and an Underlying Theory of Discrimination”, Saint LouisWarsaw Transatlantic Law Journal (1998): 63-104. . “Charter and Enlargement”, European Law Journal 8 (2002): 340-62. . ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002).
BIBLIOGRAPHY
363
Sajó, Andras, “Protecting Nation States and National Minorities: A Modest Case for Nationalism in Eastern Europe”, University of Chicago Law School Roundtable (1993): 53-74. . “Hate Speech for Hostile Hungarians”, East European Constitutional Review 3:2 (Spring 1994): 82-87. . “How the Rule of Law Killed Hungarian Welfare Reform”, East European Constitutional Review 5:1 (Winter 1996): 31-41. . ed., Western Rights? Post-Communist Application (Kluwer: Dordrecht, 1996). . ed., Out of and Into Authoritarian Law (Kluwer Law International: The Hague, 2003). . “Welfare Rights in the Post-Communist Constitutional Experience”, in Mihaela Serban Rose, ed., Constitutionalism in Transition: Africa and Eastern Europe (The Helsinki Foundation for Human Rights: Warsaw, 2003): 41-73. Salzberger, Eli M. & Stefan Voigt, “On the Delegation of Powers: With Special Emphasis on Central and Eastern Europe”, Constitutional Political Economy 13 (2002): 25-52. Scharpf, Fritz W., “Judicial Review and the Political Question: A Functional Analysis”, Yale Law Journal 75 (1966): 517-97. Scheppele, Kim Lane, “Women’s Rights in Eastern Europe”, East European Constitutional Review 4:1 (Winter 1995): 66-69. . “Democracy by Judiciary”, paper presented to the workshop “Rethinking the Rule of Law in PostCommunist Europe”, European University Institute, Florence, 28-23 February 2002 (unpublished manuscript on file with author). Schlink, Bernhard, “German Constitutional Culture in Transition”, in Michel Rosenfeld, ed., Constitutionalism, Identity, Difference, and Legitimacy: Theoretical Perspectives (Duke University Press: Durham, 1994): 197-222. Schoeman, Ferdinand D., ed., Philosophical Dimensions of Privacy (Cambridge University Press: Cambridge, 1984). Schulhofer, Stephen J., comments in the symposium “Dilemmas of Justice”, East European Constitutional Review 1:2 (Summer 1992): 17-22. Schwartz, Herman, “In Defense of Aiming High”, East European Constitutional Review 1:4 (Fall 1992): 25-28. . “The New East European Constitutional Courts”, in A.E. Dick Howard, ed., Constitution Making in Eastern Europe (Woodrow Wilson Center Press: Washington D.C., 1993): 163-208. . “The Czech Constitutional Court Decision on the Illegitimacy of the Communist Regime”, Parker School Journal of East European Law 1 (1994) 392-8. . “Panel discussion”, in Konstytucja w sáuĪbie demokracji; Constitution in Service of Democracy, conference papers (The International Centre for Development of Democracy Foundation: Cracow, March 10-12, 1995). . The Struggle for Constitutional Justice in Post-Communist Europe (University of Chicago Press: Chicago, 2000). Shapiro, Martin, Courts: A Comparative and Political Analysis (University of Chicago Press, Chicago, 1981). . “The Success of Judicial Review”, in Sally J., William M. Reisinger & John C. Reitz, eds., Constitutional Dialogues In Comparative Perspective (Macmillan: London, 1999): 193-219. . “Some Conditions for the Success of Constitutional Courts: Lessons from the U.S. Experience” in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 37-60. Shue, Henry, Basic Rights (Princeton University Press: Princeton, 1980). Shute, Stephen & Susan Hurley, eds., On Human Rights (Basic Books: New York, 1994). Šiklová, JiĜina, “Lustration or the Czech Way of Screening”, in Martin Krygier & Adam Czarnota, eds., The Rule of Law after Communism (Ashgate: Dartmouth, 1999): 248-58. Skach, Cindy, “Rethinking Judicial Review: Shaping the Toleration of Difference?”, paper presented at the workshop “Rethinking the Rule of Law in Post-Communist Europe”, European University Institute, Florence, 22-23 February 2002 (unpublished manuscript on file with author). Soeteman, Arend, ed., Pluralism and Law (Kluwer: Dordrecht, 2001). Sokolewicz, Wojciech, “Sąd Konstytucyjny w Rumunii”, in Janusz TrzciĔski, ed., Sądy konstytucyjne w Europie, vol. 2 (Wydawnictwa Trybunaáu Konstytucyjnego: Warszawa,, 1997): 145-74. Sólyom, László, “The Hungarian Constitutional Court and Social Change”, Yale Journal of International Law 19 (1994): 223-38.
364
BIBLIOGRAPHY
. “Interview”, East European Constitutional Review 6:1 (Winter 1997): 71-7. . “Introduction to the Decisions of the Constitutional Court of the Republic of Hungary”, in László Sólyom & Georg Brunner, eds., Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000): 1-64. Sólyom, László & Georg Brunner, eds., Constitutional Judiciary in a New Democracy: The Hungarian Constitutional Court (University of Michigan Press: Ann Arbor, 2000). Sonobe, Itsuo, “Human Rights and Constitutional Review in Japan”, in David M. Beatty, ed., Human Rights and Judicial Review: A Comparative Perspective (Kluwer: Dordrecht, 1994): 135-174. Sprague, Leigh, “The Russian Constitutional Court”, Parker School Journal of East European Law 4 (1997): 339-56. Stone, Alec, The Birth of Judicial Politics in France (Oxford University Press: New York, 1992). Stone Sweet, Alec, “Constitutional Dialogues: Protecting Rights in France, Germany, Italy and Spain”, in Sally J. Kenney, William M. Reisinger & John C. Reitz, eds., Constitutional Dialogues in Comparative Perspective (Macmillan: London, 1999): 8-41. . Governing with Judges (Oxford University Press: Oxford, 2000). Sullivan, Kathleen M., “Categorization, Balancing, and Government Interests”, in Stephen E. Gottlieb, ed., Public Values in Constitutional Law (University of Michigan Press: Ann Arbor, 1993): 241-68. Sunstein, Cass, The Partial Constitution (Harvard University Press: Cambridge, Massachusetts, 1993). . “Against Positive Rights”, Eastern European Constitutional Review 3 (Winter 1993): 35-9. . “Introduction: The Legitimacy of Constitutional Courts: Notes on Theory and Practice”, East European Constitutional Review 6:1 (Winter 1997): 61-3. . Designing Democracy: What Constitutions Do (Oxford University Press: Oxford, 2001). Szczerbiak, Aleks, “Dealing with the Communist Past or the Politics of the Present? Lustration in PostCommunist Poland”, Europe-Asia Studies 54 (2002): 553-72. Teitel, Ruti, “Post-Communist Constitutionalism: A Transitional Perspective”, Columbia Human Rights Law Review 26 (1994): 167-90. . “Paradoxes in the Revolution of the Rule of Law”, Yale Journal of International Law 19 (1994): 239-47. . “Transitional Jurisprudence: The Role of Law in Political Transformation”, Yale Law Journal 106 (1997): 2009-80. . Transitional Justice (Oxford University Press: Oxford, 2000). Thayer, James Bradley, “The Origin and Scope of the American Doctrine of Constitutional Law”, Harvard Law Review 7 (1893): 129-56. Tribe, Laurence, “The Puzzling Persistence of Process-Based Constitutional Theories”, Yale Law Journal 89 (1980): 1063-80. . American Constitutional Law, 2nd edition (Foundation Press: Mineola, 1988). Troper, Michel, La théorie du droit, le droit, l’État (Presses Universitaires de France : Paris, 2001). Troper, Michel & Christophe Grzegorczyk, “Precedent in France”, in D. Neil MacCormick & Robert S. Summers, Interpreting Precedents: A Comparative Study (Dartmouth: Aldershot, 1997): 103-40. TrzciĔski, Janusz, ed., Sądy konstytucyjne w Europie, vol. 2 (Wydawnictwa Trybunaáu Konstytucyjnego: Warszawa,, 1997). . “Orzeczenia interpretacyjne Trybunalu Konstytucyjnego”, PaĔstwo i Prawo 57:1 (2002): 3-14. Tsilevich, Boris, “New Democracies in the Old World”, in Will Kymlicka & Magda Opalski, eds., Can Liberal Pluralism be Exported? Western Political Theory and Ethnic Relations in Eastern Europe (Oxford University Press: Oxford, 2001): 154-70. Tucker, Aviezer, “Paranoids May Be Persecuted: Post-Totalitarian Retroactive Justice”, Arch. European Sociology 40 (1999): 56-100. Tushnet, Mark, “Policy Distortion and Democratic Debilitation: Comparative Illumination of the Countermajoritarian Difficulty”, Michigan Law Review 94 (1995): 245-301. . Taking the Constitution Away from the Courts (Princeton University Press: Princeton, 1999). van Alstyne, William W., “A Critical Guide to Marbury v. Madison”, Duke Law Journal (1969): 1-47. van Dijk, P., & G.J.H. van Hoof, Theory and Practice of the European Convention on Human Rights, 3rd edition (Kluwer Law International: The Hague, 1998). Várady, Tibor , “On the Chances of Ethnocultural Justice in East Central Europe”, in Will Kymlicka & Magda Opalski, eds., Can Liberal Pluralism be Exported? Western Political Theory and Ethnic Relations in Eastern Europe (Oxford University Press: Oxford, 2001): 135-49.
BIBLIOGRAPHY
365
Vashkevich, Alexander, “The Republic of Belarus: The Road from Past to the Past”, in Andras Sajó, ed., Out of and Into Authoritarian Law (Kluwer Law International: The Hague, 2003): 265-98. von Beyme, Klaus, “The Genesis of Constitutional Review in Parliamentary Systems”, in Christine Landfried, ed., Constitutional Review and Legislation (Nomos: Baden-Baden, 1988): 21-38. . “Constitutional Engineering in Central and Eastern Europe”, in Stephen White, Judy Batt & Paul G. Lewis, Developments in Central and East European Politics, vol. 3 (Palgrave Macmillan: Basingstoke, 2003): 190-210. Waldron, Jeremy, “Rights and Majorities: Rousseau Revisited”, in John W. Chapman & Alan Wertheimer, eds., Majorities and Minorities: Nomos XXXII (New York University Press, New York, 1990): 44-75. . Liberal Rights (Cambridge University Press: Cambridge, 1993). . “A Rights-Based Critique of Constitutional Rights”, Oxford Journal of Legal Studies 13 (1993): 18-51. . “Precommitment and Disagreement”, in Larry Alexander, ed., Constitutionalism: Philosophical Foundations (Cambridge University Press: Cambridge, 1998): 271-99. . Law and Disagreement (Oxford University Press: Oxford, 1999). . “Eisgruber’s House of Lords”, University of San Francisco Law Review 37 (2002): 89-114. Walicki, Andrzej S., “Transitional Justice and the Political Struggles of Post-communist Poland” in A. James McAdams, ed., Transitional Justice in New Democracies (University of Notre Dame Press: Notre Dame, 1996): 185-238. Wasilewski, Andrzej, “Przedstawianie pytaĔ prawnych Trybunalowi Konstytucyjnemu przez sądy (art. 193 Konstytucji RP)”, PaĔstwo i Prawo vol. 54, no 8/1999: 25-39. Webber, Jeremy, “Institutional Dialogue between Courts and Legislatures in the Definition of Fundamental Rights: Lessons from Canada (and elsewhere)”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 61-99. Weber, Renate, “Constitutionalism as a Vehicle for Democratic Consolidation in Romania” in Jan Zielonka, ed., Democratic Consolidation in Eastern Europe, vol. 1: Institutional Engineering (Oxford University Press: Oxford, 2001): 212-42. . “The Romanian Constitutional Court: In Search of Its Own Identity”, in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002): 283-308. Welsh, Helga A., “Dealing with the Communist Past: Central and East European Experiences after 1990”, Europe-Asia Studies 48 (1996): 413-28. White, Stephen, Judy Batt & Paul G. Lewis, Developments in Central and East European Politics, vol. 3 (Palgrave Macmillan: Basingstoke, 2003). Whittington, Keith E., “The Political Foundations of Judicial Supremacy”, in Sotirios A. Barber & Robert P. George, eds., Constitutional Politics (Princeton University Press: Princeton, 2001): 261-97. .“Legislative Sanctions and the Strategic Environment of Judicial Review”, I.CON International Journal of Constitutional Law 1 (2003): 446-74. Winczorek, Piotr, “Axiological Foundations of the Constitution of Poland”, Saint Louis-Warsaw Transatlantic Law Journal (1997): 59-67. Wolczuk, Kataryna, “The Constitutional Court of Ukraine: The Politics of Survival” in Wojciech Sadurski, ed., Constitutional Justice, East and West (Kluwer Law International: The Hague, 2002: 327-48. WoleĔski, Jan, “Glosa do orzeczenia TK z 28 V 1997, K 26/96”, PaĔstwo i Prawo 53:1 (1998): 88-98. Women’s Rights Centre, “Women’s Rights in the Constitution of Poland”, Polish Legal Journal 6 (2001): 87-104. Zhelyazkova, Antonina, “The Bulgarian Ethnic Model”, East European Constitutional Review 10:4 (Fall 2001): 62-6. Zielonka, Jan, ed., Democratic Consolidation in Eastern Europe, vol. 1: Institutional Engineering (Oxford University Press: Oxford, 2001). ZieliĔski, Tadeusz, “Panel discussion”, in Konstytucja w sáuĪbie demokracji; Constitution in Service of Democracy, conference papers (The International Centre for Development of Democracy Foundation: Cracow, March 10-12, 1995). Zifcak, Spencer, “Hungary’s Remarkable, Radical, Constitutional Court”, Journal of Constitutional Law in Eastern and Central Europe 3 (1996): 1-56.
INDEX Abortion, 116, 123 in Hungary, 88, 132-134 in Poland, 52, 101-103, 134-135, 324 See also Right to life Absolute entrenchment, 80-81, 317 Abstract review, See Constitutional review Ackerman, Bruce, 34, 67, 232, 310, 311, 314, 317, 322, 345 Actio popularis, 6-7 Action sua sponte, 8-11, 25-26, 322 Acuña, Carlos H., 348 Advisory opinions, 13, 76-79, 316 Affirmative action, 201, 203-204, 311312, 321 Agh, Attila, 316 Albania, xvi, 135, 211, 226, 248 death penalty, 131 lustration, 240 Aleinikoff, T. Alexander, 276, 321, 351 Alexy, Robert, 268, 309 Allan, James, 318, 322 Altimari, Frank X., 88 Amendments, constitutional, See constitutional amendments Amnesty, 261 Amnesty International, 130, 131 Arabadjiev, Aleksandr, 184 Arato, Andrew, 56, 312-313 Argentine, 261 Assembly, freedom of, 152, 156-157 Association, freedom of, 157-158 Australia, 50, 67, 108, 174, 210 Austria, 1, 13, 17 Avanesyan, Suren, 304, 328 Badinter, Robert, 49, 312 Baird, Vanessa A., 302, 309 Balamazov, Stefan, 2 Balancing, 103, 276 Balkin, Jack, 290, 353 Baltic states, 248 lustration, 240-241 Russian minorities, 61, 206, 213, 216218 See also Estonia, Latvia, Lithuania Barak, Aharon, 296, 319, 354 Baric, Sanja, 307
Barry, Brian, 342 Bartole, Sergio, 317 Beatty, David M., 350 Belarus, 40 Constitutional Court, 4, 18 death penalty, 131 Belgium, 227, 260 Bell, John, 315 Beyme, Klaus von, xi, 292, 301, 353 Bicameral systems, 92-93 Bickel, Alexander, 70, 79, 315, 316 Bills of rights, 107, 111 Bosnia and Herzegovina, 178 Constitutional Court, 4, 14 Brewer-Carias, Allan-Randolph, 41, 311 Brilmayer, Lea, 312 Brunner, Georg, 304, 308, 316, 320, 324, 330, 331, 335, 339, 345, 347 Bugaric, Bojan, 333 Bulgaria, 2, 58 Constitutional Court, 3, 8, 12, 15, 17, 58, 61, 77, 87-88, 94, 128-129, 159-160, 183-184, 220-222, 248249, 303 death penalty, 130 Disclosure of Secret Police Files Act, 248-249 freedom of expression, 159-160, 167 incitement to hatred, 160 lustration, 53, 248-249 Media Council, 167 Movement of Rights and Freedom (MRF) case, 81, 220-222 official language, 285 Panev law, 248 Pension Act, 183-184, 193, 281 prohibition of discrimination, 204 right to life, 128-129 right to work, 281 self-defence in criminal law, 128-129, 146 social security, 281 Turkish minority, 81, 206, 220-222 Bundesverfassungsgericht, See Germany, Federal Constitutional Court Burdeau, Georges, 301, 317 Burger, Warren E., 110, 321
368 Calabresi, Guido, 79, 88, 316 Caldeira, Gregory A., 302, 309 Calhoun, Noel, 346 Canada, 50, 67, 83-84, 210 Charter of Rights and Freedoms, xii, 83-84, 295 Supreme Court, xii-xiv, 83 Capital punishment, See Death penalty Cepl, Vojtech, 226, 227, 239, 304, 306, 308, 316, 331, 344-345 Cerar, Miroslav, 302 Chantebout, Bernard, xiii, Cheli, Enzo, 321 Ciemniewski, Jerzy, 175, 333 Ciobanu, Viorel, 10 Citizens' duties, See Constitutional duties Citizenship rights in CEE, 151-153 Commercial speech, 168-169 "Commonwealth model" of constitutionalism, 295 Concrete review, See Constituional review Conscientious objection, 143-144 Conseil Consitutionnel, See France Constantinescu, Emil, 308 Constantinescu, Mihai, 10, 82, 305, 317 Constitution, binding interpretations of, 11-13 prohibition on amendment, 80-82 Constitutional amendments, 13, 80-82 Constitutional complaint, 6-8, 68 Constitutional continuity, 258-262 Constitutional Courts, as deliberative institutions, 31, 119121 and parliamentary minorities, 93-96 as part of the judiciary, 33-40, 121 as protectors of minority rights, 58-62 See also Constitutional review, Constitutional rights, Constitutional dialogue, 84-85 Constitutional duties, 284-287 Constitutional remand, 79 Constitutional review abstract, xi, xiii, 5, 13, 20, 25, 42, 4950, 65-74, 78 concrete, 19, 25, 44-45, 49-50, 65-74, 78, 303
INDEX decentralised, 67 ex-ante, 13, 74-79 ex-post, 13, 25, 74-79 final, 13, 25, 29-30, 79-85 and lower courts, 19-25 and paternalism, 47-49 and statutory interpretation, 98 before World War II, 1-2 Constitutional rights, xvii, 28 articulation of, 114, 118-119, 123-124, 294 disagreement on, 59, 118-124 and judicial review, 59, 109-118 negative and positive, 172 programmatic, 172, 182-183, 332 protection of, 107-125 restrictions of, 263-284 "Constitutionally conforming" interpretation, 98-99 Corrective justice, 232 Council of Europe, accession by CEE countries, 130 Parliamentary Assembly's resolution on lustration, 232-233 Countermajoritarian difficulty, 72 Croatia conscienctious objection, 145 law on health insurance, 189 minority representation, 219 pensions law, 191, 320, 337 right of assembly, 278 right to strike, 278 Cruz Villalon, Pedro, 308, 315 Czarnota, Adam, 348 Czech Republic, Act on Civilian Service, 277 Act on Extra-Judicial Rehabilitation, 89 Act on the Lawlessness of the Communist Regime and Resistance to It, 229-231, 239, 250-251 Charter of Fundamental Rights and Basic Freedoms, 166, 334 conscientious objection, 143-142, 277 constitutional complaint, 7 Constitutional Court, 13, 14, 22-23, 39, 77, 93, 98
INDEX detention of criminal suspects law, 98 election law, 154 hate speech, 160-161 law on public sector pay, 93 lustration, 235-240, 239, 258-259 movement and residence, freedom of, 149 petition rights, 156 privacy rights, 146 Senate, 92-93, 293 salary supplement law, 93 socio-economic rights, status of, 179 statutes of limitation, 250-251, 253 statutory interpretation, 98 Supreme Court, 22, 98 "war of the courts", 21-23 Czechoslovakia Constitutional Court, pre-War, 1-2 Constitutional Court of Czech and Slovak Federal Republic, 3, 237239, 281 lustration, 226-229, 236-239 "Velvet Revolution", 226, 242 Czeszejko-Sochacki, Zdzislaw, 245, 304, 309, 310 Dahl, Robert, 27, 50, 60-61, 292, 298, 309, 312, 314, 353, 354 Dayton Accord, 4 Death penalty, in CEE states, 97, 130-131, 323 in Hungary, 101 Decommunisation definition of, 224-225 in CEE states, 223-249, 258-259 in Czech Republic, 235-240 in Germany, 233-234 Deets, Stephen, 342-343 Dehaene, Jean-Luc, 332 Deliberation within constitutional courts, 31, 119-122 Defamation, 165-167 "Democratic debilitation", 290 Democratic legitimacy, See Legitimacy of constitutional courts Devins, Neil, 69, 84, 314, 318, 321-322 Di Gregorio, Angela, 308 Discrimination, See Equality Distributive justice, 184, 192-193
369 See also Social justice DomaĔska, Aldona, 338 Donati, Filippo, 321 Dorf, Michael C., 317 Dumitru, Horatiu, 83, 305, 318, 328, 350 Dupré, Catherine, 323 Duties, See Constitutional duties Dworkin, Ronald, 109-111, 117, 119-121 192, 204, 269, 270, 321-322, 337, 340 Dziaáocha, Kazimierz, 325 Ebzeev, Boris, 312 Education, right to, 153, 177, 189-190 Eisgruber, Christopher, 56, 310, 313, 323-324, 354 Election laws, 153-156 Elster, Jon, 227, 310, 317, 333, 335, 344346 Ely, John Hart, 56, 112-113, 310, 322 Epp, Charles R., 312 Epstein, Lee, 52-53, 313 Equality in CEE constitutions, 195-197, 280 gender, 199 sexual orientation, 201-203 Estonia, constitutional review in, 13, 67, 303 Constitutional Review Chamber, 5, 910, 14, 16, 61, 67, 75, 76, 89, 97, 149-150, 157, 216-218, 271-272, 303 election law, 216-218 freedom of association, 157-158 Language Act, 61, 216-218 law on housing-cooperatives, 89 law on sale of alcohol, 271-272 movement and residence, 149-150 Non-Profit Associations Act, 157-158 privacy rights, 146 proportionality review, 271-272 Russian minority, 206, 222 Taxation Act, 146 Ex-ante review, See constitutional review, ex ante European Convention of Human Rights, 267, 276 Article 8, 202-203, 267 Article 10, 159, 166, 267 Protocol 6, 130
370 European Court of Human Rights, 267, 295 European Social Charter, 172, 185, 332 European Union, 218 accession by CEE states, xviii, 42-43, 130 Charter of Fundamental Rights, xviii, 173, 263, 276, 332, 351 European Commission, 43 Exceptionalism, 296-298 Fabre, Cécile, 322, 332 Fallon, Richard H., 310 Favoreu, Louis, 43, 311 Federalism, xi Feinberg, Joel, 351 Ferejohn, John, 311 Fijalkowski, Agata, 323 Final review, See consitutional review Fiss, Owen, 39, 310, 311, 322 Fletcher, George, 302 France, 46, 47, 227, 257 Conseil constitutionnel, xiii, 13, 38, 48, 49, 61, 75-76, 90, 120, 315-316 "fundamental principles recognised by the laws of the Republic", 75 gender equality, 61 immigration law, 82 judicial review, 90 republican form of government, 81 Fredman, Sandra, 341 Freedom of expression, 123-124, 158-169 Freedom of movement and residence, 149-151, 152 Freedom of religion, 135-146 Free speech, See freedom of expression Fudge, Judy, 301 Gadzhiev, Gadis, 21 Ganev, Venelin, 222, 305, 319, 336, 338, 344 Gardbaum, Stephen, 295 Gardocki, Lech, 304, 309 Garlicki, Leszek, 15, 57, 175, 304, 305, 306, 307, 310, 315, 317, 325, 333, 335, 343, 354 Gender equality in CEE, 199-201 Geneva Convention of 1949, 253-254 Georgia, 178 Constitution, 134, 155, 184, 292
INDEX death penalty, 130, 323 German Democratic Republic, 233 Germany Basic Law, 80-81, 264, 317 constitutional review, 5, 13, 46 decommunisation, 232-233, 258 denazification, 232 Federal Constitutional Court, xiv, 30, 79, 268, 271 prohibition of consitutional amendment, 80-81, 317 Gibson, James, 30, 302, 309 Gicquel, Jean, 312 Gillis, Mark, 306, 319, 320 Ginsburg, Tom, 308, 315 Goncz, Arpad, 253 "Government of judges", xiii, 45, 46 Graczyk, Roman, 352-353 Granat, Miroslaw, 302 Gray, John, 354 Greece, 41, 311 Greenberg, Jack, 341 Griffin, Stephen, 60-61, 68, 91, 111, 311, 313, 314, 319, 321, 322 Grimm, Dieter, 350 Grudzinska-Gross, Irena, 317, 318, 331 Grzegorczyk, Christophe, 312 Guatemala Commission on Historical Clarification, 231 Gunther, Gerald, 350 Habermas, Jürgen, 33, 59, 69, 310, 313, 314 on concrete review, 68 Halmai, Gábor, 19, 302, 307, 308, 318, 320, 323, 330, 347, 348 Hamon, Francis, 301 Hate speech, 160-164 Hausner, Roman, 309 Havel, Vaclav, 14, 22, 44, 89, 165, 237 Health care, right to, 153 Henckaerts, Jean-Marie, 332 Hepple, Bob, 186 Hofmanski, Piotr, 348 Holländer, Pavel, 23, 98, 301, 309 Holmes, Stephen, 293 House of Lords, xv Human dignity, 100-101, 128
INDEX Hungary abortion 88, 132-134 Act on the Rights of Ethnic and National Minorities, 211 affirmative action, 201, 204 association, right of, 275 Bokros package, 148, 175, 291 Constitutional Court, xv, 6, 7, 8-9, 11-12, 13, 39, 42, 51, 56, 65, 67, 75-76, 175, 181, 201, 258-259, 297 death penalty, 101-102, 131-132 decree on radio and TV, 167 defamation of politicians, 166-167 desacration of national symbols, 163 freedom of expression, 161-163 election law, 155 freedom of media, 167 gender equality, 201 hate speech, 161-163 health care, right to, 176-171 human dignity, 100-101, 129-132 lustration law, 88, 242-243, 297 minority representation, 219 minority rights, 207-208, 211 pensions law, 181 presidential powers, 92 press law, 168 preventive review, 76-77 privacy rights, 147-148 proportionality review, 272-273, 275 public prosecutor's role, 129 right to reply, 168 right to work, 186 sexual orientation, 202, 275 socio-economic rights, status of, 181 statutes of limitation, 251-157, 260261 tax evasion, 148 trade unions law, 100 Zetenyi decision, 252-253, 255, 262, 290-291 Zetenyi-Takacs Act, 252, 257 Iliescu, Ion, 308 Imholz, Kethleen, 346 Impeachment, xvii, 13, 306 Incitement to hatred, See hate speech India, 50, 67 Individual rights, See constitutional rights
371 Initiation of constitutional review, 5-10 Institutional design, 32-33, 59, 110, 120 International Convention on the Elimination of all Forms of Racial Discrimination, 160 International Convention on Civil and Political Rights, 238 Interprétation conforme, See "Constitutionally conforming" interpretation India, 50, 67 "Invisible welfare", 193 Iorgovan, Antonie, 10 Ireland, 180 Italy, 82, 92 constitutional review, 5, 43, 46 Corte costituzionale, 16, 315 right to work, 185 Japan, 46, 67, 233 Jasudowicz, Tadeusz, 282, 352 Jevgrafov, Pavlo, 304 Jowitt, Ken, 199, 338 Juan Karlos, 223 Judges, appointment of, 14-18, 26, 122, 306, 307 tenure of, 14, 26, 32, 122, 306 Judicial activism, 68, 70, 75, 87, 96-104, 195 "Judicial overhang", 117, 322 Judicial restraint, 23, 46, 97-100, 245 Judicial review, See Constitutional review Jurisdictional disputes, 13 Justiciability, 70-73 Katrougalos, George S., 333 Kelsen, Hans, xvi, 88 Kelsenian model, 1, 2, 27, 41, 43, 44, 45, 63, 99, 311 Kessler, Zdenek, 319 Kirov, Stefan, 2 Klaus, Vaclav, 22 Knight, Jack, 52, 313 Kolarova, Rumyana, 88, 248, 318, 347 Komesar, Neil K., 310, 313, 322 Konstantinov, Emil, 344 Kopecky, Petr, 319 Kostov, Milcho, 184 Kovács, János Mátyas, 193, 337
372 KoĪuch, Maágorzata, 352 Król, Marcin, 292, 353 Krygier, Martin, 301 Kymlicka, Will, 213, 341-344 Labady, Tamas, 132 Lambert, Edouard, xiii, 301 Landes, William L., 316 Language rights, See Lingustic rights Latin America, 269 Latvia, 4, 64, 201, 213, 222, 223, 303 constitutional equality, 195 minority rights, 207 Legal certainty, 47-48 Legislative apathy, 117, 291-292, 294 Legislative override, 13-14, 29-30, 79-80, 82-40, 295 Legitimacy of constitutional courts, xixii, xiv, 26-63, 108, 111, 290, 298 "Lenient legislation", 73 Letki, Natalia, 345 Linguistic rights, 212-218 Limitations of rights, See Restrictions of rights Lithuania, Centre for Research into the Genocide of the People and Resistance, 241 choice of occupation, 188-189 citizenship, 152 Constitutional Court, 13, 39, 54, 168169 criticism of officials, 279-280 death penalty 131 election law, 218 freedom of association, 158 health insurance law, 191 Laws on the Bar, 188-189 Law on the Courts, 54 Law on Officials, 279-280 lustration, 240-241, 279 prohibition of discrimination, 204 religious freedom, 137-138 restrictions of rights, 278-279 Soviet security forces, employment of, 188, 240-241 Statute on Radio and TV, 159, 168 Lucinschi, Petru, 54 Ludwikowska, Anna M., 306, 308, 311, 316, 317, 319, 324, 344
INDEX Ludwikowski, Rett R., 334, 352 Lukashenka, Aleksandr, 4 Lustration, 97, definition of, 224-225 in Bulgaria, 53, 248-249 in CEE, 224-249, 258-259 in Czech Republic, 227, 258-259 in Hungary, 88, 242-243, 259 in Poland, 57, 243-247 symbolic function of, 228 Macedonia, election law, 155 inviolability of home, 146-147 Law on Civil Procedure, 214 minority rights, 207-208 official language, 214 religious freedom, 142-143 Macey, Jonathan R., 313, 318 Magalhães, Pedro, 319 Malova, Darina, 302 Marbury v. Madison, 68-70 Marshall, John, 68-70 Martinienko, Petro F., 305, 307 Massias, Jean-Pierre, 313, 328, 344 Masternak-Kubiak, Malgorzata, 304 Mavi, Victor, 334 Meþiar, Vladimir, 236 Medushevsky, Andrey N., 319 Menéndez, Agustín José, 332 Michnik, Adam, 225-227, 344 Mihai, Lucian, 10, 90, 304, 305, 317 Mik, Cezary, 352, 353 Mikva, Abner J., 316, 323 Military service, duty of, 285 Mink, Andras, 316, 320 Minorities, ethnic, in CEE, 205-207 representation in public authorities, 218-222 Minority protection, See Minority rights Minority rights, 207-222 group rights or individual rights? 209211 judicial review of, 49-50, 58-62, 112 See also: Linguistic rights Moldova, 54, 134, 178, 226 death penalty, 130 referedum law, 54
INDEX Moore, Michael S., 28-29, 73-74, 120, 309, 315, 322 Montenegro minority rights, 208 Moral realism, 28-31 Morawski, Lech, 312 Morshchakova, Tamara, 316 Mowbray, Alastair, 350 Mungiu-Pippidi, Alina, 319 Natural rights, 128, 238 "Negative legislator", xvi, 37, 88, 99 Nenovsky, Neno,17, 58, 302, 305, 306, 307, 313, 316, 317, 319, 320, 341 Netherlands, 107, 108 Neuborne, Burt, 38, 54, 310, 313 New York Convention of 1968, 253 New York Times v. Sullivan, 166 New Zealand, 174, 210, 295 Nickel, James W., 352 Nino, Carlos Santiago, 260-261, 348 Novak, Jiri, 89 Objectivism, moral, 28, 30-31 Offe, Claus, 335 Official language, 213-218 Omissions, legislative, 7, 9, 11 Oniszczuk, Jerzy, 313, 323, 336, 337, 343, 350 Osiatynski, Wiktor, 311, 313, 333, 340, 347, 352 Ossipov, Alexander, 342 Overlapping consensus, 38 Paczolay, Péter, 335, 340, 343 Panev Georgi, 248 Pap, András László, 341, 343 Paternalism, 47-49, 275-276 Parliamentary override, See Legislative override Parvulescu, Cristian, 305, 307-308, 317 Pasquino, Pasquale, 73, 311, 315 Pensions, 182-184, 191 Perry, Michael J., 295, 318, 354 Petition, rights of, 156 Petrov, Nikolai, 340 Pettai, Vello, 61, 216, 218, 314, 340 Pettit, Philip, 32, 310 Pithart, Petr, 306, 319, 354 Pizzorusso, Alessandro, 302
373 Poland abortion, 52, 101-103, 134-135, 324 association, freedom of, 157 Broadcast Law, 62, 113, 164-165, 322 Catholic Church, 62, 139, 164, 170 Charter of Rights and Freedoms of 1992, 180 citizenship rights, 151-152 collective agreements in the workplace, 98, 99 constitutional equality, 195 constitutional complaint, 6-7, 25 constitutional interpretation, 102-103 Constitutional Tribunal, xiv, 1, 3, 9, 12, 13, 14, 16, 23-25, 29-30, 39, 42, 43, 52, 56-57, 76, 79-80, 82, 92, 96, 98, 99, 101-102, 138-141, 170, 175, 182-183, 278, 312 death penalty, 130 education, rights to, 189-190, 192 equality before the law, 197-198 gender equality, 200-201 human dignity, 128-130 Institute of National Memory, 231 interpretation of statutes, 12 judges lustration law, 57, 244-245 land development law, 270-271, 277278 legislative override, 13, 29-30, 79-80, 82, 317 lustration, 243-247, 286, 313 medical certificates law, 130 National Council of the Judiciary, 57, 244-245 official language, 215-216 Ombudsman, 130, 134, 139 ownership rights, 270, 277-278 passports law, 129 pensions law, 191 proportionality review, 270-271 presidential powers, 91-92 protection of religious sensibilities, 164-165 Rechtstaat, principle of, 3, 102-103, 135, 244, 246, 262 religion in schools, 139-141 religious freedom, 138-142 right to work, 186-187
374 Senate, 92, 247, 313 socio-economic rights, 179-180, 182183 state and religion, 138-142 Supreme Administrative Court, 270 Supreme Court, 23-25, 43, 78, 200201 vested rights, 262 "war of the courts", 24 Political question, 70-72 Portugal, 21, 95-96, 311 Positive discrimination, See Affirmative action Posner, Eric, 255, 344 Posner, Richard, 173, 316 Post, Robert, 312 Presumption of constitutionality, 98, 99 Preuss, Ulrich, 174, 333, 335 PĜibáĖ, JiĜí, 230, 303, 308, 319, 345 Privacy, right to, 146-149 Procházka, Radoslav, 302 Proportionality review, 266-276 and necessity, 266-273 and relevance, 267-268 Prosser, William, 153, 327 Protection of religious sensibilities, 163165 Pushaw, Robert J., 316 Putin, Vladimir, 78, 95 Racial vilification, See Hate speech Rapaczynski, Andrzej, 333 Raskin, Jamin B., xii Ratner, L.G., 318 Rawls, John, 37, 119, 201, 322 Reason giving, by courts, 120-122 Rechtsstaat, xi, 298 in Poland, principle of, 102-103, 135, 244, 246, 262
in Slovenia, principle of, 185 Redish, Martin, 318 Rehnquist, William, 110, 321 Reitz, John C., 46-47, 312 Religion, See Freedom of religion Restrictions of rights, 263-284 constitutional design of, 263-266, 283284 and the essence of rights, 277-278
INDEX non-discrimination requirement of, 280-281 proportionality review of, 266-276 Retroactive justice, 223-224, 249-258 Right to life, 128-135 in Bulgaria, 128 in Poland, 128 See also, Abortion, Death penalty Right to work, 185-189 in CEE constitutions, 186 Rights, See Constitutional rights Rodin, Sinisa, 307 Romania Bucharest Tram Company case, 2 Constitutional Court, 3, 8, 10, 13, 18, 79-80, 82-83, 90 election law, 154, 156 equality before the law, 198-199 foreign travel tax, 150-151, 273-274 freedom of the press, 159 Hungarian minority 61-62, 213 incitement to hatred, 160 legislative override, 79-83 local administration law, 90 minority representation, 219 movement and residence, 150-151 pensions for war veterans law, 198199 petition rights, 156 political parties, 154 presidential powers, 92 proportionality review, 273-274 sexual orientation, 202-203 Rosenfeld, Michel, 210, 256, 342, 344, 348 Rozmaryn, Stefan, 2, 302 Rubio Llorente, Francisco, 317 Russia, xi constitutional amendment, 81 constitutional complaint, 7-8 Constitutional Court, 3, 8, 9, 15, 18, 20-21, 39, 51, 52-53, 95 movement and residence, freedom of, 151 propiska, 151 restrictions of rights, 264-265 rights of criminal defendants, 265 State Secrets Act, 263
INDEX Supreme Court, 20-21 Rymarz, Ferdynand, 245, 313 Sadurski, Wojciech, 302, 310, 325, 338 Safjan, Marek, 7, 30, 247, 304, 309, 317, 347 Sajó, Andras, 175, 192-193, 303, 330, 333, 337, 340, 341 Salzberger, Eli M., 306 Scalia, Antonin, 314 Scharpf, Fritz, 72, 315 Scheppele, Kim Lane, 132, 201, 291, 318, 324, 339, 347-348, 353 Schlink, Bernhard, xiv, 302 Schulhofer, Stephen, 253, 348 Schuster, Rudolf, 166 Schwartz, Herman, 172, 175, 301, 302, 305, 317, 332, 333, 347, 353 Separation of powers, xvii, 44-45, 55, 78 Separation of state and religion, See Freedom of religion Shapiro, Martin, xiv, 35-6, 51, 69, 90, 301, 310, 313, 314, 318 Shemshuchenko, Yurii S., 307 Shue, Henry, 332 Shvetsova, Olga, 52, 313 Siklova, Jirina, 226-227, 344 Skach, Cindy, 344 Sladecek, Vladimir, 309 Slovakia, conscientious objection, 144-145, 281 constitutional complaint, 7 Constitutional Court, 13, 38-39, 76, 166, 236 defamation of politicians, 166 Hungarian minority, 205, 213-214 Law on Weapons, 251 lustration, 236 minority representation, 219-220 official language, 213-214 petition rights, 156 prohibition of discrimination, 204 privacy rights, 148 restrictions of rights, 281 socio-economic rights, status of, 179 Slovenia, 185 choice of occupation, 187-188 Constitutional Court, 9, 103-104, 185, 199
375 election law, 155 equality before the law, 199 local government law, 88-89 minority representation, 220 minority rights, 208-209 notaries Act, 188 presidential powers, 92 privacy, 147 proportionality review, 199, 266-267 Rechtsstaat, 185 reprivatisation law, 103-104, 199 war veteran benefits, 185 Smulovitz, Catalina, 348 Social security, 176-177, 179, 181-185 Social justice, 197 See also Distributive justice Socio-economic rights, 171-194 and citizenship, 152-153 constitutional catalogues of, 176-178 inclusion in constitutions, 171-176 Sokolewicz, Wojciech, 302, 315 Sólyom, László, 6, 34, 51, 56, 100-101, 131-132, 161-162, 166, 167, 181, 290, 304, 310, 312, 314, 316, 319, 320, 324, 327, 330, 331, 335, 336, 339, 345, 351, 353-354 Sonobe, Itsuo, 312 South Africa, 261 Commssion for Truth and Reconciliation, 231 Spain, 223, 261 Constitution, 180 constitutional review, 5, 21, 41, 46 Sprague, Leigh, 308, 319 Standing, rules of, 50 in US consitutional review, 70-71, 312 stare decisis, 44-45 "State based on law", See Rechtsstaat Statism, 46-47 Statutes of limitation, retroactive extensions of, 249-258, 260-261 Staugaitiene, Teodora, 331 Stone Sweet, Alec, 45, 61, 75, 82, 87, 9192, 94, 311, 312, 314-319, 354 Strashun, Boris, A., 304, 308, 311, 319 Strict scrutiny, 57, 104 Sullivan, Kathleen, 282, 352
376 Sunstein, Cass, 176, 293, 313, 322, 333, 353-354 Sweden, 67, 76, 109 Law Council, 74 Switzerland, 311 Szabo, Andras, 161-162 Szczerbiak, Aleks, 346 Taney, Roger B., 110, 321 Teitel, Ruti, 39, 40, 44, 226, 256-257, 290-291, 293, 296, 310, 344, 346, 348, 353-354 Thayer, James Bradley 100, 320 "Third chamber" model, 14, 37, 39, 75 Todorov, Todor, 94, 167, 184, 302, 304, 305, 306, 307, 316, 320, 331, 336 Toleration regimes, 222 Transitional justice, 223-224, 258-262 Triadic model of courts, 35-36 Tribe, Laurence H., 314, 332, 352 Troper, Michel, 301, 312, 317 TrzciĔski, Janusz, 39, 302, 304, 309, 310, 353, 343 Tsilevich, Boris, 342 Tucker, Aviezer, 344-348, 348 Turkey, 212 Tushnet, Mark, xii, 116-117, 290, 318, 322, 336, 353 Tychyj, Wolodimir, 303, 307 Uitz, Renata, 324, 330 Ukraine, 47 Constitutional Court, 4, 8, 15, 17, 50, 65, 94, 135, 184, 226, 303 election law, 155 Unconstitutional conditions, 281-282 United Kingdom, 109, 111, 295 Human Rights Act, 295 United States constitutional review, xii-xiii, 41-42, 45, 47-48, 50, 100, 109-111, 276, 283 First Amendment to the Constitution, 109, 161, 163 hate speech, 163 judicial review, See constitutional review minority rights, 209-210 Supreme Court, xii, 37-38, 44, 50, 5556, 60, 66-73, 75, 84, 91, 109-111, 120, 276, 291-292, 312
INDEX unconstitutional conditions, 281 welfare rights, 173-174 USSR, 184, 212, 222, 249, 266, 271 Van Alstyne, William, 313, 314 Van der Jeught, Stefaan, 332 Van Dijk, P., 350 Van Noof, G. J. H., 350 Varady, Tibor, 211, 342 Vashkevich, Alexander, 303 Vermeule, Adrian, 255-256, 344, 348 Veto points, 91, 94-95 Voigt, Stefan, 306 Vote, right to, 152-155 Wald, Patricia M., 353 Waldron, Jeremy, 31, 60, 109-113, 123, 310, 313, 314, 322-323 Walicki, Andrzej, 347 WaáĊsa, Lech, 44, 180, 244 Warren, Earl, 110, 292, 321 "Wars of the courts", 21-24, 26 Washington, George, 316 Wasilewski, Andrzej, 311 Webber, Jeremy, 83, 318 Weber, Renate, 12, 302, 305, 319 Welfare rights, See Socio-economic rights Wellington, Harry H., 79, 316 Welsh, Helga A., 346 Whittington, Keith, 291, 294, 318, 353354 Winczorek, Piotr, 332 Wolczuk, Kataryna, 303, 305, 314 WoleĔski, Jan, 102, 321, 324 Yaroslavtsev, Vladimir, 18, 78, 305, 308, 316, 317, 320 Yeltsin, Boris, 18, 52, 305, 308 Yugoslavia Constitutional Court, xiv, 1, 3 Yuzhkov, Leonid, 4 Zdyb, Marian, 245, 313, 247 Zhelev, Zhelyu, 94 Zhelyazkova, Antonina, 340, 344 ZieliĔska, Eleonora, 338 ZieliĔski, Tadeusz, 171, 183, 332, 336 Zielonka, Jan, 319, 337, 338 Zifcak, Spencer, 313, 348, 353 Zirk-Sadowski, Marek, 312 Zlinszky, Janos, 132
INDEX Zoll, Andrzej, 12, 92, 317, 318 Zorkin, Valery D., 3
377