LIST OF CONTRIBUTORS Richard Belfield
Industrial Relations Department and Centre for Economic Performance, London School of Economics, UK
John T. Delaney
Katz School of Business, University of Pittsburgh, USA
Hristos Doucouliagos
School of Accounting, Economics and Finance, Faculty of Business and Law, Deakin University, Australia
Tony Dundon
Centre for Innovation and Structural Change, National University of Ireland Galway (NUIG), Ireland
Jack Fiorito
Department of Management, College of Business, Florida State University, USA
Brian Harney
Judge Business School, University of Cambridge, UK & Centre for Innovation and Structural Change, National University of Ireland Galway (NUIG), Ireland
Elizabeth A. Hoffmann
Department of Sociology and Anthropology, Purdue University, USA
Paul Jarley
Center for a Sustainable Aluminum Industry, Gatton College of Business, University of Kentucky, USA
Dae Yong Jeong
Department of Sociology, University of Illinois at Urbana-Champaign, USA
Bruce E. Kaufman
Department of Economics and Beebe Institute, Georgia State University, USA
Patrice Laroche
Institut Universitaire de Technologie, Universite´ Nancy, France vii
viii
LIST OF CONTRIBUTORS
John Lawler
Institute of Labor & Industrial Relations, University of Illinois at Urbana-Champaign, USA
David Lewin
Anderson School of Management, University of California at Los Angeles (UCLA), USA
Jianwei Li
Research Division II, National Institute for South China Sea Studies, Hainan, China
John Logan
Industrial Relations Department, London School of Economics, UK
David Marsden
Industrial Relations Department and Centre for Economic Performance, London School of Economics, UK
David Metcalf
Centre for Economic Performance, London School of Economics, UK
Chris Riddell
School of Policy Studies, Queens University, Canada
Roland Zullo
Institute of Labor and Industrial Relations, University of Michigan, USA
INTRODUCTION Volume 15 of Advances in Industrial and Labor Relations (AILR) contains 10 papers, four of which deal with human resource management and six with unionization. Six of the papers were originally presented in ‘‘Best Papers’’ sessions at the 57th and 58th annual meetings of the Labor and Employment Relations Association (LERA). In keeping with AILR’s global perspective and global sourcing of leading research, the studies contained in these papers draw on data from the United Kingdom, France, Asia, Canada, and the United States. The first paper, by David Marsden and Richard Belfield, presents findings from a study of performance pay for public school teachers in Great Britain, where new national performance pay and performance management initiatives were introduced in 2000.1 Data for this study were obtained from three waves of surveys conducted among samples of primary- and secondary-school classroom teachers and head teachers in England and Wales, with the first survey conducted just prior to the introduction of the new performance pay initiatives and the second and third surveys conducted during the first and third years, respectively, following introduction of these initiatives. Marsden and Belfield find that as with other public sector workers, teachers are not particularly motivated by the financial incentive aspects of the new pay system, but are motivated by the goal-setting and appraisal aspects of this system. As a result, say the authors, the new system has resulted in closer alignment among teacher, school, and national policy objectives, with goal-setting in particular being positively related to increasing pupil (student) academic performance. These findings are of interest well beyond Great Britain, perhaps especially in the United States, which has also recently initiated new national policies aimed at improving public school teacher and student performance. The second paper, by Elizabeth A. Hoffman, analyzes workplace dispute resolution in two types of organizations, namely, cooperative and hierarchical (see footnote 1). For this purpose, Hoffman draws on field data in the form of 128 open-ended employee interviews conducted during a three-year period at pairs of works sites in three industries, namely, coal mining, taxicab driving, and organic food distribution, variously located in Wisconsin (US), London, Wales, and Halifax (UK). These interviews as well as the ix
x
INTRODUCTION
author’s site observations were tape-recorded and transcribed, and the data were then analyzed using NUD*IST Vivo (or NVivo) software. Hoffman finds that in the worker cooperatives, employees typically use both informal and formal mechanisms and channels to raise and resolve a wide variety of workplace issues. In contrast, employees in the hierarchical organizations, each of which had a formal grievance procedure in place, tended to pursue grievance resolution either formally in the case of coal miners or informally in the cases of taxicab drivers and organic food distribution, but not both ways. More compelling, perhaps, Hoffman also finds that compared with employees in cooperative work sites, employees in hierarchical work sites were far more likely to cope with felt grievances by tolerating them rather than pursuing their resolution formally or informally – a phenomenon that other researchers have dubbed ‘‘suffering in silence’’ (Boroff & Lewin, 1997; Lewin & Boroff, 1996; Lewin & Peterson, 1999). Hence, Hoffman concludes that members of worker cooperatives possess and use more dispute resolution strategies than their conventionally employed counterparts. The third paper, by Hristos Doucouliagos and Patrice Laroche, uses extensive survey data to analyze the impact of human resource (HR) practices and unionization on the organizational efficiency of French enterprises. Eschewing a one-dimensional productivity measure, these authors operationalize organizational efficiency in two dimensions, namely, technical efficiency or actual outputs in relation to what is achievable, and scale efficiency or the scale of factors employed in the operations of a firm. For their empirical work, the authors drew HR data from a nationally representative sample of 2,381 French establishments with more than 20 employees, added establishment financial performance data from other sources, and then used Data Envelopment Analysis (DEA) to calculate technical and scale efficiency for each establishment. Unionization is essentially treated as a control variable operationalized in a dichotomous union present (1)–union absent (0) format. On the basis of their Tobit regression estimates, Doucouliagos and Laroche conclude that when introduced on their own, high involvement type HR practices have negative effects on technical efficiency but positive effects on scale efficiency. When introduced into unionized workplaces, by contrast, these HR practices seem to have positive effects on technical efficiency and negative effects on scale efficiency (though this finding varies notably by type/index of HR practices). While they conclude that ‘‘On balance, HR practices appear to have a positive effect on productivity in French industry,’’ the main contribution of this paper is to show how sensitive this conclusion is to the particular empirical specification employed – whether using data from France or other nations.
Introduction
xi
The fourth paper, by Brian Harney and Tony Dundon, uses data drawn from 18 case studies, conducted during 2002–2003, of Irish small and medium size enterprises (SMEs) to analyze the factors shaping human resource management (HRM) practices in these enterprises.2 Applying a 20-point HRM profile to their case studies but focusing in particular on HR planning, recruitment and selection, training and development, and performance appraisal and reward mechanisms, the authors document the wide variety of practices in each of these areas that prevails among Irish SMEs. Harney and Dundon then show how these practices are shaped by various external factors, such as product market structure, industry and local labor supply, as well as by internal factors such as ownership structure, managerial style, and unionization. One factor that did not appear to shape these SMEs’ HR practices is size of enterprise, yet this ‘‘non-significant’’ factor plays a key role in this paper. In particular, Harney and Dundon frame their study with and use their empirical findings to further support a holistic, open systems view of HRM, which they argue runs strongly counter to received HRM theory, especially HRM ‘‘best practices.’’ Stated differently, the authors believe that HRM research, especially research on HRM and business performance, has a strong large firm bias that makes it unsuitable for analyzing and understanding the formation and use of HRM policies and practices in SMEs. This is a provocative hypothesis and one that readers of AILR can assess for themselves in light of the evidence that Harney and Dundon bring to bear upon it. The fifth paper, by Dae Yong Jeong and John Lawler, is an ambitious effort to develop a new theory of enterprise unionism based on a study of union structure in nine Asian nations (see footnote 1). Using a comparativehistorical research approach, the authors focus heavily on the interplay among national political forces and union structure in each of these nations. Three main types of union structure are identified in this research: an enterprise union type structure that is highly stable and dominant in some nations, such as Japan, Thailand, and Malaysia, and less stable and frequently changing in other nations, such as Korea and The Philippines; a non-enterprise or laissez faire union structure of the type found in Indonesia and Hong Kong; and a middle-ground or transitional union structure of the type that characterizes Taiwan and Singapore. In essence, Jeong and Lawler treat union structure as the main dependent variable and show how that structure has been differentially shaped by exogenous factors in each nation, in particular, national political policy and power. For this purpose, the authors use ‘‘critical junctures’’ analysis in which the initial period of collective bargaining as shaped by state labor policy is key, but is then followed
xii
INTRODUCTION
by periods of transition and sometimes disjuncture in which state labor policy changes markedly, with consequent effects on union structure and union-management relations. Jeong and Lawler not only interpret their comparative-historical evidence to support their theory of enterprise unionism, but in the process also reject other theoretical explanations, including national culture theory and internal labor market theory. The sixth paper, by David Metcalf and Jianwei Li, is interesting enough on its own but is especially so when considered in conjunction with the Jeong and Lawler paper summarized above. In essence, Metcalf and Li contend that Chinese trade unions, which collectively have more members than all of the world’s other trade unions combined, are basically powerless when it comes to representing their members in negotiations with employers. To support this forceful claim, the authors rely on data drawn from detailed case studies conducted during 2004–2005 in Hainan Province. On the basis of this research, Metcalf and Li judge Chinese trade unions to have very little independent bargaining power, as reflected in their lack of impact on enterprise wage inequality and workplace safety. Further, only about half of Chinese trade union members are actually represented in collective bargaining, and efforts by trade union representatives to broach let alone resolve worker grievances are brushed off or otherwise rejected by government and/employer officials (including officials of US-based companies operating in China). Indeed, the Chinese Ministry of Labour often competes with and dominates the All Chinese Federation of Trade Unions (ACFTU) when it comes to labor relations and dispute resolution involving Chinese workers. And, say the authors, such union impotence has not fundamentally changed despite the recent enactment of new laws that, on the surface, promote collective bargaining and tripartite workplace dispute resolution. Whether further development of competitive product and labor markets in China will alter the role of unions in such a way as to increase their power and influence is a question worth pondering, but on the basis of Metcalf and Li’s analysis this likelihood seems small indeed. The seventh paper, by Chris Riddell, develops several hypotheses regarding union raiding and tests these hypotheses using data drawn from British Columbia, Canada during the last quarter of the 20th century (see footnote 1). A raid is defined as a certification attempt involving a union applying for bargaining rights in a unit in which another union holds certification and, as the author notes, such raiding has attracted only a modicum of scholarly attention. In developing the conceptual framework for this study, Riddell analogizes union raids to corporate hostile takeovers, positions company management as favoring incumbent unions, and argues that it is unclear, a
Introduction
xiii
priori, whether a labor federation’s (specifically, the Canadian Labour Congress’) no-raiding policy actually reduces union raiding. Using a unique data set compiled by the British Columbia Labour Relations Board (BCLRB), Riddell analyzes 606 union raids involving approximately 57,500 workers that occurred between 1978 and 1998. On the basis of Probit regression estimates, he finds that raiding is much more likely to be successful when an existing union has underachieved in collective bargaining, that employers do favor incumbent unions and are able to influence worker voting in this regard, and that independent unions are about as likely as larger national and international unions to carry out successful raids. Taken together, these findings suggest that union raiding is an efficient mechanism in facilitating competition for union services – a finding that has important implications for unionization in other Canadian provinces as well as in other nations. The eighth paper, by John Logan, provides a comprehensive historical analysis – the first of its kind – of the US Labor-Management Reporting and Disclosure Act (LMRDA) requirements covering employers and consultants. The LMRDA was enacted in 1959 to deal with the then well publicized problem of union racketeering and corruption, and much enforcement and much research on enforcement of the LMRDA as it pertains to unions have occurred since then.3 But the LMRDA also contains provisions that require financial disclosure from employers and consultants who persuade employees to reject unionization, and as Logan shows these provisions have been only meagerly enforced. While such lack of enforcement may have mattered little during the first two decades of the LMRDA’s existence, it has mattered quite a lot since then in light of greatly expanded employer use of consultants to thwart employee unionization. Logan’s analysis shows that not only the US Congress, which of course enacted the LMRDA, but also the US Department of Labor and the courts have through their administrative rulings and judicial decisions, respectively, permitted aggressive consultant campaigns to flourish and failed to expose such activities to public view despite the ostensible intent of the law. Further, and although enforcement of these particular LMRDA provisions has waxed and waned somewhat on the basis of whether a Democrat or Republican administration held sway, the secular trend is away from rather than toward financial disclosure by employers and ‘‘labor’’ consultants. Hence, the story told by Logan in this paper adds to our understanding of a key legal and political factor in the continuing decline of US unionism. The ninth paper, by Roland Zullo, provides a contextual and theoretical analysis of the formation in September 2005 of Change to Win, a new USbased labor federation and rival to the AFL-CIO. For this purpose, Zullo
xiv
INTRODUCTION
traces the rise and evolution of decentralized, economically oriented business unionism in the US, emphasizing how the legal development of industrial relations strongly contributed to this union strategy. He then identifies four key characteristics of business unionism: organizing workers capable of securing negotiated agreements, voluntarism or neutral political affiliation, limiting internal democracy in favor of organizational efficiency, and promoting ‘‘American style’’ capitalism. The core of Zullo’s analysis, framed by political theory, is that the internalization – globalization – of market capitalism has exposed the vulnerabilities and contradictions of these underlying characteristics of business unionism which, in turn, has pressured the AFL-CIO and especially certain of its affiliates to consider a new strategy, namely, the socialization of labor–capital conflict or social movement unionism. While Change to Win was ostensibly created for this purpose, Zullo shows that even in its very early stage this new labor federation is nevertheless adopting certain business unionism-type tactics (analogous to the old CIO in its early days), and also that the AFL-CIO itself is adopting certain elements of social movement unionism. Hence, concludes Zullo, even if Change to Win is successful in changing certain elements of US labor law and in achieving broader-based unionism, it too will ‘‘default’’ to business unionism. For readers of AILR, perhaps the key issue embedded in Zullo’s research is whether this new inter-labor federation competition will spur a resurgence of US unionism akin to the period in which the CIO competed with the AFL. The tenth paper, by Jack Fiorito, Paul Jarley, and John T. Delaney, makes for a suitable companion to the Zullo paper by using extensive quantitative analysis of survey data to illuminate the specific choices facing Change to Win as well as the AFL-CIO in their efforts to reverse the declining fortunes of the US labor movement (see footnote 1). Ironically, as these authors point out, union leader and member dissatisfaction with the AFL-CIO led in 1995 to the election of insurgent president John Sweeney, who undertook financial reallocation and organizational decentralization initiatives aimed at de-emphasizing services to members in favor of enhanced union organizing activity – initiatives that, following Zullo, appear well within the tradition of business unionism. These initiatives met with little success, however, so that the present challenge of revitalizing the US labor movement seems at least as daunting as when Sweeney took over the reigns of the AFL-CIO. Fiorito, Jarley, and Delaney’s analysis of 1990 National Union Survey (NUS) and 1997 Survey of Union Information Technology (SUIT) data suggests that organizing success requires a few larger unions willing to innovate with new organizing methods and that operate with relatively
Introduction
xv
decentralized organizational, decision making, and governance structures. Stated differently, Fiorito, Jarley, and Delaney’s regression findings imply that both Change to Win and the AFL-CIO would do well to embrace the contemporary market capitalism notion of getting closer to the customer – in this instance, the customer who may potentially buy union membership and associated union services. If this constitutes further movement toward rather than away from business unionism, one may say ‘‘so be it.’’
NOTES 1. Winner of and originally presented at the Advances in Industrial and Labor Relations/Labor and Employment Relations Association Best Papers session, 58th Annual Meeting of the Labor and Employment Relations Association, Boston, MA, January 8, 2006. 2. Winner of and originally presented at the Advances in Industrial and Labor Relations/Labor and Employment Relations Association Best Papers session, 57th Annual Meeting of the Labor and Employment Relations Association, Philadelphia, PA, January 9, 2006. 3. Indeed, under the administration of President George W. Bush, 48 new fulltime staff have been added to the U.S. Department of Labor specifically for the purpose of conducting financial audits of labor unions. This has occurred during a period in which U.S. unionization continues to decline and in which there is no evidence of a rise in union racketeering or corruption.
REFERENCES Boroff, K. E., & Lewin, D. (1997). Loyalty, voice and intent to exit a union firm: A conceptual and empirical analysis. Industrial and Labor Relations Review, 51(October), 50–63. Lewin, D., & Boroff, K. E. (1996). The role of loyalty in exit and voice: A conceptual and empirical analysis. Advances in Industrial and Labor Relations, 7, 69–96. Lewin, D., & Peterson, R. B. (1999). Behavioral outcomes of grievance activity. Industrial Relations, 38(October), 554–576.
David Lewin and Bruce E. Kaufman
PAY FOR PERFORMANCE WHERE OUTPUT IS HARD TO MEASURE: THE CASE OF PERFORMANCE PAY FOR SCHOOL TEACHERS David Marsden and Richard Belfield ABSTRACT The introduction of performance-related pay with performance management in the state school sector of England and Wales represents a considerable change in the school management system. After 2000, all teachers were subject to annual goal setting performance reviews. Experienced teachers were offered an extended pay scale based on performance instead of seniority, and to gain access to the new upper pay scale, teachers had to go through a ‘threshold assessment’ based on their professional skills and performance. This paper reports the results of a panel survey of classroom and head teachers which started in 2000 just before implementation of the new system, and then after one and after four years of operation. We find that both classroom and head teacher views have changed considerably over time, from initial general scepticism and opposition towards a more positive view, especially among head teachers by 2004. We argue that the adoption of an integrative bargaining approach to performance reviews explains why a growing minority of schools have achieved improved goal setting and improved pupil attainments as they have implemented performance management. Pay for performance has Advances in Industrial and Labor Relations, Volume 15, 1–34 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15010-6
1
2
DAVID MARSDEN AND RICHARD BELFIELD
been one of the measures of organisational support that head teachers could bring to induce changes in teachers’ classroom priorities. We argue that the teachers’ case shows that a wider range of performance incentives than previously thought can be offered to employees in such occupations, provided that goal setting and performance measurement are approached as a form of negotiation instead of top-down.
1. INTRODUCTION In theory, school teachers should be among the least suitable groups for linking pay to performance. Nevertheless, despite strong initial opposition among teachers and their unions, and equally strong scepticism from academic economists (e.g. Richardson, 1999; Dolton, McIntosh, & Chevalier, 2003), the pay for performance system for school teachers in England and Wales, introduced in 2000, appears to be gaining acceptance among teachers. In schools where it is practiced systematically, there is evidence that it is leading to improved goal setting, and faster improvements in pupil performance than in other schools. This article seeks to explain how this came about. In doing so, it argues that we need to broaden current theories linking pay to performance, and consider the way performance is defined, through goal setting and the potential element of individual negotiation in this process. One of the key limitations of existing theory is that it has focused on a narrow range of types of output-based pay, such as for semi-skilled workers, sales staff and senior managers, where performance is more easily measured. In his Journal of Economic Literature review of current research on the subject, Prendergast (1999) urged us to look beyond such groups for a wider understanding of pay and incentives for other kinds of employees. In contrast to the former occupations, and like many other public service workers, school teachers, of whom there are nearly half a million in the UK, have proved a challenge for motivation by conventional forms of pay for performance. The nature of their work is ‘imprecise’. It does not comprise a set of well-defined techniques that have to be consistently applied, and there is, at best, a loose relationship between particular actions and their students’ learning (Murnane & Cohen, 1986). In this respect, their work could be said to differ fundamentally from that which has provided the strongest demonstrations of how output-based pay can boost performance, such as Lazear’s windshield replacers at Safelite (Lazear, 1996). Not only is their work ‘imprecise’, but frequently it involves a range of different kinds of activities,
Pay for Performance Where Output is Hard to Measure
3
some of which are more amenable to measurement than others, for example, students’ national test results versus educating future citizens. Their work often involves a great degree of interdependency. Even if one can isolate an individual teacher’s classroom contribution, which many teachers believe improbable, there remain important areas of team-work within schools (Dolton et al., 2003). Moreover, the presence of a ‘collegiate ethos’ (Adnett, 2003) diminishes the likelihood that teachers will ‘shirk’ because their effort is hard to measure, and so reduces the relevance of one common argument for tying pay to performance. Thus if we were to stick to the conventional forms of output-based pay, which as Prendergast notes, have been analysed extensively within Personnel Economics, we should expect to find few successful examples of pay for performance in schools. This was indeed the case until fairly recently. Despite a proliferation of schemes in the US in the early twentieth century (Murnane & Cohen, 1986), and a long-running scheme in England in the latter part of the nineteenth century (Dolton et al., 2003), pay for performance virtually disappeared from school teaching in both countries for most of the twentieth century. Teaching in schools was barren-ground for any kind of payment by results, even in private schools (Murnane & Cohen, 1986). Theory and practice seemed to agree, and the affair was closed. Nevertheless, some characteristics of teachers’ work have meant that the issue never quite disappeared, and in recent years, it has undergone a revival in the US in a number of school districts with performance pay schemes, in Israel (Lavy, 1999) and in Japan.1 It has done so most dramatically in Britain, where a national scheme, ‘Performance Management’, was introduced in 2000 in state primary and secondary schools in England and Wales, the subject of this paper. The reasons for the sustained interest are easy to understand. It is widely accepted that there are considerable variations in teacher effectiveness, a point touched on in the British government’s recent teacher recruitment campaign ‘everyone remembers a good teacher’,2 and widely acknowledged by both classroom and head teachers in the first wave of the survey on which this article is based (Marsden, 2000). Governments have also retained an interest for budgetary reasons. Faced with a need to raise educational standards, and tight restrictions on public spending increases, several national governments have sought to use pay systems more actively as an instrument to achieve these ends. Notwithstanding the academic arguments against performance pay, many educational practitioners have observed the extent to which existing pay rules were manipulated in order to recruit and retain, and the perverse effects of paying teachers for management activities
4
DAVID MARSDEN AND RICHARD BELFIELD
in order to enhance their pay. It was widely felt too that the then existing salary system was failing to motivate teachers because so many (about 60%) were stuck on the top pay grade, and was encouraging them to take on additional non-teaching duties for more pay, which diverted them from the classroom, and so made raising standards even harder to achieve. There was also a feeling in British government circles that even though teachers work very long hours during term time, and experience a strong professional ethos, there were few financial inducements to encourage them to accept the changes in work priorities sought by the government.
2. THE TEACHERS’ PERFORMANCE PAY AND PERFORMANCE MANAGEMENT SYSTEM The performance management system for state school teachers, introduced by the government in 2000, sought to address the need to reward teachers better for excellence in the classroom. Performance management (PM) sought to combine goal setting and appraisal with performance pay, thus extending the growing practice of other parts of the British public services. It made annual salary progression dependent on performance. The new pay scale comprised two parts. For the lower part, which was, roughly speaking, the old main pay scale, the annual seniority increments were retained, covering roughly the first few years of a teacher’s career. Progression to the new ‘upper pay scale’ required passing the ‘threshold assessment’, based partly on teachers’ professional development, and partly on the progress of their pupils. Progression along the upper scale was based entirely on performance. There was also provision for accelerated increments for high performers along the lower part of the scale. The associated financial rewards could be considerable, potentially taking teachers’ earnings to 25% above the top of the lower scale, where about 60% of teachers had been clustered in 2000. Unlike some private sector bonus systems, the teachers’ performance-related increments are permanent and count for their pensions. In exchange for the new structure of rewards, all teachers were required to have an annual performance review to set goals and appraise performance, including those on the lower part of the new pay scale. The most controversial element of the new performance system has been the concept that performance should include an element of pupil progress. On the whole, teachers, and their unions, have been receptive to rewards for improved inputs, such as for improved skills. However, pupil progress relates to their outputs, and this caused widespread opposition, with the
Pay for Performance Where Output is Hard to Measure
5
largest of the teachers’ unions opposing it on the grounds that it amounted to payment-by-results. The other unions, which were more supportive of performance review, were also strongly opposed to any form of paymentby-results, and were deeply suspicious that in some schools, head teachers would implement this part of the new scheme in a mechanistic fashion. Although the Education Department provided many ‘good practice’ illustrations of how pupil progress could be applied in a constructive fashion, there was widespread suspicion among teachers at the outset that the scheme was really about using performance pay in order to save money and restrict pay increases to those who toed the line. As will be seen, our study shows that this perception declined over time, but has nevertheless remained a significant under-current among teachers. To assist implementation, on its website, the Education Department provided schools with numerous examples of good practice concerning how pupil progress could be integrated into performance reviews. Many of these illustrate a problem-solving approach to goal setting and appraisal. One example, which is discussed later in this article, shows the objective setting part of the review involved a joint analysis of strategies to tackle a student learning problem, how to improve attainments of boys in a particular subject, and agreement to implement them during the year. The next review in the cycle would then assess how successful the implementation had been, and the contribution of these strategies to the school’s own goals. Despite such positive examples, many teachers, and their unions, feared that practice in schools would be rather different. It was feared that pressure on head teachers to raise their school’s academic performance would lead to payment-by-exam-results. One such pressure, which we explored, was that of educational ‘league tables’. As a result of changes in funding arrangements and the devolution of a greater management autonomy, schools now have to compete to attract pupils who bring funding. Schools which fail to do so, will experience falling numbers and income, and may eventually be closed. A crucial signal to parents in this quasi-market is provided by comparative tables, league tables, displaying each school’s national exam results. These are publicly available on the Education Department’s website, and are frequently published in local newspapers, so that parents know which schools are getting more pupils through their exams, and can benchmark them against local and national standards. A second aspect of diversity among schools in the implementation of performance management was suggested to us in an interview with the former head of Cambridge Educational Associates,3 the organisation responsible for monitoring implementation of the threshold. From early on, it
6
DAVID MARSDEN AND RICHARD BELFIELD
seemed that some schools had approached performance management as a means of improving how schools are run, to achieve better coordination between teachers’ activity in the classroom and the school’s wider objectives. In contrast, initially at least, the majority of schools were using the new scheme as a means of getting what was felt to be a long overdue pay increase for teachers, and which should ease staff retention problems, especially in high living cost areas such as London. Later on, we characterize these two strategies as ‘reformer’ and ‘firefighter’. This diversity of strategies will make it possible later on to test the effects of the scheme on pupil-performance outcomes.
3. THE ANALYTICAL APPROACH: PERFORMANCE MANAGEMENT AS ‘INTEGRATIVE’ NEGOTIATION The examples of good practice performance reviews provided by the Education Department include a special feature that is under-developed in the discussions of agency, expectancy, and goal-setting theory: namely a view of goal setting as a problem-solving process. The problem is how to align individual employees’ work goals with the changing goals of their organisation. In professional work, management is dependent on the knowledge and expertise of their staff in order to define appropriate performance goals, and especially to identify the steps necessary for their achievement. There needs therefore to be an exchange of information both about objectives to ensure these are realistic, and about the means to reach them. The nature of this asymmetry is illustrated from a study by the government’s school inspectorate, which quoted one teacher from a school with a good appraisal system which included an element of classroom observation: My classroom observation (by the head) was useless. For one thing he only came once, and there was no proper feedback. The head was completely out of touch with recent developments in modern language teaching, and so unable to engage in the type of debate I hoped for. (Ofsted, 1996, y15)
The appraiser in this case lacked knowledge of both the subject and its teaching methods, and would be hard put to ‘cascade’ objectives top-down, and hope that they would be successfully implemented. There may also be disagreements about the goals that constitute performance. To what extent should the teacher prioritise exam passes versus developing an understanding of, and an affection for, the language being taught, and there might need to be compromise from both sides in order to sustain commitment to agreed goals.
Pay for Performance Where Output is Hard to Measure
7
One framework for thinking about such processes is provided by what Walton and McKersie (1965) describe as ‘integrative’ bargaining or negotiation. However, in this case, it would take place at an individual rather than at a collective level, between individual teachers and their line managers. According to these authors, we should distinguish ‘integrative’ from ‘distributive’ bargaining, the latter being about shares of the pie, and the former about problem-solving in order to achieve a larger pie. They present four stages or components of integrative bargaining: identifying the problem; searching for alternative solutions; selecting the best alternative; and commitment to implementation. Whereas the first two may seem relatively technical, the third, involves also reference to the preferences of the two parties, or as the two authors stress, to their utility functions. The final component brings us back to issues of re-contracting and commitment to deliver by both parties. Thus, if performance management is to be seen as involving a significant element of integrative bargaining, then one would expect it to show evidence of these components, which will be explored in the empirical part of the paper. However, before that, it is helpful to compare the integrative bargaining approach with other theoretical approaches to the analysis of incentives and performance appraisal. Several elements of the process of integrative bargaining are recognised individually by the other main theories in the area, but they are not treated together. In discussing the optimal design of incentives, agency theory deals extensively with issues of monitoring performance, and gaining agreement to a mutually satisfactory design of incentives, otherwise the job offer will be rejected. Mostly, the emphasis has been on new hires rather than on negotiation with incumbent employees (Prendergast, 1999; Tzioumis, 2005). Agency theory has also looked at issues of renegotiation, particularly in relation to pay when firms are faced with changes in market conditions, and it has dealt with the problems of ‘hold-ups’ in such relationships, whereby one party threatens to end the relationship in order to force the other to make large concessions (Gibbons, 1998; Malcomson, 1997; Teulings & Hartog, 1998). However, their emphasis has been very much upon distributive rather than integrative bargaining, on changing rewards rather than adapting elements of performance. The psychological theory of goal setting, as recently reviewed by Locke and Latham (2002), has recognised that participative goal setting, with an input from employees, often leads to better results, although the emphasis, as Locke and Latham (2002) observe, has been primarily on information exchange, and to a lesser extent on goal commitment. Expectancy theory emphasises the social exchange between managers and staff, and the need for perceived good faith from management in the operation of performance
8
DAVID MARSDEN AND RICHARD BELFIELD
evaluations on which performance pay depends, but on the whole, management is left to determine the nature of the nature of performance and its link with rewards (Furnham, 1997). A third strand of psychological theories does consider problems of contracting and breach of perceived agreement, namely, ‘psychological contract’ theory (Rousseau, 1995). For sound methodological reasons, Rousseau rejects the idea that organisations can be party to psychological contracts: organisations are not psychological entities. An integrative negotiation approach draws on elements of these theories by offering a stronger focus on adapting the content and priorities of work performance to changing needs of the principal, in the light of the expertise and the preferences of the agent, and on the need to advance by mutual consent on the basis of give and take by both sides. The next section of the paper presents the empirical results. The panel results document the change from initial hostility to performance management among classroom teachers, and also many head teachers, towards declining scepticism about its value among teachers and an increasing belief in its usefulness among school heads. It then looks at performance management and head teacher views on its contribution towards improved school management, and argues, on the basis of survey replies, that it functions in many schools as form of integrative negotiation between school management and classroom teachers. The article then seeks to provide rough estimates of the proportion of schools in the sample which could be said to practise integrative negotiation systematically in performance management, and to show that in these schools, pupil attainments in national tests have increased more than in other schools.
4. RESULTS 4.1. The Nature of the Survey The primary data source for this study is a panel questionnaire survey of classroom and head teachers in the same schools, conducted at intervals between 2000 and 2004 by the Centre for Economic Performance, and based on a random sample of state schools in England and Wales.4 It has been carried out in close consultation with the teachers’ unions, professional organisations and the Education Department. Nevertheless, it is an independent study. The panel has now completed three waves: the first in February–March 2000, before the new scheme was implemented (see Marsden, 2000); the second in May–June 2001, after the first round of ‘threshold’ results were known; and
Pay for Performance Where Output is Hard to Measure
9
the third in May–June 2004, after the results of the fourth round of the system, including potential progression to point 3 on the new upper pay scale. For classroom teachers, the questionnaire probes their attitudes and experiences in relation to performance management, and seeks to measure aspects of their work patterns and priorities alongside other variables such as their degree of organisational commitment and assessment of the working atmosphere in their schools. A copy of the questionnaire used in the first wave can be found in Marsden (2000). For head teachers, it asks also about the operation of performance management in their schools, and whether they believe it has assisted them in their management duties. In addition to the ‘before and after’ element of the panel, it is possible to link replies from classroom teachers with those of their head teachers, and with other information on their schools concerning the impact of performance management on work patterns, and some educational outcomes based on Education Department’s school performance data. Initially, the panel included replies from about 4,000 teachers and 1,000 heads. Accounting for sample attrition, it is possible to link replies from about 1,000 teachers and 300 heads over time through the panel. We conducted a number of checks to see whether the panel results reported here differ statistically significantly from the simple cross-section results for each wave, and found that on the whole they do not. In combining all these different types of information, this study goes much further than the previous UK public service studies, such as those reviewed in the government’s Makinson (2000) Report, by tracking the same individuals over time, and by combining multiple points of observation and types of data. These studies, which led Makinson to conclude that the schemes practised in the British public services were not working, gave no indication as to how employee attitudes changed over time, and lacked independent data on performance outcomes, weaknesses the present study has been designed to overcome. 4.2. From Initial Hostility to Increasing Acceptance of Performance Management: 2000–2004 As one would expect from the earlier discussion, at the start, teachers were more sceptical of the principle of linking pay to performance and its feasibility in their area of work than other public service employees (see Marsden & French, 1998). This is reflected in the first wave responses of both classroom and head teachers in 2000 (see Table 1 below). Initially only a fifth of classroom teachers and a third of heads agreed with the principle of linking teachers’ pay to performance, with comparably low percentages of both groups, thinking it could lead to a fairer allocation of pay, or that it was fair to link
Classroom Teachers’ Judgements of Performance Management: Before and After Implementation (Balanced Panel Results).
Abbreviated Questions
Appropriateness of PM to teachers’ work Hard to relate the work done in schools to individual performance Teachers have too little autonomy Managers will reward favourites PM causes jealousies Improved goal setting by PM Targets set more clearly Made me more aware of my school’s targets (Secondary schools only) More incentive to focus on pupil attainments Used as a means to make staff better informed about objectives within the school
Head Teachers
Wave 1 % Agree
Wave 2 % Agree
Wave 3 % Agree
Change
Wave 1 % Agree
Wave 2 % Agree
Wave 3 % Agree
Change
23
29
37
34
37
50
6
13
19
28
45
63
22
22
37
25
45
57
89
83
75
83
69
44
– 54 88
– 43 60
– 19 37
–
36 – 81
30 – 50
9 – 29
– –
42 32
60 37
52 –
84 –
82 –
– –
22 –
35 –
– 40
– 38
– 45
– NS
–
16
31
–
33
41
NS
NS
–
+
–
DAVID MARSDEN AND RICHARD BELFIELD
Acceptance of the basic principles of PM The principle of relating teachers’ pay to performance is a good one PM will lead to a fairer allocation of pay Fair to reward for pupil progress
Class Teachers
10
Table 1.
Organisational support through PM Used to help better identify teachers’ professional development needs (Secondary schools only) Incentives offered by PM Salary levels>Threshold are too low to make me want to work harder (% agree) Salary levels>Threshold are too low to make me want to work harder ( % disagree) PM simply a device to get more work done (Secondary schools only) There is a budgetary quota on Threshold and UPS pay increases PM helped to ensure teachers in your school get their pay increase (Secondary schools only)
–
15
37
–
44
70
–
25
43
–
54
63
NS
–
–
–
–
61
74
NS
33
35
33
–
–
–
–
15
25
28
–
–
–
–
58
58
49
–
–
–
–
66 82
65 28
56 79
– –
– –
– –
– –
–
81
76
NS
–
72
55
NS
–
79
82
NS
–
64
46
Pay for Performance Where Output is Hard to Measure
(Secondary schools only)
Number of balanced panel observations: class teachers: c 2300; head teachers: c 100 for most questions. Note: Results are based on the balanced panel, that is those replying in each of the three waves. The results of the balanced panel mostly differ only slightly from those of the three cross-sections. Scores in the table are weighted by sample fractions (smplwt). Change over time significant at: NS, not statistically significant; NA, not applicable. o1%; o2%; o5%; +o10%.
11
12
DAVID MARSDEN AND RICHARD BELFIELD
pay to pupil progress. Yet, over time, there have been significant increases in support for these views, roughly doubling over the duration of the panel. It should be emphasised that these are balanced panel results, tracking individuals who replied to all three waves, and so represent changes of opinion by individual teachers. By 2004, between half and two-thirds of heads, and between one and two-fifths of classroom teachers had come to accept these principles. There have also been equally significant changes in teachers’ beliefs about the appropriateness of performance management to teachers’ work. Initially, 90% of classroom and 80% of head teachers considered that PM was inappropriate because one could not relate the work done in schools to the performance of individual teachers, a point echoed by the economist critics of the scheme (e.g., Richardson, 1999; Dolton et al., 2003). By 2004, this had dropped to 75% among classroom teachers, and to only 44% among school heads. Likewise at the outset, over a third of heads thought that teachers had too little autonomy in their jobs to be able to vary their performance, but this had fallen to under 10% by 2004. Among classroom teachers, fears that managers would use PM to reward their favourites, a deep-rooted problem in other British public sector performance pay schemes, had fallen away from over half in 2000 to less than a fifth by 2004. Thus, even though many teachers maintained their opposition to PM, substantial numbers had changed their assessment of it by 2004. Closely related to these changes in teachers’ attitudes appears to be the growing perception that goal setting has been working better than initially expected, as 60% of classroom teachers and 80% of heads report that management now sets targets more clearly, and among secondary schools in particular, 40% of classroom teachers and 70% of heads say that PM is now used as a means to make staff better informed about objectives within the school. By 2004, nearly 40% of teachers reported that PM had made them personally more aware of their school’s objectives, the increase being particularly strong in secondary schools. Another significant change has been the increased perception that PM provides organisational support to teachers, helping to identify their professional needs, up from 25% to over 40% between 2001 and 2004 among class teachers, and up to over 60% among head teachers by 2004. Among head teachers, there has also been modest growth to 45% in 2004 of those who believe that PM has encouraged teachers to focus on pupil attainments. Nevertheless, the salience of pay has remained strong, and it lends some support to the position of the largest teachers’ union, the National Union of Teachers, that PM is primarily about paying teachers for results, especially in the sense that it represents what its members believe to be the underlying
Pay for Performance Where Output is Hard to Measure
13
reality of the scheme. In the first wave, two-thirds of classroom teachers thought the scheme was ‘simply a device to avoid giving a pay rise to all teachers’ (not shown in the table), and even after four years of operation, about half of them still believed the scheme was basically about getting more work done. Despite the subsequent evidence of high pass rates, a very high percentage of teachers, and a substantial number of heads, doubt the government’s commitment to continued funding of performance increases, and fear that they will become very selective if constrained by school budgets. Initially, and again in 2004, around 80% of classroom teachers believed school budgets would impose a quota on performance pay increases, and even after four years of operation, just under half of all head teachers believed the same. Thus, although the unions have pressed hard to ensure that an absolute standard of performance is applied, most teachers believe that budgetary considerations force PM to focus on relative performance, and restrict rewards to the best performers. Apart from who gets incentive pay, there is also the question of how far teachers value the size of rewards offered. Although there seems to be a stable group of about one-third who believe the new salary levels are too small to make them want to work harder, there has been a growing minority who disagree, and find the rewards attractive: up from 15% in 2000 to nearly 30% in 2004. Thus, although other surveys have shown that teachers are not strongly motivated by financial incentives, pay being more a source of discontent than of positive motivation (Varlaam, Nuttall, & Walker, 1992), there is clearly a group of teachers within the sample who have found the new pay scales attractive. An important category among these appears to be new entrant teachers who value the long-term prospect of earnings on the extended pay scale (Marsden, 2000). Finally, a key question for the interpretation of PM in schools is how far teachers believe it is operated primarily as a means to ensure teachers get their pay increase. This perception was much stronger among classroom than among head teachers, with around four-fifths of classroom teachers holding this view in 2004, but only half of head teachers doing so. This suggests that classroom teachers are keenly aware of the link between increased pay and their performance reviews. Although success rates at both the threshold assessment and upper pay scale progression were extremely high between 2000 and 2004, at over 90%, these were known ex post. Ex ante, the results show both a strong consciousness that pay increases require success in the review process, and continuing uncertainty as to whether pass rates will, in practice, be determined by absolute criteria, comparable to a ‘driving test’ as one teachers’ union official described it, or whether they will be constrained by a quota or budgetary restrictions. A good
14
DAVID MARSDEN AND RICHARD BELFIELD
measure of the uncertainty in teachers’ minds about the selective nature of the progression linked to performance is shown by their responses between waves to the question about whether performance awards are subject to a quota. The second wave took place in spring 2001, just after teachers had been informed of the results of the first round of threshold assessments. There was a sharp fall, from over 80% to less than 30%, in the percentage thinking that restrictions would be applied to the number of successful applications. However, that percentage bounced back to 80% in the third wave. The reason for this, suggested to us when presenting preliminary results to the teachers’ unions, was that the Education Department had floated proposals shortly before the third wave that the money available would be capped, so that future increases would be constrained by school budgets. On that occasion the proposals were withdrawn, but the continued uncertainty has meant that teachers believe the scheme has real teeth. Indeed, there was a long-running tension between the government, whose policy is that the scheme should be selective, and the unions, which have sought to minimise selectivity.5 The head teacher judgements are noteworthy for the decline in numbers who think of PM primarily as a means to get pay increases for their teachers. This was especially true in secondary schools which have more developed management systems than in primary schools. This would be consistent with an increased emphasis on using PM as a means of improving goal setting within schools noted earlier on, and is consistent with the point to be developed later on, that the number of schools using PM to reform school management has been growing. Thus, the first impression from these results is that the predicted failure of performance reviews and performance pay in schools has not occurred, and that by 2004, there was every sign that it was taking root in schools, providing more motivating rewards for a significant number of teachers. Most important, the procedures for defining the kinds of performance that management wanted to achieve were proving moderately effective: goal setting was believed to have improved, teachers felt management was providing some support to them as a result of PM, and there was an increased awareness of the priority to be given to pupil attainments. 4.3. Performance Review in Schools as ‘Integrative Negotiation’ Clues as to the reasons for the moderate success of PM by 2004 can be found by looking more closely at the process of performance review that lies at the heart of PM. This section explores evidence that, in the best cases, performance reviews have functioned as a form of integrative negotiation of
Pay for Performance Where Output is Hard to Measure
15
teachers’ work objectives, and notably in the ‘reformer’ schools. It does so by considering in turn each of the four stages of integrative bargaining identified by Walton and McKersie. 4.3.1. Identifying the Problem In the first stage of integrative negotiation, the two parties have to identify the problem which brings together the priorities of the organisation and the capacities and orientations of its members. In the case of performance review, this means relating the contribution of individual teachers to the wider objectives of the school and their adaptation to new circumstances. From the management side, in recent years, schools have come under a large number of external pressures which have made their comparative performance with other schools a key concern. This information is easily available to parents of potential pupils, and with the funding changes that attach school funding to pupils, schools compete in a ‘quasi market’ (Glennerster, 2002). The Education Department also benchmarks school performance, and there is also richer and more qualitative information available to all online from government inspectors (Ofsted, 1996). Within schools, this pressure has become symbolised by the educational league tables of local schools based on comparative academic performance at national tests. To capture this element of pressure on head teachers, we enquired about the influence of league tables on their schools in the third wave. In that, 60% of heads reported increased parental pressure as a result of educational league tables. We also enquired about the methods used to produce better academic performance in their schools, and notably, whether schools sought to learn about educational practices used at comparable schools scoring strongly in school league tables. Fifty percent of teachers, and sixty percent of heads, replied that their schools did so. Heads were also asked whether they had shifted resources in response to league table performance issues: 60% reported greater academic content in courses covered by tests; 41% that significant resources were devoted to teaching test skills; and 30% reported more resources being provided to test subjects, such as maths, at the expense of other subjects. Thus, the mixture of benchmarking and quasi market had generated pressure on heads to change priorities and shift resources in their schools. Most schools use a range of decision-making methods to address such issues, but the performance review has a special place because it provides an opportunity for individual-level discussions about individual work objectives and how to relate them to school objectives. The results of our survey indicate that classroom teachers enter this process with different priorities, which have also been taken up by their
16
DAVID MARSDEN AND RICHARD BELFIELD
unions. For example, when questioned about the reasons for the way they allocated their discretionary working time between different activities, many teachers emphasised their desire to give a ‘high quality of education’ as the most important reason, particularly for such activities as lesson preparation and feedback. The same can also be found in some of the written-in comments. As one classroom teacher put it: the new system is based on a ‘narrow perception of education as measured by attainment, whereas schools are also about learning to be members of a community’.6 Indeed, some heads expressed similar views in their own written-in comments, but it was their job to make the new system work. Thus performance management can be seen as involving a process of problem identification from two sides, which are not necessarily in opposition to one another, but they do have distinct points of view that need to be reconciled. The effectiveness of this process is revealed by the growth in the share of both head teachers and team leaders7 who believed that PM had made teachers more aware of school objectives, both around 60% by wave 3, and had made teachers think more systematically about their work priorities, both around 50% by wave 3 (Table 2). Many also thought that it had increased the importance of good middle-management in their schools. Table 2. Contribution of Performance Management to Improved Goal Setting in Schools: Head Teachers’ and Team Leaders’ Views (Balanced Panel).
Performance Management has: Made more teachers aware of the school’s objectives in the School Improvement Plan Made teachers think more systematically about their work priorities Increased the importance of good middle management.
Wave
Head Teachers Agree %
Team Leaders Agree %
2
41
47
3
57
65
2
39
37
3
54
50
2
55
45
3
60
42
Note: Sample numbers in the balanced panel are 104, and in the cross section about 420 and 290 respectively in waves 2 and 3. Statistical significance of change between waves 2 and 3: 1%; 5%. Results weighted by sample fractions.
Pay for Performance Where Output is Hard to Measure
17
Given the difficulty for head teachers to be acquainted with the detail of their colleagues’ work, except in small schools, and their inevitable reliance on this intermediate management level, this suggests that in many cases PM has raised awareness of the need to strengthen the linkage between different levels of management within schools. 4.3.2. Searching for Alternative Solutions The second stage of integrative negotiation involves identifying strategies for solution, and how to implement them, including support given to teachers by their schools. The survey did not probe directly the search for alternative solutions, but when considered with other sources, there is indirect evidence that this occurs in the best cases. A first source is the good practice case study material provided by the Education Department at the outset, which drew on existing practice in some schools, and sought to show others how PM could be used (see Box 1 below). The spirit of these is to suggest
Box 1. Objective Relating to Pupil Performance: Class Teacher (Year 5, approximately ages 9–10). Background: Evaluation of optional Year 4 test data and the Year 4 teacher’s own assessments shows a clear gender gap in attainment in English and a weakness in writing across the class. In the judgement of the Year 4 teacher, by the end of the year, about 60% of children in the class were not able to do all that the Literacy Framework expects. In his view, about 70% of boys fell into this category as opposed to 50% of girls. The objective: By the end of Year 5, to increase substantially the percentage of the class as a whole that will be able to do almost all of what the Literacy Framework states that they should be taught over the year, in writing as well as reading, and to reduce the gender gap from the present 20%. In discussion: The teacher and team leader might discuss the importance of remedying identified weaknesses in children’s knowledge of spelling conventions, sentence construction and punctuation, and teaching strategies known to be helpful for boys such as clear shortterm targets or use of non-fiction writing exercises. The review meeting would discuss the outcomes achieved by the children and review comparative achievement by boys and girls. Source: DfEE (1999).
18
DAVID MARSDEN AND RICHARD BELFIELD
open enquiry in the search for solutions rather than a strict top-down approach. The Education Department’s guidance stated: Objectives would typically emerge from a discussion between team leaders and individual teachers about the priorities for the coming year and the particular ways in which the teacher can help the pupils he or she teaches. This might involve targeting the progress of a group of pupils which is not meeting expectations or a small number of named pupils. Or it might be a priority to implement new policies or develop approaches or techniques which will help pupils to progress e.g. better classroom management skills. DfEE (1999, p. 10)
Thus, having identified a problem, the two parties work together to devise a strategy to tackle it. The same spirit was illustrated in a number of case studies included in the same document, and which fleshed out these general principles. Other examples provided by the Education Department include the provision of support to teachers to enable the objectives to be achieved, most notably, further professional development to provide additional skills where needed, but they also included other types of support, as identified in the objective setting process. An important part of the search for alternative solutions involves consideration of the means required for their implementation. Management controls many of these resources, so it is to the head teachers’ replies that one has to turn for evidence of this (Table 3). The questions that comprise the table reflect the factors identified by heads in the first wave as potential causes of variations in the effectiveness of experienced teachers in their schools, and which therefore, are likely to be addressed in the development of new strategies for the school. Thus, we asked heads whether PM had Table 3.
Performance Management as a Means of Supporting Teachers: Head Teacher Views, Balanced Panel.
PM has helped the school assist those teachers: With difficult or inappropriate workloads Whose professional development needs are greatest Whose morale was low Who had difficulty motivating their pupils
Wave 2
Wave 3
% agree
% agree
11 37 18 12
35 36 23 21
Weighted by sample fractions. Number of observations: balanced panel 103, and in the cross section, wave 2 c. 415, and wave 3, c. 290. Change between wave 2 and wave 3: statistical significance: 1% level; 10% level.
Pay for Performance Where Output is Hard to Measure
19
enabled them to address some of these problem factors. By wave 3, benefits were identified in a substantial minority of cases, between 20 and 35%, the most important being assistance in identifying teachers’ professional development needs, and workload problems. There was a notable increase between the second and third waves for some issues. By 2004, 35% of heads in the balanced panel reported that PM had helped them identify and assist teachers whose workload problems might have inhibited their performance. 4.3.3. Selecting the Best Alternative An important part of integrative negotiation lies in the involvement of both parties in selecting the best alternative: otherwise it is hardly a negotiation. Teachers were asked about the conduct of their own performance reviews. A first test is whether teachers themselves thought the agreed priorities reflected the wider objectives of their schools, as in certain key management documents, such as the School Development Plan (SDP) (Table 4). They might for example have disagreed with the question because they thought Table 4. Teachers’ Reports on the Nature of Their Most Recent Performance Review. (Balanced Panel).
Reference of Individual Goals to School Goals Did they relate to the wider objectives of the school (as in SDP or team plans)? Did they include indicators of pupil progress? Teacher influence on goal setting Did you have the opportunity to discuss them with your team leader? Could you influence the objectives chosen? Did they take account of your professional needs? Evidence of commitment to agreed goals Were your objectives clear and measurable? Do you understand how they will be monitored and reviewed? Are you in a position to achieve your objectives? Did you agree your objectives with your team leader?
Wave 2
Wave 3
% Agree
% Agree
91
89
NA
81
96
91+
92 79
89 78
94 81
88+ 87
93 97
87 NA
Not every question was asked in both waves to minimise the demands placed on respondents. Significance of change wave 2 to wave 3: 5%; +10%.
20
DAVID MARSDEN AND RICHARD BELFIELD
the head had imposed some personal objectives unrelated to those of the school, or because the teachers were ignorant of the school’s objectives. In either case, one would conclude that the resulting solution was not seen by both parties as the ‘best alternative’. Although some of the written-in comments by teachers mentioned objectives being imposed on them in their performance reviews, the answers to other questions in Table 4 also confirm that around 90% of teachers felt fully involved in the process: they had the opportunity to discuss objectives, to influence those chosen, and account was taken of their professional needs. Confirming the ‘golden thread’ that the government established in the scheme, in the great majority of cases, the performance reviews included indicators of pupil progress. On this evidence at least, it would seem that teachers were involved in the selection of the best alternative strategies to achieve their objectives, and thus participated in the third stage of integrative negotiation. 4.3.4. Commitment to Implementation The final stage of any integrative negotiation concerns commitment to implementation. A necessary part of this is that objectives should be clear and measurable, that there should be verifiable criteria of fulfilment, and an understanding of the monitoring procedures. Without these, the process can easily become an empty exercise. The overwhelming majority of teachers (c. 90%) reported that they had clear and measurable objectives, and a slightly smaller percentage said that they understood how progress would be monitored and reviewed, and that they were in a position to achieve their objectives (Table 4). They also considered their objectives were realistic, and nearly all said they had agreed them. As with choosing objectives, so with monitoring, the Education Department gave considerable guidance as to how statistical and other evidence could be used to help define and monitor achievement of objectives. Considering these replies at face value, there is a strong case for concluding that integrative negotiation is an important component of performance management as it is practised in many schools. Inspection of the written-in comments underlines this, but it also raises some questions about the significance of the very high level of positive reports. On the positive side, one head teacher wrote: ‘It has helped in the move towards being a selfevaluating school and to establish the need for more challenging objectives for staff. The previous system was ‘‘cosy’’. In similar vein, another head said: ‘[PM] has made them more aware of the need for objective evidence to prove pupil progress’. On the negative side, some classroom teachers expressed adverse comments, of which, one wrote: ‘PM has resulted in
Pay for Performance Where Output is Hard to Measure
21
teachers doing more admin, tasks analysis, statistical work and less time given to creating interesting vibrant lessons’. One explanation of such divergent written-in views, apart from their relatively small number compared with the main replies, is that head teachers had a somewhat different perception of performance management than their classroom teachers. For the former, PM involved use of data and other materials to diagnose problems faced by their schools. For the latter, in schools where the message had not been well communicated, statistical data were irrelevant to the task of ‘creating interesting vibrant lessons’. One measure of whether PM was mere form-filling and agreeing a paper exercise lies in the extent to which both head and classroom teachers share the view that PM has led to improved goal setting in their schools. Where classroom teachers experience PM as a lot of meaningless statistics they are unlikely to have been inolved in the problem-solving dimension of goal setting. To this we now turn.
5. INSTRUMENTALITY OF PM IN LINKING CLASSROOM TEACHER AND SCHOOL-WIDE OBJECTIVES Integrative negotiation requires two parties, and so it is important to demonstrate that the schools in which this was best developed were also those in which teachers had greater awareness of school objectives. This requires a linking of replies from classroom and head teachers. Ideally, we should compute average scores for teacher awareness of objectives in a particular school, and then compare these with the judgements of the head teacher on the quality of the different stages of integrative negotiation within the performance review process. Unfortunately, the response rate across schools was too variable, and so we tackle the question the other way round: whether the teachers who said PM had made them personally more aware of their school’s objectives were working in schools in which the heads reported that PM was effective? Although there appears to have been a considerable improvement in the operation of PM between waves 2 and 3, we limit this analysis to wave 2 because it contained a wider range of descriptive questions to head teachers about the conduct of performance reviews. To gauge the quality of performance reviews as judged by the head teachers, we conducted a factor analysis which boiled 18 questions down to three dimensions: whether PM had led to clearer goals being set; whether objectives were agreed to by teachers in their schools; and whether the
22
DAVID MARSDEN AND RICHARD BELFIELD
school provided support, such as for professional development. These three indexes were then included in an ordered probit regression on classroom teacher replies as to whether their performance review had raised their awareness of school objectives. The latter variable included a five point scale from disagree strongly to agree strongly. We also included some descriptive questions from the teacher questionnaire relating to aspects of their performance reviews, plus a set of control variables. The results are shown in Table 5. Although the pseudo R2 is low, it is statistically significant at the Table 5.
Factors Raising the Awareness of Class Teachers of Their School’s Objectives. Coef.
Head teacher policies on PM in their school Clearer goals 0.1346927 Objectives agreed 0.1482757 Provision of support 0.0278599 Teacher views of their latest Performance Review Specific objectives set in PR 0.3596218 Clear objectives set in PR 0.1137584 Objectives related to SDP 0.3119143 Objectives take account of 0.1723101 professional needs School provides a mentor for 0.1068846 applications Teacher eligible for threshold 0.2677386 School uses PM to inform about 0.4737709 objectives School used PM to assist profession 0.3853758 development
Significance
Std. Err.
Variable type
0.0547 0.0388 0.0489
Factor score Factor score Factor score
0.1379 0.1214 0.0941 0.0829
Dummy Dummy Dummy Dummy
0.0764
Dummy
0.1178 0.1286
Dummy Dummy
0.0998
Dummy
Dependent Variable: ‘PM has made me more aware of the targets set in the school development plan.’ (Ordered Probit, Values from 1, Disagree Strongly to 5, Agree Strongly.) Control variables on teacher and school characteristics included dummies if female, aged over 30, part-time, had a degee, union member, secondary school, and number of school pupils. Of these, only ‘secondary school’ was significant, at 0.1%, with a coefficient of 0.3. Pseudo R2 ¼ 0.045, p ¼ 0.000 Wald w2 ¼ 158.7, n ¼ 1699. Data analysed using STATA 9. Factor analysis of head teacher views was based on an analysis of 18 variables on different aspects of PM in their schools. These scores have a mean of zero and a standard deviation of unity. The variables were adjusted by setting missing values to neutral values to ensure the regression was based on full coverage of the response for wave 2. Total teacher response in wave 2: 1792, used replies in regression: 1699. Significance: o1%; 2%; 5%.Robust standard errors used.
Pay for Performance Where Output is Hard to Measure
23
0.1% level. The important conclusion from the table is that when head teachers judge performance management to provide clearer goals and, those goals are agreed, classroom teachers are more likely to report that it has increased their awareness of their school’s objectives. Likewise, classroom teachers are more likely to find PM increases their goal awareness when the performance review is carried out systematically, when there are specific and clear goals, their own needs are taken into consideration, and when they see their school using PM to inform and to support its teachers. A slightly different angle on the same question is provided by examining the degree to which head teachers, the school leadership group and classroom teachers all share the view that PM has caused goal setting to become clearer and more effective in their schools. Again the motivation for this analysis is that a consistent set of views is much more likely if goal setting is being run systematically in a school, and in a spirit of integrative negotiation. If the spirit is all ‘take’, or that goals are imposed on teachers, or just that heads have unrealistic views about how the process is running in practice in their schools, then one would expect no relationship between head, leadership group and classroom teacher views. In the absence of a coordinated approach, individual teachers may have good or bad experiences with their individual department heads, but these would not be consistent within the same school. On the other hand, if there is coordination, one would expect the views of these three groups to converge. We approach this issue by combining a number of questions to head teachers on the contribution of PM in their schools to more effective goal setting, and computing an index based on a factor analysis, and then tabulating these against classroom teacher judgements about goal setting. This provides an indication of the degree to which the two sets of judgements coincide (Table 6, Panel A). Being a factor score, the mean is zero and the standard deviation is unity, so that roughly two-thirds of cases should lie in the range between plus and minus one. Thus, a positive score for school ‘x’ indicates that the strength of the head’s assessment of goal setting is above the average for schools in the sample. Thus in Panel A, we observe that in schools where classroom teachers agree strongly that PM has improved goal setting, the heads’ scores are also above average and in the positive direction (at 0.439), indicating a convergence of their views. Conversely, in schools where the teachers judge PM has not improved their awareness of school goals, we observe lower scores for the head teacher assessments. Thus, as with the ordered probit analysis, we find evidence that, where heads report PM as working effectively, so too do classroom teachers.
24
Table 6.
DAVID MARSDEN AND RICHARD BELFIELD
Comparing the Effectiveness of PM as Seen by Head Teachers and Classroom Teachers in the Same School.
Panel A: Index related to class teacher views of effects of PM (all classroom teachers including those in the leadership group)
Agree strongly Agree Disagree Disagree strongly Neutral/no view
PM has Made Me More Aware of School Targets
PM has Given Me Incentive to Work beyond Job Requirements
PM Means Good Work Now Rewarded
PM makes Managers Set Clearer Targets
0.439 0.100 0.046 0.075 0.214
0.268 0.160 0.101 0.104 0.118
0.202 0.133 0.107 0.057 0.088
0.194 0.108 0.040 0.001 0.143
Panel B: Index related to the leadership group views on the effects of PM on classroom teacher performance PM makes Class Teachers More Aware of School Targets Agree strongly Agree Disagree Disagree strongly Neutral/no view
0.402 0.111 0.105 0.359 0.210
PM promotes Better Work Priorities among Class Teachers
PM increases Importance of Good Middle-Mgt in the School
0.300 0.202 0.127 0.302 0.054
0.181 0.188 0.095 0.217 0.131
The cells show an index of effective goal setting as judged by head teachers, based on factor scores for questions related to improved goal setting. The cells combine scores for waves 2 and 3. They have a mean of zero and a standard deviation of unity.
We are able to take this a step further by examining how far both head teachers and the leadership group (departmental heads and team leaders) coincide in their assessments of the effect of PM on goal setting for classroom teachers (Table 6, Panel B). This is more a measure of consistency of view within the managerial hierarchy in schools. One would expect to find such consistency in schools in which integrative negotiation is a part of PM, which is shown by the rising value of the head teachers’ index as we move towards schools where the leadership group has a positive assessment of PM.
Pay for Performance Where Output is Hard to Measure
Table 7.
25
Class Teacher Views of Changes to PM.
Class teacher responses (all classroom teachers) Now less focus on pupil progress in PM All good teachers can now reach point 3 on UPS School looks to education practices of schools higher up league tables
Yes
No
Neutral/No View
0.0484 0.1205 0.1259
0.1819 0.0890 0.0360
0.0623 0.0950 0.1505
Cells: head teacher scores of whether PM aids goal setting in their school. All class teachers combined.
Finally, we turn to the relationship between heads’ judgements of the effectiveness of goal setting and class teachers’ reports concerning the content and conduct of PM in the same schools. As already mentioned, pupil progress has been the litmus test for the government as to whether or not ‘performance’ is real. As can be seen in Table 7, in schools where head teachers judge PM to be effective, classroom teachers are more likely to report a continued emphasis on pupil progress within PM. Secondly, echoing the element of organisational support to teachers where PM has been well-implemented, teachers are more likely to be confident that merit will be rewarded – ‘all good teachers’ can expect to progress on the upper pay scale. Finally, reflecting the external pressures on school management, it is notable that where head teachers report effective goal setting, class teachers report that their school actively benchmarks its educational practices on other good-performing schools.
6. HOW MANY SCHOOLS HAVE ADOPTED THE ‘REFORMER’ STRATEGY? By now it is clear that behind the general picture of growing acceptance and increasing positive assessment of PM shown in Table 1, there is considerable diversity in the way in which the national scheme has been implemented in individual schools. Thus, there would seem to be schools where it is quite reasonable to accept that the PM has functioned like a process of integrative individual-level negotiation, whereas in others, this seems to be far from the case. This diversity was reflected in some of the written-in comments. One teacher wrote: ‘Teachers feel obliged and pressurised into making unrealistic targets at the beginning of the year which they then cannot fulfil during the year’. On the other hand, one head wrote: ‘PM has improved the positive attitude to professional development and self-improvement within the overall picture of school improvement’.
26
DAVID MARSDEN AND RICHARD BELFIELD
There will clearly be a large grey area of PM practice in between the cases of imposed goals and those where there is a real spirit of negotiation. Therefore, a precise allocation of schools into one or the other category is impossible and any measure has to be taken with a grain of salt. As schools change over time, the manner in which they apply PM may also evolve. Therefore, it is better to talk of strategies adopted by schools and which may change, rather than to seek to characterise individual schools. Thus, a first strategy, hinted at earlier, is the ‘reformer’ strategy, where the school seeks to use PM as an opportunity to improve management and to use goal setting to improve coordination within the school. On the other hand, the ‘firefighter’ strategy groups two broad approaches: to use PM to get teachers their pay rise, and so reduce staff recruitment and retention problems; and to use PM to impose targets on teachers in an attempt to raise performance. Both of these are likely to be driven by crises. A school may well switch strategies, from being a ‘fire-fighter’ to being a reformer for example, once retention problems have been resolved, and there is time to move on to other issues. One way to identify the ‘reformer’ strategy might be to enquire whether goal setting is integrated within a school, and in the spirit of the previous discussion, to explore whether both classroom and head teachers’ agree that PM has improved goal setting in their schools. We characterise schools with the ‘reformer’ strategy as those in which head teachers give above average scores using the index of goal-setting quality in Tables 6 and 7, and where the classroom teachers agree that it has improved goal setting. On these rough and ready estimates, we may characterise between 15 and 25% of schools in the sample as pursuing the ‘reformer’ strategy (Table 8). This can be seen in the top row, which uses the largest number of observations available by pooling responses for 2001 and 2004. A similar analysis was carried out for the other indicator used in the earlier tables, namely, whether PM had increased one’s personal awareness of school targets. The effects were somewhat smaller, but so too was the number of positive replies to that question. These estimates are consistent with those of early case studies by Wragg, Haynes, Wragg, and Chamberlin (2001), and by Mahony, Menter, and Hextall (2003), who found modest effects of the new system in changing the way teachers perform. However, what these early studies could not capture is the growth in the number of schools adopting the ‘reformer’ strategy. Even though the small number of effective observations beckons caution, whichever measure is used, there has been an increase of about ten percentage points in such schools: from 10–20% in 2001 to 20–30% in 2004. The growth in the numbers of ‘reformer’ schools was confirmed by
Pay for Performance Where Output is Hard to Measure
27
Table 8. Percentage of Schools with ‘Joined-up’ Goal Setting: Schools in which class teachers respond ‘agree’ given that both head teachers and the leadership group agree that PM has improved goal setting in their schools. PM has Led to Improved Goal Setting in My School (a)
(b)
(c)
(d)
% Agree
% Agree
% Agree
% Agree
26.0 273
15.6 224
14.3 26.8 70 82
8.2 20.3 85 74
Cross section analysis pooled across waves Waves 2 & 3 using pooled responses 28.7 26.5 Schools with a complete set of observations 397 219 Panel analysis using wave-specific responses Wave 2 Wave 3 No complete obs. wave 2 No complete sets of obs. wave 3
20.4% 35.8% 206 137
27.2 38.5 81 65
Note: Column headings: (a) Heads and all classroom teachers. (b) Heads and leadership group. (c) Leadership group and non-leader class teachers. (d) Heads, leaders and non-leader class teachers. The heads’ judgements on goal setting were based on the same index as that used in Tables 6 and 7, but taking the median value as dividing positive from negative judgements as to whether PM had improved goal setting. The leadership group values relate to those who agreed that PM had improved goal setting in their schools, as did the class teacher question. Members of the leadership group were identified from the biographical data obtained in the first wave. The number of observations in the panel was reduced by the need to have responses from both heads and the relevant group of classroom teachers for each school (Although we had information from 424 schools for the balanced panel analysis, this yielded only 214 in wave 2 and 139 in wave 3 with information on the head’s assessment of improved goal setting. The numbers of schools with a complete set of observations dropped further when tabulating these replies against those of classroom and leadership group teacher. This will tend to eliminate smaller schools disproportionately. Results are unweighted and do not correct for over-sampling of secondary schools).
computing a transition matrix between such schools in 2001 and 2004, which showed strong movement from ‘fire-fighter’ schools in 2001 to the ‘reformer’ category in 2004, with only small number of moves in the opposite direction. Thus, one can say that the number of schools adopting the ‘reformer’ strategy has increased quite strongly since the inception of the new system.
28
DAVID MARSDEN AND RICHARD BELFIELD
7. SCHOOL PERFORMANCE Finally, we turn to the effects of PM on school performance as reflected in the government’s league table performance indicators. It is impossible to test whether the new system has led to improved pupil results for the whole country because there are too many other factors at work. However, it is possible to examine whether the schools which implemented performance management most thoroughly achieved greater improvements in their results than those which did not. Comparing improvements in goal setting with improvements in academic performance over time across schools enables us to factor out many of the possible competing variables. It should be added that national tests are externally assessed so that although there has been debate about grade inflation across the system as a whole, the scope for any individual schools to inflate their grades is small. It is quite likely that using a relative measure understates the possible full effect of PM on school performance, but we have no simple way of adjusting for the unknown amount of general grade inflation. To gauge these we take changes in the academic results of schools at Key Stage 2 (age 10) for primary schools, and for GCSE8 (age 15) for secondary schools (Table 9). We identify those schools which improved their relative academic performance as reflected in the results for school years ending 1999–2000 and 2002–2003, and compare this group with those reporting an improvement in goal setting between waves 2 and 3. The sample numbers of those changing between waves 2 and 3 are quite small, so the results can be only tentative, but we found a positive and statistically significant relationship: schools whose heads report improvements in goal setting were more likely also to have improved their academic results. Confidence in these results is boosted by similar findings by researchers at the CMPO, Bristol University, using a completely different methodology, which show a positive effect of threshold assessment on the academic Table 9. (Column %) Improved Use of Goal Setting in School
No Yes
Goal Setting and School Performance. Improved Relative Academic Performance at School No
Yes
53 47
40 60
Note: Table shows changes between waves 2 & 3. Goal-setting factor score as in Tables 6 and 7. Based on the 169 observations for schools that could be matched. w2 ¼ 3.1; P ¼ 0.079.
Pay for Performance Where Output is Hard to Measure
29
performance of pupils in classes taught by eligible teachers (Atkinson et al., 2004). The CMPO study examines the change in academic achievements of pupils in classes taught by teachers who were, or were not, eligible to pass the threshold. They found that pupils with teachers eligible to pass the threshold were more likely to improve their performance. Although the CMPO study emphasises the financial incentive of passing the threshold, their evidence is consistent with this paper. Their statistical results could equally well derive from improved goal setting as opposed to simple financial incentive. Although the findings of our own study suggest that most teachers are not strongly motivated by the extra financial rewards offered, it is clear that going through the threshold application procedures would make them more amenable to focusing on the goals agreed with their head teachers.
8. CONCLUSIONS This paper has shown that despite initial academic scepticism and the hostility of many teachers and their unions, performance management had, by 2004, taken root in many state schools in England, and was contributing to improved goal setting. Although such schools were still in the minority by 2004, in those where PM had become well established, it had also contributed to improved pupil performance. The reason the initial academic scepticism was misplaced was that insufficient account was taken of how the problems of performance definition and monitoring could be resolved by appropriate goal-setting measures. This suggests that if we take up Prendergast’s challenge to look at performance incentives across a wider range of occupations, we need to give more consideration to mechanisms for defining and agreeing performance goals. Although some of the leading exponents of goal setting theory have systematically downplayed the importance of financial incentives (e.g. Locke & Latham, 1990), they have assumed a key supporting role in the teachers’ case. For teachers, the link with pay has been of central importance. At one level, it was the ‘bribe’ to gain acceptance for the new system: no performance management, no large pay increase. But there is also a sense in which it has functioned as performance-related pay, and so contributed to the negotiation between school management and classroom teachers. Heads can decide whom to put forward for the threshold assessment, and they can choose when to propose their colleagues for movement along the upper pay scale. The evidence from a teacher’s performance review is one of the key items in the submissions for pay progression. Up until the time of the latest
30
DAVID MARSDEN AND RICHARD BELFIELD
wave, in practice and ex post, success rates at the threshold and for upper pay scale progression have been very high, and some have argued it is virtually automatic. However, this is to miss that the power heads have to advance or postpone proposals for upper pay scale advancement, and the impact of the long-running uncertainty over central government funding for performance increases. Ex post the pass rates may be high, but ex ante, in the eyes of most teachers and their heads, according to the evidence of the CEP surveys, the outcome is uncertain. It is surely the ex ante prospect that drives behaviour rather than the ex post knowledge. The link with pay introduces another element, namely that of negotiation. It has been argued in this paper that an appropriate framework for considering the effective cases of goal setting within PM is that of integrative negotiation. The rates of pay are fixed by national pay scales, and schools are constrained by Education Department rules as to how they allocate their budgets. Nevertheless, putting a teacher forward for the threshold or for upper pay scale progression earlier rather than later is one of the measures of organisational support available to head teachers when trying to work out solutions to the kind of problems that are the focus of PM: alignment of individual and collective goals, and renegotiating work priorities. Walton and McKersie were careful to retain a reference to the parties’ utility functions within their analysis of integrative bargaining. Problem-solving has perhaps always a technical component, but the chosen solution nearly always also affects the welfare of both parties. Putting a teacher forward for progression is one of the elements of support alongside professional development, and adjustment of inappropriate workloads that can be used to help persuade individual teachers to adjust their work priorities to those needed by their schools. Could this be achieved without the link with pay? A previous Conservative government introduced a national teacher appraisal scheme in 1991, yet it was not widely implemented in practice (Ofsted, 1996), and what was left of it was replaced in 2000. In contrast, PM has been almost universally implemented, and as shown in the CEP surveys, its design in most schools, as concerns written objectives and so on, follows the key Education Department’s guidelines. Although it is difficult to probe in this survey, there are signs of important differences between the ‘fire-fighter’ schools, which have just used PM as a form-filling exercise to get teachers their pay increase, and the ‘reformer’ schools. The former dissociated the two processes, and have paid the price as PM appears to have been less effective than in the ‘reformer’ schools, and arguably less beneficial in terms of pupil attainments.
Pay for Performance Where Output is Hard to Measure
31
One piece of evidence that would have helped to clinch the argument about changing priorities and practices by classroom teachers as a result of PM was lost in the course of the survey. It had been hoped to track how teachers allocated their discretionary time, this being considered a better measure than questions about changes of priorities and practices. Unfortunately for the study, measurement of this variable was disturbed by conflicts over teachers’ working hours, and government action to ‘remodel’ teachers’ working time. It also proved impossible to capture teachers’ working time use at the same point in the school year while at the same time surveying attitudes at the same point in the performance management cycle. As a result, there proved to be too much ‘noise’ in this key variable. Nevertheless, we know from the replies of head teachers that many of them have shifted resources and teaching priorities towards more academic and test subjects in response to league table pressures, and these are more likely to include ‘reformer’ schools. With this, the good practice examples of performance reviews and the evidence of improved pupil attainments in ‘reformer’ schools, we can infer that many teachers have adjusted their work priorities as a result of performance management where it has been well run. By the time of the last wave in 2004, it was clear that effective performance management was spreading as a result of more schools adopting a systematic approach rather than diffusing evenly by the same amount across all schools. Hence the interest in identifying those with the ‘fire-fighter’ and ‘reformer’ strategies, and tracking those that switch. Adopting the ‘reformer’ strategy opens up not just a one-off change in performance for the school, but rather provides it with the means of addressing continuously changing priorities. Professional groups may bring a high degree of motivation by virtue of their professional ethos, but that can also prove conservative in the face of such changes. The cases of systematic performance management suggest that by approaching goal setting as a form of integrative negotiation, a way is opened to addressing changing school priorities on a long-term basis. Recognising this is important, because professional groups, such as teachers, bring expert knowledge to their work that is not always accessible to management. Even when the manager is another teacher, there may be differences in subject knowledge and teaching methods, so that it is hard to identify and impose top-down new objectives and the strategies to achieve them. Approaching the process as one of integrative negotiation enables both parties to engage in a dialogue on their respective objectives, the means to achieve them, and the measures of organisational support needed. In such cases, agreement to goals is arguably the more appropriate means of gaining commitment to fulfil them.
32
DAVID MARSDEN AND RICHARD BELFIELD
ACKNOWLEDGMENT We acknowledge the financial support of the Leverhulme Foundation Future of Unions project, the ESRC, and STICERD. We also wish to thank the teachers’ unions, the Education Department, and all the classroom and head teachers whose support and comments have made this study possible. We also wish to thank the participants at the 2006 LERA conference, and the referees for their valuable comments. We should also like to thank our colleagues at CEP for their generous comments, and Maria Koumenta and Ettore Ricciardi for their invaluable research assistance in preparing the data on school performance. Thanks also for comments given by colleagues at the CEP Tuesday workshop, and the CIPD Professional Standards Conference, 2005.
NOTES 1. We are indebted to Shushi Okazaki of the Akahata newspaper for information about schools in Japan. 2. The advertising slogan was used by the government’s Teacher Training Agency in the early 2000s. 3. We are indebted to Mike Chapman, former head of Cambridge Educational Associates, for first drawing our attention to the different ways in which schools were approaching the new system. The CEA was in charge of the system of schools’ external advisors for the first wave of threshold assessments. 4. Data for schools in Wales were not included in this analysis. 5. This is described in Marsden and Belfield (2005a). 6. Respondent #622 Classroom teacher-wave 1. 7. Team leaders are classroom teachers exercising middle management responsibilities. 8. General Certificate of Secondary Education.
REFERENCES Adnett, N. (2003). Reforming teachers’ pay: Incentive payments, collegiate ethos and UK policy. Cambridge Journal of Economics, 27(1), 145–157. Atkinson, A., Burgess, S., Croxson, B., Gregg, P., Propper, C., Slater, H., & Wilson, D. (2004). Evaluating the impact of performance-related pay for teachers in England. Working Paper 04/113, Centre for Market and Public Organisation, University of Bristol, p. 61. DfEE. (1999). Performance management framework for teachers: Consultation Document, Department for Education and Employment, London, September 1999 Dolton, P., McIntosh, S., & Chevalier, A. (2003). Teacher pay and performance. London: Bedford Way Papers, Institute of Education.
Pay for Performance Where Output is Hard to Measure
33
Furnham, A. (1997). The psychology of behaviour at work: The individual in the organization. Hove and New York: Psychology Press. Gibbons, R. (1998). Incentives in organizations. Journal of Economic Perspectives, 12(4), 115–132. Glennerster, H. (2002). United Kingdom Education 1997–2001. Oxford Review of Economic Policy, 18(2), 120–136. Lavy, V. (1999). Evaluating the effect of teachers’ performance incentives on pupils’ achievements. Working Paper. Hebrew University of Jerusalem. Lazear, E. (1996). Performance pay and productivity. NBER Working Paper #5672. Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance. Englewood Cliffs, NJ: Prentice-Hall. Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist, 57(9), 705–717. Mahony, P. Menter, I., & Hextall, I. (2003) The impact of performance threshold assessment on teachers’ work: Summary report. Roehamton, University of Surrey, Research Report on ESRC project: The impact of the performance threshold assessment on teachers’ work (ESRC 0002239286). Makinson, J. (Chair) (2000). Incentives for change: Rewarding performance in national government networks. Public Services Productivity Panel, HM Treasury, London. Malcomson, J. M. (1997). Contracts, hold-up, and labor markets. Journal of Economic Literature, 35(4), 1916–1957. Marsden, D. W. (2000). Teachers before the ‘threshold’. Centre for Economic Performance, Discussion Paper 454, London School of Economics. Marsden, D. W., & Belfield, R. (2005a). Unions and performance-related pay: What chance of a procedural justice role? In: S. Fernie & D. Metcalf (Eds), Trade unions: Resurgence or demise. London: Routledge. Marsden, D. W., & Belfield, R. (2005b). Performance pay for teachers: Linking individual and organisational-level targets. Centre for Economic Performance, Discussion Paper 703, London School of Economics. Marsden, D. W., & French, S. (1998). What a performance: Performance related pay in the public services. Centre for Economic Performance Special Report, London School of Economics, London, available online from www.cep.lse.ac.uk Murnane, R. J., & Cohen, D. K. (1986). Merit pay and the evaluation problem: Why most merit pay plans fail and a few survive. Harvard Educational Review, 56(1), 1–17. Ofsted.(1996). The appraisal of teachers 1991–1996: A report from the Office of Her Majesty’s Chief Inspector of Schools. Ref: HMR/18/96/NS, Office for Standards in Education (Ofsted), London, 26 pp. Prendergast, C. (1999). The provision of incentives within firms. Journal of Economic Literature, 37, 7–63. Richardson, R. (1999). Performance related pay in schools: An assessment of the Green Papers. A Report for the National Union of Teachers. National Union of Teachers, London. Rousseau, D. (1995). Psychological contracts in organisations: Understanding written and unwritten agreements. Thousand Oaks, CA: Sage. Teulings, C., & Hartog, J. (1998). Corporatism or competition? Labour contracts, institutions and wage structures in international comparison. Cambridge: Cambridge University Press. Tzioumis, K. (2005). Essays on managerial incentives: Relative performance evaluation, stock option introduction, union effect and politics in CEO compensation in the United States. Doctoral thesis, London School of Economics, University of London.
34
DAVID MARSDEN AND RICHARD BELFIELD
Varlaam, A., Nuttall, D., & Walker, A. (1992). What makes teachers tick? A study of teacher morale and motivation. Clare Market Papers, No. 4, Centre for Educational Research, London School of Economics, London. Walton, R. E., & McKersie, R. B. (1965). A behavioral theory of labor negotiations: An analysis of a social interaction system. New York: McGraw-Hill. Wragg, E. C., Haynes, G. S., Wragg, C. M. & Chamberlin, R. P. (2001). Performance related pay: The views and experiences of 1,000 primary and secondary head teachers’, Teachers’ Incentive Pay Project, Occasional Paper No. 1, University of Exeter.
APPENDIX. SAMPLE DESIGN AND QUESTIONNAIRE The questionnaire was developed in close consultation with the teachers’ unions and the Department for Education and Skills, and piloted on groups of lay representatives. For the first wave, a random sample of 1,675 schools was drawn from the Register of Educational Establishments for England and a similar register for Wales, and packages of questionnaires were sent to head teachers. Heads received a covering letter explaining the nature of the study, that it had the support of the head teachers’ associations, and that it had been developed in consultation with the teachers’ unions and the DfEE. They were asked to distribute the questionnaires. In small schools with under 35 teachers, every teacher was sent a questionnaire, and in larger ones, heads were asked to select every nth teacher off the school’s staff list depending on the size of the school. Heads were asked to complete a special questionnaire. Being a panel study, the initial respondents to wave 1 were approached again for waves 2 and 3. The overall response rate to the first wave was about 20%, which comprises a double response: whether the head teacher agreed to distribute the questionnaires in the first place, and then whether the teachers themselves chose to reply. In many schools, head teachers have a policy of not distributing questionnaires in their schools in order not to add to the workload on their teachers. Initially, the panel included replies from about 4,000 teachers and about 1,000 heads. Accounting for sample attrition, it is possible to link replies from about 1,000 teachers and about 300 heads over time through the panel. We conducted a number of checks to see whether the panel results reported here differ statistically significantly from the simple cross-section results for each wave, and found that on the whole they do not. The questionnaire used for classroom teachers and for head teachers in wave 3 can be found in Marsden and Belfield (2005b). That for wave 1 can be found in Marsden (2000).
DISPUTE RESOLUTION IN COOPERATIVE AND HIERARCHICAL WORKSITES$ Elizabeth A. Hoffmann ABSTRACT This study compares dispute resolution strategies of workers in hierarchical, conventional businesses with those of members of worker cooperatives – organizations in which all workers co-own and co-manage the business. Drawing on data from three industries (coal mining, taxicab driving, and food distribution), this study finds some support for predictions in the literature that assert that the cooperative’s flattened structure and egalitarian ideology will affect workers’ grievance resolution. Although the data do not indicate a single pattern in dispute resolution strategies (i.e., with all members of the cooperatives resolving their disputes one way and all non-cooperative employees using a different strategy), the data do demonstrate that, when comparing matched cooperative and conventional businesses within each industry, the worker cooperative members possess more dispute resolution strategies than their conventionally employed counterparts.
$
This research was supported by a National Science Foundation grant (SBR-9801948).
Advances in Industrial and Labor Relations, Volume 15, 35–66 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15003-9
35
36
ELIZABETH A. HOFFMANN
INTRODUCTION The complex dynamics involved in workplace grievance behavior is an important subject of industrial relations scholarship. Previously, however, researchers have relied almost exclusively on observations of traditional, hierarchical organizations – even when making broad statements about the fundamental character of workplace disputes. My research moves beyond this conventional wisdom to consider the question of how formal and informal workplace disputes are raised in the absence of formal hierarchy. At a theoretical level, this research is important because it helps to disentangle the impacts of hierarchy and power, while at an applied level it provides insights into the feasibility of a key plank in many progressive platforms – worker involvement and ownership. Specifically, this study compares employees’ grievance resolution strategies at matched pairs of worker cooperatives and conventional businesses in three very different industries: coal mining, taxicab driving, and wholefood distribution. Extant research suggests that organizational structure, ownership, and ideology might greatly affect how employees address their problems at work, i.e., their grievance behavior (e.g., Bailyn, 1997; Hochner et al., 1988; Hoffmann, 2001b; Rothschild & Whitt, 1986; Tjosvold, Morishima, & Belsheim, 1999; Tucker, 1999; Whyte, Tove, Christopher, Reed, & Stern, 1983). Because this study draws on several literatures, it addresses various predictions on dispute resolution in worker cooperatives. Some industrial relations literature cautions that worker cooperatives might not be a viable alternative to the conventional, hierarchical business. This research suggests that these worker cooperatives may be less efficient and less likely to succeed as organizations. However, if these businesses do struggle into existence and succeed, their workers might enjoy such benefits as greater respect and recognition and less labor–management conflict (e.g., Hochner et al., 1988). Indeed, some literature on worker cooperatives suggests that evenly distributed power and greater worker participation should produce workers who are very able to assert their needs and raise necessary grievances. The grievance behavior research asserts that greater trust and shared goals among workers and worker–managers facilitate easier and more successful dispute resolution; one might imagine that increased trust and common goals will be more prevalent in worker cooperatives, where inclusion, equality, and worker participation are officially encouraged (e.g., Tjosvold et al., 1999). Thus, worker cooperatives present a stark contrast to conventionally organized businesses in that the cooperatives attempt to evenly distribute
Dispute Resolution in Cooperative and Hierarchical Worksites
37
power, encourage worker control through egalitarian ideologies and flattened management structure, and engage in concerted efforts to minimize power imbalances. Nevertheless, they do continue to have many of the same goals of conventional businesses, such as profits and efficiency, and, thus, allow for instructive comparisons. Although this study does not find a worker cooperative impact across industries – e.g., I did not find that all worker cooperative members easily resolved disputes, while all conventional business employees had greater difficulty – I did find a worker cooperative effect within each industry. Specifically, I found that, when comparing workers within each industry, the members of the worker cooperatives had more ways to resolve workplace disputes than their conventionally employed counterparts in the same industry.
THEORETICAL CONTEXT AND PREVIOUS RESEARCH Workplaces are often sites of conflict and, accordingly, much literature has examined workplace grievance resolution. Some literature suggests that, by equalizing power, many difficulties of workplace grievance resolution – such as accessibility and mobilization – might be successfully resolved. Other literature rejects this possibility, implying that redistribution of power is not sufficient to circumvent fundamental problems of grievance resolution. Other theorists assert that worker cooperatives are simply not viable, robust business alternatives. Worker cooperatives offer a type of organization that experiments with all of these theories. As collectively owned, flattened hierarchies, they minimize official power inequities. Through egalitarian ideologies these workplaces attempt to equalize unofficial power. By focusing on mutual trust, cooperation, and worker empowerment, they endeavor to create a different environment for dispute resolution. The Inevitability of Hierarchy According to Weber (1946, p. 27), the operation of an organization without a hierarchy would be ‘‘utopian’’ and impossible to achieve in modern society. He believed that eliminating hierarchy from large-scale organizations would be utopian because this would require radical structural changes, not merely alterations in the distribution of power (Rothschild & Whitt, 1986). Moreover, the absence of hierarchy would be utopian because of its strong
38
ELIZABETH A. HOFFMANN
link to bureaucracy, a link which Weber holds is inevitable in modern society. Hierarchy provides both the apparatus that bureaucracy needs and the encouragement of beliefs on which bureaucracy bases its claims to legitimacy, facilitating greater efficiency (Weber, 1946). Like Weber, Robert Michels doubted the possibility of large-scale, non-hierarchical organizations (Michels, 1981, pp. 38, 43). His famous statement: ‘‘Who says organization, says oligarchy’’ (Michels, 1981, p. 49), vehemently expresses the classical disbelief in collective management. Arguing more strongly than Weber for this inclination, he stated that ‘‘organization implies the tendency to oligarchy y. As a result of organization, every party or professional union becomes divided into a minority of directors and a majority of directed’’ (Michels, 1981, p. 42). Because most organizational theorists assume Michels’s ‘‘Iron Law of Oligarchy’’ as a given in their research, few studies seek to explore the functionings of organizations where hierarchy and oligarchy are not present. Several studies have addressed hierarchy, yet these, too, conclude that hierarchy is unavoidable. In their well-known 1956 study of a democratic labor union, the International Typographers’ Union, Lipset, Trow, and Coleman argue that hierarchy is necessary (Lipset, Trow, & Coleman, 1956, p. 361). In other research, scholars assert that hierarchy is not just inevitable, but beneficial. Hannan and Freeman (1989) suggest that the likelihood of success is diminished if an organization operates without a hierarchical structure, although they, too, do not identify hierarchy specifically. They argue that the possibility of structural innovations may cause a loss of technical efficiency and may create costs in legitimacy for the organization with regard to its institutional environment (Hannan & Freeman, 1989). Similarly, institutionalist theorists argue that hierarchy is an important rational myth. If an organization operates without the structure of hierarchy, it may lose the legitimacy from others that it needs to succeed. The institutional legitimacy of hierarchy makes social action more easily reproduced (Zucker, 1977). However, institutional theorists do not agree with Weber and others on the technical ‘‘benefits’’ of hierarchy. Other scholars do not perceive hierarchy as beneficial. Some classical theorists have perceived hierarchy negatively, the most famous of these being Marx. Understanding organizations as power systems that are designed to maximize profit and control, Marx argued that hierarchy is utilized to ‘‘deskill’’ workers, rather than to improve efficiency. Thus, hierarchy is not a rational system of coordination, as Weber believed, but ‘‘an instrument of control and a means of accumulating capital through the appropriation of surplus value’’ (Scott, 1992, p. 115).
Dispute Resolution in Cooperative and Hierarchical Worksites
39
However, while Marxists assert that the use of hierarchy in the workplace is to control and oppress, they do not necessarily support systems of collective management. Marxism argues that cooperative management is misguided because it fails to challenge the fundamental exploitative nature of organizations. Marx himself felt that cooperatively held property would precede socialism. However, the type of social property arrangement he addressed would be formally developed on a national scale and government sponsored (Marx & Engels, 1986). Additionally, efforts to flatten organizational hierarchy may retard broader-scale efforts for change as well as impair the workings of the organization itself. Cooperative re-organization may harm the organization both in its technical environment, with its emphasis on the production of a product or service, and in its institutional environment, with its emphasis on conformity to elaborate rules and requirements. Although technical and institutional environments operate with different logics and demand different criteria, combinations of both types of environments exist for nearly all organizations. Since all organizations must interact with both environments (Scott, 1992), adopting a non-hierarchical means of operating could have perilous effects on the success of an organization. Organizations without Hierarchy Despite these pessimistic assertions by many scholars, collectively run organizations attempt to escape from the inevitabilities of hierarchical control. Cooperatives hold themselves to be organizations operating without Weber’s assumption of domination in that there is no one person or small, select group holding the ultimate authority (Rothschild et al., 1986). Some contemporary theorists suggest that, in less conventional, alternative groups, hierarchy does not necessarily follow from what would inevitably produce hierarchy in most situations. These researchers argue that, because these subgroups interpret conditions differently from mainstream society, ‘‘hierarchy-related’’ conditions may not necessarily lead to hierarchy for them. Collins and other scholars have argued that subgroups within society differ in their orientations toward assumed, ‘‘taken-for-granted’’ norms and in how they invest cultural resources (Collins, 1981). The inevitability of hierarchy, therefore, might not be assumed by people who hold less conventional beliefs, especially with regard to formal organizations. In addition, Scott has argued that cultural frames establish which means are appropriate and which ends are desirable (Scott, 1992). In this way, people with cultural
40
ELIZABETH A. HOFFMANN
frames that differ from those of mainstream society may view means and ends that are commonly deemed utopian, such as a non-hierarchical workplace, as highly possible and worthwhile. Thus, the ‘‘given’’ of the hierarchical workplace may not be embraced by subgroups with different cultural frames or orientations, because they have different goals and norms and accept alternative means and mandates. Rothschild and Whitt (1986) found that interviewees in their work on various forms of cooperatives often held non-conventional beliefs, especially with regard to hierarchy and efficiency, the capitalist economic system, and government actions. Many people in their study identified themselves as Marxists or anarchists, or aligned themselves with such ideologies. These ideologies will encourage contrasting and opposing values from mainstream society and deny some ‘‘taken-for-granted’’ norms. Thus, even though certain conditions may exist which would generally be considered to mandate a hierarchical organization, hierarchy is not automatic within these groups. Powell and DiMaggio (1991, p. 29) argue that institutional models will not be imported in their entirety into systems that are very different from the ones in which they originate. In addition, despite these subgroups’ alternative ideologies and commitment to eliminating certain de-valued or rejected aspects of conventional organizations, they will nevertheless continue to operate using some elements of the system they oppose. Thus, these subgroups may have some aspects of mainstream organizational culture, such as paychecks, regular operating or business hours, and income tax deductions; while not necessarily including all aspects, namely hierarchy.1 Scott argues that, under some conditions, institutional environments may reduce rather than increase the quantity or elaborateness of organizational structure. He explains that cultural controls can substitute for structural controls in some circumstances. ‘‘When beliefs are widely shared and categories and procedures are taken for granted, it is less essential that they be formally encoded in organizational structures’’ (Scott, 1987, p. 507). This suggests that when the members of an organization share goals, rather than having to negotiate the opposing goals through formal mechanisms, the structures of the organization can be less rigid. This allows for the possibility of operating without needing a hierarchy to impose elaborate structure and control on the members of the organization. This may be more likely in organizations where members not only share goals, but also more subtle cultural understandings and symbols which provide order, directly influencing the beliefs and behaviors of individual participants. Thus, ‘‘(o)rganizational structures may only be required to support and supplement those cultural systems that exercise a direct
Dispute Resolution in Cooperative and Hierarchical Worksites
41
influence on participants’’ (Scott, 1987, p. 507). This argument suggests that hierarchy could be avoidable in organizations with a high level of homogeneity, where goals, beliefs, conceptual language, and symbols are shared among the members. A type of organization that tends to be less hierarchical is the ‘‘clan’’ form of organization. Ouchi (1980) has developed a model of organizational efficiency, using three mechanisms: markets when the actors do not share the same objectives or goals, yet there is no ambiguity as to what each actor will do to fulfill her/his part of the bargain; bureaucracies when there are high levels of both goal incongruence and performance ambiguity; and clans when goals are shared among actors even though performance ambiguity remains high.2 While markets rely on the immediateness of the transaction to ensure compliance and performance and bureaucracies employ legitimate authority and rules to ensure long-range performance, clans rely on socialization to ensure performance and boundary-enforcement. Like the bureaucracy, the clan form has the benefit of orchestrating longterm performance. However, rather than dealing with difference between the employers’ goals and the workers’ goals by maximizing power imbalances and supervision, clans create goal congruence among workers and employers.3 The clan develops an organic solidarity among all members of the organization. This high level of goal congruence removes the need for a hierarchy of power and supervision, since ‘‘[w]here individual and organizational interests overlap y rewards can be achieved at a relatively low transaction cost’’ (Ouchi, 1980, p. 136). In this way, clans do not have to rely on Weber’s rational legal form, but instead call on the ‘‘[c]ommon values and beliefs [that] provide the harmony of interest’’ for the organization (p. 138). Thus, the clan form does not work around its goal incongruence, which would necessitate creating a hierarchy, but rather it eliminates the goal incongruence, removing the need for hierarchy. Socialization might produce this high level of goal congruence. By exerting informal power to teach common values and eliminating differences in goals, hierarchy is avoided. While Ouchi and his collaborators primarily focused on clans operating in modern mainstream industrial organizations, his concept of the clan can also be applied to the worker cooperative. Ouchi used his clan concept to briefly discuss the business side of the utopian communities researched by Kanter. He compared his explanation of the workings of clans with Kanter’s discussion of the cooperative communities she studied (Ouchi, 1980). Kanter found that a key to a collective’s success was a widely shared belief that individual interests are best served by an immersion of the individual interests into those of the greater whole (Kanter, 1972). The importance of
42
ELIZABETH A. HOFFMANN
shared goals and beliefs that run counter to those of mainstream society has been supported by the findings of Collins (1981) and Zucker (1977). Additionally, Swidler’s work on ‘‘free schools,’’ which attempted to minimize the level of authority, hierarchy, and power differences in the classroom, demonstrates that the perpetuation of a ‘‘common purpose’’ is essential for ideological appeals to be effective (Swidler, 1979). Rothschild and Whitt, in their work on a variety of collectivist organizations, also emphasize the necessity of ideological appeals as the main means of control in order to avoid using centralized, hierarchical authority (Rothschild et al., 1986). However, these studies either centered around types of cooperatives other than worker cooperatives, or, if they did study worker cooperatives, they did not focus on dispute resolution. For example, while both Kanter’s (1972)4 work on communes and Swidler’s (1979) 5 work on free schools provide indepth analyses of non-hierarchical institutions, neither study examines cooperative workplaces nor addresses grievance resolution more than briefly and tangentially. Rothschild and Whitt (1986),6 Iannello (1992),7 and Hacker (1989)8 do examine worker cooperatives, but do not explore dispute resolution. Obstacles to Dispute Resolution Despite its absence in the worker cooperative literature, dispute resolution is an important issue, since dispute resolution is often difficult for workers. Research in hierarchical, conventional organizations demonstrates that individuals who bring grievances in their workplace institutions often face the significant disadvantage of being parties with few or no other experiences with the formal grievance procedure at hand. The opponent, the management, is likely to have past and possibly on-going experience with the arena (Galanter, 1974). Indeed, even when an organization experiences turnover among its managers, the institutional memory and collective experience of the remaining managers still gives the management side of any grievance an advantage – a structural advantage, not an advantage linked to individual managers’ personal abilities. Thus, management has an idea of what to expect and how to plan and strategize to maximize any possible advantage (Galanter, 1974). Unlike individual disputants whose only focus must be the specific case at hand, management can use each case as a way to play, not only for the specific outcome, but also, to influence the rules – both the general rules and content-specific ones. In other words, management can ‘‘play the odds’’ in that the stakes are less for it than for each individual disputant, so it can afford to lose or settle a few strategically unimportant
Dispute Resolution in Cooperative and Hierarchical Worksites
43
cases, as long as the more important victories are won. However, to the individual disputant, the case at hand is the only chance to win her/his issue; s/he cannot cut her/his losses over the next several cases because there may be no other cases for her/him (Galanter, 1974). However, this absence of concern about long-term effects gives the individual disputant a key advantage; s/he can ‘‘do her/his damnedest’’ without fear of reprisal the next time s/he goes to court or brings a grievance against the other party because there will be no next time (Galanter, 1974). One key reason Bumiller, in her research on workplace grievances, provides to explain why the people in her study did not pursue their claims is that they ‘‘legitimized their own defeat’’ (Bumiller, 1988, p. 29). Many did this by characterizing the struggle against perpetrators as ‘‘unwinnable’’ and ‘‘me against the corporation’’ (Bumiller, 1988, p. 52). For example, many of the women interviewed by Bumiller preferred to rely on their own interpersonal skills rather than on assistance from formal avenues. Moreover, the belief that authority is benevolent contributed to a more passive and accepting attitude by victims toward mistreatment by superiors. Assuming a type of paternalism, the victim falls back into an acceptance of conditions and that acceptance inhibits her/his ability at each stage of the grievance process (Bumiller, 1988). Dispute resolution in worker cooperatives might be free of many of these obstacles (e.g., Hochner et al., 1988; Hoffmann, 2001a; Rothschild & Whitt, 1986; Tjosvold et al., 1999; Tucker, 1999; Whyte et al., 1983). For example, because the cooperatives emphasize equality and attempt to empower workers, they might not see their cause as ‘‘unwinnable,’’ as Bumiller’s subjects did. Many of the typical worker–management struggles might be eliminated (Hochner et al., 1988; Tjosvold et al., 1999). Alternatively, members of cooperatives might be even more susceptible to abandoning their grievances in the belief of a paternalistic benevolence on the part of the cooperative or fellow workers (Rothschild & Whitt, 1986; Tucker, 1999). Additionally, the cooperative itself might retain its structural advantages (Galanter, 1974). Worker Cooperatives Cooperatives see themselves as organizations operating without Weber’s assumption of domination in that no one person or select group holds all authority (Rothschild & Whitt, 1986). Thus, they assert that many of the barriers to raising grievances should be greatly lessened, if not altogether absent, in worker cooperatives (Cornforth, Chris, Alan, Jenny, & Roger,
44
ELIZABETH A. HOFFMANN
1988; Thornley, 1981). Because all workers are owners of the business, each should be empowered to assert her/his needs without the fear of having to face any ‘‘tyrannical power’’ or of her/his struggle being ‘‘unwinnable.’’ The cooperative structure and ideology should enable the members to raise issues and concerns, even unpopular ones (Hochner et al., 1988; Hoffmann, 2005). Linehan argues that ‘‘(b)y participating in cooperatives, workers acquire new skills in organization and in self-management. Together they achieve what none of them could do alone. In this way, workers’ cooperation allows people an opportunity to gain self-confidence and become more self-reliant.’’(1983, p. 18). The reduced reliance on control in worker cooperatives allows for greater worker initiative and for team cooperation in problem-solving, in contrast to top-down monitoring,9 which encourages conventional, hierarchical firms to organize production technically and bureaucratically in order to minimize the task complexity and to maximize the ease of monitoring (Putterman, 1982, p. 147). However, to what extent such dogma is realized in the day-to-day operations of worker cooperatives is controversial. Commitment to cooperative management and ownership must be matched with a similar commitment to being a stable business (Cowling, 1943; Rock, 1991). This delicate balance is often difficult to achieve. Unlike conventional businesses which view their services and products merely as essentials of the profit-making process rather than as ends in themselves, worker cooperatives, ideally, produce goods and services to benefit both those who use them as well as those who produce them. For this reason, many production worker cooperatives have refused to make goods that are not ecologically sound or are war-related materials (Linehan & Tucker, 1983). Also, market pressures can exacerbate tensions between the cooperative’s goals of self-management and the need to survive economically in a competitive economy. In addition, being a member of a worker cooperative places great time demands on individuals beyond their working hours. The hours demanded for committee meetings, self-education, and decisionmaking place additional burdens on workers’ time. Such a demanding lifestyle, especially burdensome when the cooperative is newly formed, can severely dampen the appeal of the non-hierarchical workplace (Linehan & Tucker, 1983). Dispute Resolution in Worker Cooperatives Some scholars assert that structural rules, more so than qualities inherent within individuals, promote activism (Pateman, 1970). Pateman argues that
Dispute Resolution in Cooperative and Hierarchical Worksites
45
workers in oligarchic organizations will be apathetic and passive, while workers in organizations which foster participation will respond with greater activism. Pateman maintains that people have a natural desire to control their own destiny, and, therefore, naturally prefer activism over passivity. People lack activism in oligarchic settings when they have not learned the necessary skills through prior participation in democratic organizations. She argues that, despite workers’ natural tendency toward activism, without the necessary skills of democratic participation, they will not demand participation (Pateman, 1970). Additionally, an employer’s dispute resolution procedures can severely limit employees attempts to raise concerns (see e.g., Edelman, Uggen, & Erlanger, 1999; Felstiner, Abel, & Sarat, 1980–1981; Grillo, 1991; Hirschman, 1970; Hodson & Sullivan, 2002; Hoffmann, 2003; Miller & Sarat, 1981; Morrill, 1995; Silbey & Sarat, 1989). For example, in her study of women disputants, Grillo (1991) demonstrated that dispute resolution procedures themselves often heighten the employer’s advantages by inhibiting employees from challenging the employer. Edelman, Christopher, and Howard’s (1999) extensive work on human resource departments in organizations across the country demonstrates that, by symbolizing legality, the employer’s dispute resolution processes provides legitimacy to the employer and her/his practices and diminishes the strength of employees’ rights. Within the organization, this legitimation makes it more difficult for employees to raise internal grievances because the internal dispute process serves to reaffirm the employers’ position as unquestionably correct. Even if employees are aware of problems or would-be grievances, they might be reluctant, wary, or unmotivated to actually bring the grievances forward (e.g., Gaventa, 1980; Hirschman, 1970; Tucker, 1999). Additionally, externally, employees may find it more difficult to convince external agencies or officials that they have been wronged by their employer because of the external legitimation that the internal resolution process strengthens (Edelman, Erlanger, & Lande, 1993). However, dispute resolution in worker cooperatives might have a different dynamic (Hochner et al., 1988). Some researchers emphasize that interest-based, rather than rights-focused, bargaining significantly increases successful grievance handling (Brett & Goldberg, 1983 in Tjosvold et al., 1999). One way interest-based bargaining occurs is when both sides see themselves ‘‘on the same side’’ and working toward the same goals. This perception is sometimes referred to as cooperative interdependence. Tjosvold et al. (1999) identified three types of goal interdependencies: cooperation, when people believe their goals are positively linked; competition,
46
ELIZABETH A. HOFFMANN
when people believe that goal attainment by others diminishes the likelihood of their own goal attainment; and independence, when people believe that their goals are unrelated. Workers with cooperation interdependency expect each person to work hard toward the shared goals; therefore, these workers are more likely to exchange information and support each other and to have stronger interpersonal relationships. Workers in competition are more likely to restrict information, distort communication, and promote their own interests to the detriment of others’ interests. Independent workers act similarly to the workers in competition, but with weaker effects and with greater indifference to fellow workers (Tjosvold et al., 1999). Tjosvold et al. argue that cooperative goals – in contrast to competitive and independent goals – promote open-minded discussions of disputes that result in resolutions that are mutually benefiting to both supervisors and employees. Tjosvold et al. (1999, p. 59) found that employees and supervisors with cooperative interdependence ‘‘were able to negotiate openmindedly to develop integrative solutions that successfully handled their informal grievances’’ while workers in competitive interdependence were ‘‘closed-mouthed, inefficient, dismissed new ideas, and developed solutions that worked against employee interests.’’ The cooperative supervisors and employees were often willing to compromise, assist each other, and work for a successful resolution to problems. Some researchers of worker cooperatives (e.g., Cornforth et al., 1988; Honigsberg, Kamoroff, & Beatty, 1982; Linehan & Tucker, 1983; Mackin, 1997; Rothschild & Whitt, 1986; Thornley, 1981; Tucker, 1999) assert that worker cooperatives might affect the type of interdependence members experience. Specifically, these scholars argue that workers and managers in cooperatives might be more likely to share similar goals and enjoy cooperative interdependence. In contrast, employees and managers in more conventional, hierarchical settings might be more likely to be in competition with or independent from one another, having antagonistic or indifferent relationships with their co-workers.
SAMPLING AND METHODS I used the comparative case method to explore dispute resolution strategies and attitudes (Ragin, 1987). I interviewed and observed workers in three industries: coal mining, taxicab driving, and organic food distribution. Within each industry, I studied a worker cooperative and a matched conventional business. The industries in this study offered a range of workplace
Dispute Resolution in Cooperative and Hierarchical Worksites
47
cultures, gender balances, and business objectives. I visited each business twice, observing as well as interviewing workers and achieving variation in interviewees on many dimensions. The duration of the visits ranged from a few days to two weeks. Sample Over a period of three years, I conducted a total of 128 interviews: 18 at HealthBite Distributors, 35 at Organix Coop, 14 at Private Taxi, 20 at Co-op Cab, and 41 at Coal Cooperative/Valley Colliery. (Coal Cooperative and Valley Colliery were the same physical mine, but under different ownership and management systems, as explained below.) For each site, Table 1 provides summary statistics on the interviewees as well as on the organizations, themselves. I did not identify a specific group of workers whom I knew to have had ‘‘disputes’’ but spoke to all interviewees about their workplace experiences generally. I interviewed a wide variety of workers to maximize the range of problems and experiences as well as the variety of solutions and expectations to be included in this study. My sample included present and former employees as well as managers and worker–managers. Interviewees also differed in terms of length of employment, sex, race, age, level of education, socioeconomic status, and section of the particular business. I am confident that my findings are well triangulated and valid because (1) the interviewees repeated similar themes, voicing comparable statements as earlier interviewees, indicating that the data collection had reached a point of having gathered all perspectives, and (2) I had carefully sampled the workers, ensuring that the study included many different types of workers, Table 1.
Summary of sites and interviewees.
Industry
Valley Colliery Coal Cooperative
Coal mine Coal mine
Private Taxi
Taxicab driving
Co-op Cab
Taxicab driving
HealthBite Distributors Organix Coop
Organic food Organic food
Type of Organization
Location
Conventional Worker cooperative Conventional Worker cooperative Conventional Worker cooperative
Number of Workers
Number Interviewed
Wales (U.K.) Wales (U.K.)
252 239
38** (15%) 41** (17%)
Wisconsin (U.S.) Wisconsin (U.S.) London (U.K.)
120
14 (12%)
150
20 (13%)
32
18 (56%)
Halifax (U.K.)
50
35 (70%)
48
ELIZABETH A. HOFFMANN
and, hence, captured all possible perspectives. Although these interviewees are not statistically representative of all the workers at their individual organizations, the diversity of this sample is helpful in developing conceptual models. Generally, I approached interviewees myself, rather than requesting volunteers to come forward. Sometimes I would approach a group of people, ask to talk with one of them, and schedule interview times with the others. Other times I would approach people who were off by themselves. Since a significant focus of this study is the raising of grievances, interviewing only those inclined to step forward could create an unrepresentative sample of perspectives on grievance behavior. The assertiveness and extroversion necessary to volunteer to be interviewed by a stranger may be correlated with both attitudes on raising grievances and ability to resolve disputes. I arranged certain interviews in advance with key people and workers from underrepresented groups within the organization whom I wanted to be certain to include. Most of the interviews were conducted in public places or in private spaces at the companies themselves. All interviews were conducted in person, using a set of open-ended questions as initial probes on a wide variety of work-related topics. No survey instruments were used, since, following the initial probes, many questions were posed in response to each interviewee’s own statements. The interviewees were drawn from six worksites in three industries: (1) coal mining, (2) taxicab driving, and (3) organic food distribution. The industries ranged from the coal mining industry where the workers are very pro-union, to the non-conformist- or loner-oriented taxicab industry, and the progressively oriented organic food industry. The industries also ranged from having a predominantly male workplace culture, such as coal mining, to being less explicitly gendered, such as the organic food industry. These differences in workplace culture are somewhat reflected in the industries’ different gender balances: 50:1 men to women at the coal mines, 5:1 in taxi driving, and 1:1 in organic food. Table 1 provides a summary of the organizational attributes of each business. All businesses in this study met several key criteria. First, the company needed to have a formal system for grievance resolution. Second, it had to be sufficiently large that a formal grievance system was necessary. For this study, the minimum size of an organization was 30 workers. Third, each business had to be a stable organization with established procedures; none was less than two years old. Fourth, no organization could be part of a larger organization. Additionally, each cooperative included in the study had to be a true worker cooperative – with all employees being equal
Dispute Resolution in Cooperative and Hierarchical Worksites
49
shareholders and no outside shareholders – not merely an Employee Stock Option Plan (ESOP) company. Within each industry, I compared a worker cooperative (non-hierarchical workplace in which all workers are co-managers and co-owners) to a conventional, hierarchical business matched in size and gender ratios. The coal mining and organic food distribution were studied in the United Kingdom (U.K.); the taxicab industry was studied in the United States (U.S.). Because these two cultures are sufficiently similar, no cross-cultural comparison is included in this study. Admittedly, people are more class-conscious in the U.K., but fundamental disputing culture, as seen in the two legal systems, is sufficiently similar (Wheeler, Klaas, & Rojot, 1994). The businesses are summarized in Table 1. I looked at two organic food distributors: Organix Coop, a worker cooperative located in the mid-North of England, and HealthBite, a conventional business located near London. As organic food distributors only, they do not produce any of their products: organic produce or foods made from organic produce and with minimal processing. While some workers in the industry describe the attraction of these jobs as simply the need for a paycheck, others spoke of their dedication to the organic and whole food movement and saw the jobs as a type of activism. Organix Coop was begun over 20 years ago by progressive college students who wanted to create a better, healthier, more egalitarian work environment. This consciousness of the worker cooperative ideology still permeates the business. Workers at Organix Coop became members after completing a probationary period and being voted into membership by the current members. Once they became members, they received their part of the company’s profits, as well as wages, and became ‘‘vested’’ in the company, with each worker owning a single share of stock, regardless of tenure. When they left the cooperative, they would have to sell their share back to the company, generating a type of severance pay. I selected HealthBite Distributors for this study because it was similar to Organix Coop in many ways, including the gender and racial balance of its workforce, the hours of the business, and the business’s focus. HealthBite was formed when two individually owned organics wholesale businesses merged in the early 1990s. One acted as more of a wholesale warehouse store; the other mainly as a distributor to individual homes and organizations. Today, HealthBite is primarily a distribution company, although individuals can come to the warehouse and buy off the skids. These two owners, now partners, share the management of the business in the new London location. Both the conventional taxicab company (Private Taxi) and cooperative taxicab company (Co-op Cab) are located in the same Midwestern town,
50
ELIZABETH A. HOFFMANN
whose alias is Prairieville. Co-op Cab was begun over 20 years ago by cab drivers who were out of work due to strikes at two of the city’s main taxicab companies. Possibly affected by its location in a Big Ten university town, known for its progressive politics (both cab companies reputably had ‘‘overly-educated’’ drivers, often with advanced, even doctoral, degrees (Langway, 1997)), Co-op Cab embraced the worker cooperative ideology in trying to create a better workplace, although not as strongly, uniformly, or dogmatically as Organix Coop above. Workers at the cooperative, Co-op Cab, became members once they had successfully completed a probationary period as determined by the membership committee. Once members, they shared in the profits of the company in addition to their wages. Finally, Valley Colliery and Coal Cooperative were ‘‘deep-pit’’ mines, meaning deep underground mining, as opposed to strip mining. The two coal mines in this study were actually the same physical mine under two very different systems of ownership and management, with interviews for both companies conducted several years after the re-opening of the mine as a cooperative. Therefore, since these miners were only interviewed after the conversion of the mine into a cooperative, some of the data rely on their recollections of the organization three years earlier, when it was owned by British Coal. This gives a high level of credence to the interviewees’ memories. However, this concern is offset by their ability to identify differences before and after the worker buy-out, thus, providing important insights regarding how worker attitude toward raising grievances changed with the conversion into a worker cooperative. Moreover, by studying the same company, this study avoids many comparability problems that arise from comparisons of two or more companies, such as variation in the composition of the workforce, different challenges at the separate work sites, and dissimilar institutional histories. The alias ‘‘Valley Colliery’’ refers to this mine when it was nationally owned by the British Coal Board, while ‘‘Coal Cooperative’’ refers to the mine once it became a worker cooperative. This mine, located in Wales, U.K., was the last deep pit in Wales and one of the few left in the U.K. As such, employment at the mine – both when it was still part of British Coal and after it became a worker cooperative – held important cultural significance for the miners, who deeply identified with the mining occupation. During the period between the closing of the mine by the Coal Board and its reopening as a cooperative, some out-of-work miners took factory jobs, the only other jobs in the area. They described them with much contempt, often saying that they would rather go on government assistance than work there again. Once the mine was re-opened as a cooperative, workers had to
Dispute Resolution in Cooperative and Hierarchical Worksites
51
become members before they could begin work at the mine. In order to participate, each worker had to buy a single share of the cooperative at approximately $13,000. As with the other two cooperatives, this share entitled the member to profit sharing as well as wages. When the worker left the coop, this share would be bought back by the company. Methods One of the key benefits of qualitative studies is the high validity possible: the researcher can understand the greater context, obtain a large overview, and can triangulate the accounts of differently situated interviewees with various bases of knowledge. I employed a qualitative comparative case method (Ragin, 1987) to study three very different industries, each with one cooperative and one conventional organization. In gathering data for this study, I interviewed workers; observed behavior; read related documents and articles; attended companies’ business meetings and, when possible, grievance hearings; and participated in aspects of some businesses (e.g., went down into the coal pit, rode along in the taxicabs). The interviews ranged from 20 minutes to over 5 hours, with most lasting between 20 and 90 minutes. At least one especially lengthy interview occurred at each business. The main focus of the interviews was how the interviewee would handle potentially grievable circumstances. Interviewees were asked mostly general, open-ended questions, but with some direct questions, especially as follow-up inquiries. In discussing grievance resolution strategies, respondents would often draw on examples from their past experiences. All interviews and most site observations were tape-recorded and transcribed, so all quotes used here are direct quotes. These data were analyzed using the qualitative data software NUD*IST Vivo, often referred to only as NVivo. NUD*IST is the acronym of Non-numerical, Unstructured Data: Indexing, Searching and Theorizing. NUD*IST Vivo is the most recent version of this software program from Sage. Using NVivo, I began by coding the transcribed interviews for various themes. Some of these themes were responses to explicit questions (e.g., ‘‘In what ways is your job difficult?’’). However, many others were extracted from the responses of interviewees to broader questions (e.g., ‘‘How would you describe your job?’’ ‘‘How would you recommend/criticize your job to another worker in the same industry?’’ ‘‘What would you change about your job if you could just snap your fingers and it would be different?’’) or to follow up questions to other responses. Thus, a portion of the codes were not the result of a direct
52
ELIZABETH A. HOFFMANN
question or set of questions, but were produced by careful analysis of interviewees’ responses to various questions, as facilitated by using NVivo. The many benefits of using computer-assisted qualitative data analysis software include greater speed in coding and analysis, more complex ways of studying relationships in the data, text searches, and formal structure for developing analysis (Barry, 1998). I used NVivo for cross-case analysis, to gather data together and explore them, and to re-order the hierarchy of codes. Discussions of dispute resolution at other jobs were labeled with ‘‘disputes,’’ with one or more sub-categories, as well as with the category ‘‘other jobs.’’ This allowed me to extract discussions involving only disputes at present jobs and yet also have the option of including reports of past grievance behavior at earlier jobs. The category ‘‘other jobs’’ was also used for any discussion of other positions the interviewee had had; it was used alone, such as for simple descriptions of other jobs, or with other categories, such as with ‘‘management’’ or ‘‘gender,’’ for contrasts on specific aspects with the previous workplaces. I began by coding on very basic concepts: disputes, workplace culture, bad, good, et cetera. I then ‘‘coded on,’’ refining those codings into more specific sub-categories. For example, text coded as ‘‘disputes’’ was broken down into ‘‘formal,’’ ‘‘informal,’’ ‘‘procedures,’’ ‘‘legal consciousness,’’ ‘‘lumping,’’ and ‘‘whining.’’ Some passages of interview text received more than one sub-category. These sub-categories were further refined; sometime breaking them down into even smaller categories and, other times, merging two or more categories. For example, I later collapsed ‘‘lumping’’ and ‘‘whining’’ into one category ‘‘toleration’’ that included both stoic acceptance and vociferous whining; both involved acknowledging a problem but taking no action to resolve it. Thus, the category of ‘‘toleration’’ is different from simply ‘‘silence’’ in that ‘‘toleration’’ encompasses both the unhappy, truly silent employee and the less quiet employee who simply refuses to take an action to resolve the grievance. Using NVivo’s attribute function, I assigned attributes to each interviewee based on her/his company, industry, cooperative/non-cooperative status, and gender. Using NVivo’s matrix and search functions, I compared responses from interviewees within and between companies, industries, and organizational structures. I contrasted their various responses regarding job involvement, dispute resolution, workplace power, gender dynamics, ideological commitment, and workplace socializing. NVivo also can generate tabulations of the number of occurrences of any coded item. For example, it can produce a chart showing how many interviewees at each business mentioned addressing disputes through formal means, informal means, toleration, or exit.
Dispute Resolution in Cooperative and Hierarchical Worksites
53
The reader should note that these tabulations are not generalizable in any way. These tabulations are not from a quantitative random sample, but are descriptive statistics, illustrative of the balance of qualitative evidence.
FINDINGS The data did not show a single clear pattern between the worker cooperatives and the conventional, hierarchical businesses with regard to dispute resolution strategies. Members of the worker cooperatives did not all address their grievances one way, while members of the more hierarchical organizations all used a different grievance strategy. However, when comparing each matched set of businesses (one cooperative to one conventional, hierarchical business) within each industry, a pattern does emerge. Workers in the cooperatives had more dispute resolution strategies at their disposal. In each industry, members of the worker cooperatives described more dispute resolution options than their counterparts in the conventional, hierarchical company. While co-op members often described having both formal and informal strategies at their disposal, their conventionally employed counterparts often mentioned only either formal or informal means. Additionally, employees in the more hierarchical businesses were more likely to simply adopt coping mechanisms (‘‘toleration’’) rather than raise their grievances either formally or informally. In Table 2, I summarize the percentage of workers at each business who discussed each dispute resolution strategy. Table 2.
Percentage of Workers who Described Each Dispute Resolution Strategy.
Valley Colliery Coal Cooperative Private Taxi Co-op Cab HealthBite Distributors Organix Coop
Formal Processes (%)
Informal Processes (%)
Toleration (%)
90 46 14 55 0 74
3 92 43 60 56 69
11 3 36 5 61 29
Notes: Percentages sum to greater than 100% because the categories are not exclusive; some interviewees mentioned more than one grievance resolution strategy. A ‘‘no response’’ category does not exist. All interviewees provided at least one strategy (formal processing, informal processing, toleration, or exit) that they anticipated trying to address workplace problems.
54
ELIZABETH A. HOFFMANN
Below, I describe how this dynamic existed within each industry. I will discuss formal and informal processes together since they both are strategies for directly addressing the problems in the workplace. I will discuss toleration last, and separately, since toleration could be understood as merely a coping strategy, rather than a means for actually resolving disputes. Formal and Informal Processes Coal Mining Miners in the more hierarchical mine, Valley Colliery, only described resolving grievances formally. They explained that informal means were not effective and so were rarely tried. Many employees, such as the one quoted below, explained that raising formal grievances was their only avenue because managers refused to engage in any informal grievance resolution. One of the control room workers explained how relations between managers and workers were so strained that informal negotiation was difficult and unlikely: The manager used to come down and he wouldn’t talk to you. He’d probably tell somebody else who would tell you to do something. They felt they were some super human! We were down there and they were up at the top like. It was all, ‘Do this!’ You know? They tell you rather than ask you. There was no talking to them. [094]
Another miner explained that the response to a dispute could be a walk-out by the miners or a lock-out by the management – both extreme forms of formal action and exercises in official power. [If we had that dispute], we’d probably be going home. That was their attitude: if [Valley management] couldn’t have their own way, they’d send the men home. So they would close the mine down. Yes. Or we would take the decision down ourselves and go home. [050]
Indeed, formal action, whether on the part of the employees or the management, was the only way that disputes were addressed. In contrast, the members of the worker cooperative coal mine, Coal Cooperative, spoke of their ability to resolve grievances both through informal means and through formal grievance procedures. Many issues that had been dealt with only formally before the conversion to a cooperative could now be handled informally. One miner who had been at the mine for seven years before the worker buy-out provided this recent example of informal grievance resolution: It was about a disagreement over work that was left for the weekend that I didn’t think was satisfactory. I channeled it through my foreman who didn’t agree with me. So I then
Dispute Resolution in Cooperative and Hierarchical Worksites
55
took it to the manager and put my case forward. Obviously, he’s had a word with my foreman and I sorted it out with the foreman anyway since. If you don’t bring the point up, it’ll just keep on happening. You find most people around here are like that now. They’ve got their idea, and they put the better idea forward. [153]
The availability of informal means did not mean that the formal procedures were abandoned. Indeed, formal grievance procedures were also used at Coal Cooperative. One electrician, for example, spoke of how issues that would have been deemed not sufficiently important or inappropriate to bring as a formal grievance when the mine was run by British Coal were acceptable to be raised in the cooperative. He described a formal grievance brought soon after the mine reopened as a cooperative: a grievance about the toilet paper. Another thing they wanted changed when we came back as a cooperative was the toilet paper. The toilet paper [the miners used], they were the old government bloody thick paper. A simple thing like that. And the managers, under British Coal, their toilets up there, they had the soft, bloody soft, pink paper. The things like that. Silly little things. But it matters. It says, ‘I’m no better than that manager over there and he’s no better than me.’ [129]
Safety concerns, also, continued to be a key area where workers filed formal grievances. Certain safety risks, if left unattended, could lead to injury, death, or economic loss for the mine. Taxicab Driving In contrast to the coal industry – in which the conventional mine workers could resolve grievances formally but not informally, while those in the cooperative mine could raise disputes both formerly and informally – workers in the more hierarchical taxicab company, Private Taxi, explained that they could raise some disputes informally, but rarely included formal procedures among their dispute–resolution strategies. A Private Taxi driver explained that employees would try whatever means they could, with whomever they could, when they attempted to resolve grievances informally. Everyone goes to whoever they think they’re gonna get some satisfaction from. If someone has a good relationship with [the owner], then they’ll probably go to [him] and see what they can do. If someone has a good relationship with [the manager], at this point in time, they’ll go to [him] to see what they can do. In some cases, people will go to the dispatcher and say, ‘Hey, you know, listen, I got this problem with this guy that’s doing this other shift, and I feel like he’s screwed me over. Or whatever.’ [063]
Each employee at Private Taxi had to negotiate on his or her own, seeking informal resolution without predictable outcomes. Often, this dynamic led many employees to simply learn to tolerate problems, rather than attempting either formal or informal resolution, as discussed later.
56
ELIZABETH A. HOFFMANN
While the Private Taxi workers had barely one means of resolving grievances, the members at Co-op Cab spoke of using both formal and informal ways of resolving workplace disputes. Many members explained that a key advantage of a worker cooperative was that formal grievances were more socially acceptable and easier to raise. For example, a woman who had been at Co-op Cab for about two years said: People aren’t afraid to bring grievances if they feel they’ve got one. We’re encouraged to use the Workers’ Council if we feel that we have a grievance y I think there’s a sort of a sense that there’s very few jobs where you have that opportunity, so make the most of it. [128]
The following quote from a driver and dispatcher who had worked at Co-op Cab for about six years explained that, since he had never had a dispute that he could not resolve informally, he never had had the need to use the formal grievance procedures. I guess my first priority interpersonally, if I had a problem with another employee, would be to work it out with them. If I couldn’t work it out with them I would be in a new kind of situation. I’ve usually been able to work it out. [148]
His statement is typical of most men at Co-op Cab. These men said they were comfortable raising grievances formally or informally, but usually preferred informal routes. In contrast, the woman’s quote, above, is typical of the women at Co-op, who did not feel that informal routes were available to them. Although they did not have access to informal options, the women were comfortable with, anticipated using, and had used, their cooperative’s formal grievance procedures. Much of this dynamic is due to the somewhat sex-segregated socializing at the cab cooperative and the fact that most of the worker– managers, who had the power to informally negotiate grievances, were men. This created an environment in which the women lacked the social networks to enable them to resolve grievances informally (see Hoffmann, 2004; Hoffmann, 2005 for detailed discussion of this dynamic at Co-op Cab). Nevertheless, although not all members of Co-op Cab were able to use both formal and informal means, both strategies for dispute resolution were available to the members. Food Distribution Members of Organix Co-op also talked about both formal and informal dispute resolution strategies. For example, one worker described a recent
Dispute Resolution in Cooperative and Hierarchical Worksites
57
formal grievance: [It’s] just a difference of opinion in some cases. If somebody has just purely a difference of opinion on whatever it might be. That can sometimes turn nasty and can end up being taken to grievance. When you say, ‘Turn nasty’ what do you mean? Well, just maybe each other’s working practices are y I can’t think of an example y Oh, one taken to grievance was over machinery, the way somebody operated machinery and the person didn’t agree with [that person’s] working practice. [125]
Others preferred resolving disputes through informal routes, as this longtime member from Organix Coop explained: I tend not to go to the meetings. I like to chat things up in the pub. That’s just the way I like to do it. [007]
Often workers who did feel able to resolve disputes both informally and formally, expressed a preference for the informal route. As at Co-op Cab, some members of Organix Co-op were more able to use both formal and informal routes than other. Rather than falling along gender lines, at Organic Co-op, members who felt less able to resolve grievances informally were members who were less formally educated, of lower class status, and were non-white (Hoffmann, submitted). However, these lessempowered members, like the women at Co-op Cab, still were very able to bring formal grievances. Nevertheless, despite this inequitable access to informal routes, no Organix Co-op member felt unable to resolve disputes effectively, one way or another, in contrast to workers at HealthBite, the more hierarchical food distribution company. At HealthBite, most workers felt that neither option was very promising. Employees at HealthBite were similar to the employees in the hierarchical taxicab company who most anticipated informal dispute resolution options. In fact, the employees at HealthBite relied even more exclusively on informal routes, with no one anticipating using the formal dispute mechanisms that were available at the company. Indeed, the formal procedures were not an option that many considered, and those who did consider it thought it was not a worthwhile path. For example, one worker from HealthBite mentioned the formal grievance procedures, but emphasizes his belief that this route is ineffective. We’ve got a complaint procedure we go to if we got a complaint about something or anything like that. They try and solve it. But, really, we’re on our own. [086]
58
ELIZABETH A. HOFFMANN
This quotation, representative of others, expresses that, in practice, employees at HealthBite were left to their own informal strategies if they were going to try to resolve problems. One senior woman from HealthBite explained how she engaged in informal grievance resolution on behalf of other workers as well as herself. I’m the strong character, so I pretty much stand up for myself y. There’ve been occasions when there’ve been female members here on some occasions when they felt that there’ve been injustices and they’ve come to me. And I’ve sort of pushed on their behalf because I’ve worked more closely with [the two owners] and maybe I know them better than some of the other staff. And maybe the staff feel they can’t approach the directors or they’re not sure how they’ll take something or whatever. So I’ll quite happily go in there with them or go in there on their behalf and say ‘So-and-so is not happy with this,’ [or] ‘They feel that they should be on a higher rate because of that.’ [143]
As the quote, above, indicates, many employees felt unable to raise grievances formally or informally. These workers would often choose to develop ways of coping with their problems at work, rather than trying to resolving them formally or informally. This is discussed in the following section. Toleration Toleration10 was another way that workers handled disputes. In that ‘‘toleration strategies’’ did not involve actually addressing the disputes at hand, they technically are not ‘‘dispute resolution strategies.’’ Nevertheless, these toleration strategies did provide ways that the workers could continue in their workplaces by learning to cope with the various problems. Sometimes the toleration strategy involved simply saying nothing and swallowing one’s aggravation, while other times workers who engaged in toleration voiced their frustration, but only to uninvolved co-workers. In each industry, employees at the more hierarchical businesses were more likely to talk about dispute strategies that involved toleration (see Table 2). Workers who could not or would not raise their disputes either formally or informally, and who wished to remain at the organization, were left in the position of having to learn to tolerate the problem. Although this is not a strategy to actually resolve disputes, it is a way that some workers dealt with their workplace problems. (For a discussion of the decision to exit, see Hoffmann, 2006.) Taxicab Driving Often having become frustrated with past attempts to address grievances, many workers developed coping strategies, learning ways to adapt to various problems rather than to resolve them. For example, this veteran cab
Dispute Resolution in Cooperative and Hierarchical Worksites
59
driver at Private Taxi explains that now he tries to ‘‘just stay real calm’’ and not let problems bother him. [Now] how do I handle it? I used to complain to the dispatcher. He said, ‘I’ll get you another ride.’ I said, ‘That doesn’t settle nothing!’ I told management about it. They didn’t do nothing about it. They didn’t reprimand the driver or anything like that. So after a couple of years, I just ignored it. If somebody’s stealing my ride, I’d just say, ‘Well somebody else got the ride.’ I just stay real calm because I know it ain’t gonna do no good to complain on it. You can’t get uptight about it. [107]
As were many at the hierarchical cab company, this driver was proud of his ability to ignore potential grievances and to not let such things bother him. Often, workers doubted that their managers would be responsive to any attempts to resolve their grievances, formal or informal, as this cab driver explains below. Unfair sh*t is always going to happen. That’s why it’s good to go out the airport. When you’re waiting for rides at the airport, you can hang out with other drivers and complain about the bad call you got from that dispatcher, or how you didn’t get the shift you were supposed to, or how some other driver cheated you out of ride. You get to get all that stuff out, off your chest. And that’s real good to do, because that’s usually all you can do. [070]
Believing that management would resist their attempts to address their workplace disputes, many of these employees – both in the hierarchical cab company and the hierarchical food distribution company – were resigned to simply tolerate their unraised grievances. Food Distribution For example, an employee at HealthBite explained that often his preferred path is to do nothing. If someone else isn’t doing their work and I’m doing it all, what I’ve learned in the past, is to just shut your mouth and keep doing it. ‘Cause that’s how it works. I just shut up and keep doing it. [059]
His emphasis on simply pushing onward past the problem and accomplishing his work in spite of any potential grievances was echoed by many similarly situated employees. Coal Mining Interestingly, the employees at Valley Colliery present an exception; they seldom mentioned toleration. This is because they were part of the very active N.U.M. miners’ union. As discussed above, these miners were more likely to bring their grievances to their union representatives, rather than learning to quietly tolerate their problems. (For a more detailed discussion of
60
ELIZABETH A. HOFFMANN
this dynamic in the coal industry, see Hoffmann, 2006; for a more information about Valley Colliery and Coal Cooperative, see Hoffmann, 2001a.)
DISCUSSION AND CONCLUSIONS This study compared workplace dispute resolution strategies of workers at two very different types of companies: conventional, hierarchical businesses, and worker cooperatives. The data demonstrate that the cooperative structure and ideology can have an impact on the dispute resolution strategies of its members, enabling them to have more dispute resolution options than their counterparts in conventional businesses. In the coal mining industry, miners at the conventionally organized mine had only formal routes through which to resolve their disputes, while those at the cooperative mine could use both formal or informal avenues. In the taxicab industry, employees at the conventional company most often anticipated resolving disputes informally, and even then, could only occasionally engage in dispute resolution at all. In contrast, members of the taxicab cooperative regularly resolved disputes both formally and informally. In the food distribution industry, workers at the conventional company rarely tried to resolve their disputes – more often relying on their own toleration strategies – but when they did attempt dispute resolution, they worked informally rather than through formal mechanisms. Their counterparts in the worker cooperative were more able to resolve disputes, through both formal and informal means. Interestingly, however, in two of the three cooperatives, a portion of the members were unable to utilize both formal and informal routes of dispute resolutions. While the demographics of members of these less-empowered groups were different between the two cooperatives, this finding suggests that power structures, such as class and gender, present in the larger society surrounding the individual cooperatives also must be considered. The data in this study confirm some of the predictions from the extant literature. Members of the cooperatives had greater access to more avenues to resolve their disputes. As Pateman (1970) argued, in organizations that foster democratic participation, workers’ activism will be greater . Indeed, each cooperative in this study emphasized democratic organization and encouraged workers’ participation in the management of the organization. Through inclusion in work groups, serving on Workers Councils, voting for members of the elected bodies (such as boards of directors), and attendance at regular membership meetings, these workers learned to be participatory members in a workplace democracy.
Dispute Resolution in Cooperative and Hierarchical Worksites
61
This study also strengthens the contention that organizations can work well without an extensive hierarchical power structure, questioning the assertions of Weber, Michels, Hannan et al., and Lipset et al. While this is interesting at a theoretical level, it also is important on an applied level. These findings lend strength to organizational innovations that call for greater employee control of, input into, and ownership of their own work. While few businesses will re-organize themselves into worker cooperatives, many may consider various plans of heightened worker involvement and ownership, but then might dismiss such plans as impractical. This research speaks to that debate and lends credence to these potential workplace innovations. Additionally, this study offers some tentative implications for cooperative interdependence. Workers in cooperative interdependence expect each person to work hard toward the shared goals; therefore, these workers are more likely to exchange information and support each other and to have stronger interpersonal relationships (Tjosvold et al., 1999). Members of the cooperatives – although they would disagree sometimes on various specifics of managing the cooperative or of day-to-day activities – were united in wanting the cooperative to succeed. The realization of cooperative members’ shared goals possibly had the dual effect of enabling members to speak up when they perceived a problem as well as lessening the need to silently tolerate problems or develop coping skills. However, as Pateman also predicted, this worker activism was not automatic (Pateman, 1970). New workers often had to learn how to be active members of a cooperative. The cooperatives were conscious of what they saw as the necessary ideological transition from being an employee to being a worker–owner. The following quote from a woman at the taxicab cooperative illustrates this nicely: We’re also working on a position called head training coordinator y that will make people who are coming in who aren’t from a union shop background or a cooperative background get the idea that if they don’t like something they don’t have to put up with it just because they like their job. There is probably a reason why they don’t like it and it could be fixed. I don’t think people come in understanding that they can speak their minds without being retaliated against. [011]
Additionally, some members of the cooperatives realized that the greater empowerment and opportunities for participation available in their worker cooperatives were not equally accessible by all members. The women at Coop Cab and the working-class, less-educated, and non-white members of Organix Coop were not as able as their co-workers to resolve grievances informally. While the cooperative structure and ideology enabled them to
62
ELIZABETH A. HOFFMANN
use the formal grievance procedures, their difference in social status from their higher-status co-workers left them with less access to the informal power and networks necessary to resolve grievances informally. (For a full discussion of these differences in dispute resolution strategies at Coop Cab and Organix Coop, see Hoffmann 2004, Hoffmann 2005, Hoffmann submitted.) Focusing more on organizational theory, this study showed that workers in cooperatives did not experience as many barriers to dispute resolution as did comparable workers in the same industries whose experiences were similar to those found by other scholars who studied hierarchical businesses (e.g., Edelman et al., 1993; Bumiller, 1988). Unlike most studies of dispute resolution which are situated within conventional, hierarchical organizations, this study revealed that members of cooperatives were less likely to see a grievance as ‘‘unwinnable’’ (Bumiller, 1988) or be stymied by their individual disputant status (Galanter, 1974). Additionally, this research demonstrated that the structure and ideology of the cooperative organization can have an effect on how employees resolve their workplace disputes. Rather than inhibiting grievances – e.g., placing their own needs behind their concern for the greater common good, as some literature suggested (Rothschild & Whitt, 1986; Tucker, 1999) – workers in cooperatives were often able to embrace more dispute resolution strategies. Members of cooperatives had more options at their disposal for resolving workplace disputes than conventional employees in the same industry. However, the industries discussed here are either male-predominant (coal mining and taxicab driving) or gender neutral (food distribution). The findings of this study should be further explored in industries which are comprised mostly of women, perhaps even with mostly women managers, as well as women employees. Some extant literature suggests that women sometimes engage in different dispute resolution strategies – and, indeed, the data from Coop Cab demonstrate that women might be able to avail themselves of all the dispute resolution options that their male co-workers have. Therefore, future research should explore dispute resolution in conventional, hierarchical businesses and matched worker cooperatives in femalepredominant industries.
NOTES 1. Some organizational scholars (e.g., Scott and Zucker) have not examined hierarchy specifically, but rather the conditions that are believed to demand hierarchy,
Dispute Resolution in Cooperative and Hierarchical Worksites
63
questioning whether these conditions themselves are avoidable. In this way, these scholars suggest the possibility of an organization without hierarchy by removing the certainty of the presence of those conditions that are linked with hierarchy. However, none of these theorists specifically addressed the absence of hierarchy as a possibility. 2. Ouchi uses the term ‘‘clan’’ similarly to Durkheim’s (1893/1984) concept of an organic association which resembles a kin network but may not include any biologically related individuals. 3. This could be either via careful selection of new members or socializing of new members. 4. Kanter describes the stresses and struggles of membership in collective and cooperative organizations, but does not explore dispute resolution within these organizations. 5. The closest Swidler comes to addressing dispute resolution is her discussion of informal social control, through teachers’ and leaders’ charisma, attempts at group solidarity, and egalitarian ideology. Actual negotiation and resolution of disputes is not discussed. 6. Rothschild and Whitt explore worker cooperatives, examining such aspects as structural features, internal and external conditions supportive of cooperatives, and worker satisfaction, but not dispute resolution. 7. Iannello addresses decision-making at (1) a consensual, (2) a ‘‘modifiedconsensual’’ organization, and (3) a hierarchical women’s organization. 8. Hacker, in her study of technology and women at Mondragon (Basque cooperatives), does mention some group action taken by women, but includes no discussion of dispute resolution. 9. Conventional businesses are consciously organized hierarchically, with decisions that are top down and information flowing up. Workers lack control over their conditions at work and only highly filtered information reaches the rank and file; managerial decisions are made higher in the hierarchy, not based on local knowledge or hands-on expertise (1997). 10. My category ‘‘toleration’’ is similar to Hirschman’s (1970) concept of ‘‘acquiescence’’ in that the workplace problem is perceived but no action is taken by the employee. Toleration includes both complaining to co-workers and keeping silent. Regarding the behavior of silence, Boroff and Lewin reported that the employees they studied would sometimes react to unfair workplace treatment with silence. Specifically, their study demonstrated that employees who were particularly loyal to their workplace were more likely to ‘‘respond by suffering in silence’’ (Boroff & Lewin, 1997, p. 50). Bryant’s (2003) research also found workers who reacted to potentially negative circumstances with silence. His work on workplace changes suggests that silence is a more constructive response to organizational change and is more likely to advance the worker’s career, than any use of voice.
REFERENCES Arthur, H., Cherlyn S. G., Judith G., Elaine, S., & Eileen, A. (1988). Worker buyouts and QWL. W. E. Upjohn Institute for Employment Research.
64
ELIZABETH A. HOFFMANN
Bailyn, L. (1997). O1_MRKO1_MRKTransforming work: How do we live with it? Perspectives on Work, 1, 10–13. Barry, C. A. (1998). Choosing Qualitative Data Analysis Software: Atlas/ti and Nudist Compared. Sociological Research Online, 3. Boroff, K. E., & Lewin, D. (1997). Loyalty, voice, and intent to exit a union firm: A conceptual and empirical analysis. Industrial & Labor Relations Review, 51, 50–63. Bryant, M. (2003). Persistence and silence: A narrative analysis of employee responses to organisational change. Journal of Organizational Change Management, 17, 578–592. Bumiller, K. (1988). The civil rights society: The social construction of victims. Baltimore: The John Hopkins University Press. Collins, R. (1981). On the Micro-Foundations of Macro-Sociology. American Journal of Sociology, 6, 984–1014. Chris, C., Alan, T., Jenny, L., & Roger, S. (1988). Developing successful worker co-operatives. London: Sage. Cowling, E. (1943). Co-operatives in America: Their past, present and future. New York: Coward-McCann, Inc. Durkheim, E. (1893/1984). In: W. D. Halls (Trans.), The Division of Labor in Society. New York: Free Press. Edelman, L. B., Christopher, U., & Howard, S. E. (1999). The endogeneity of legal regulation: Grievance procedures as rational myth. American Journal of Sociology, 105, 406–454. Edelman, L. B., Erlanger, H. S., & Lande, J. (1993). Internal dispute resolution: The transformation of civil rights in the workplace. Law and Society Review, 27, 497–534. Edelman, L. B., Uggen, C., & Erlanger, H. S. (1999). The endogeneity of legal regulation: Grievance procedures as rational myth. American Journal of Sociology, 105, 406–454. Felstiner, W. L. F., Abel, R. L., & Sarat, A. (1980–1981). The emergence and transformation of disputes: Naming, blaming, claiming y. Law and Society Review, 15, 631–654. Galanter, M. (1974). Why the ‘Haves’ come out ahead: Speculations on the limits of legal change. Law & Society Review, 9, 95–127. Gaventa, J. (1980). Power and powerlessness. Urbana, IL: University of Illinois Press. Grillo, T. (1991). The mediation alternative: Process dangers for women. Yale Law Journal, 100, 1545–1610. Hacker, S. (1989). Pleasure, power and technology: Some tales of gender engineering and the cooperative workplace. Boston: Unwin Hyman. Hannan, M. T., & Freeman, J. (1989). Organizational ecology. Cambridge: Harvard University Press. Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to declines in firms, organizations, and states. Cambridge: Harvard University Press. Hodson, R., & Sullivan, T. A. (2002). The social organization of work. Belmont, CA: Wadsworth/Thomson Learning. Hoffmann, E. A. (2001a). Confrontations and compromise: Dispute resolution at a worker cooperative coal mine. Law & Social Inquiry, 26, 555–596. Hoffmann, E. A. (2001b). Compromise, confrontation, and coercion: Formal and informal dispute resolution in cooperative and hierarchical worksites. Doctoral Dissertation, University of Wisconsin–Madison, Madison, WI. Hoffmann, E. A. (2003). Legal consciousness and dispute resolution: Different disputing behavior at two similar taxicab companies. Law & Social inquiry, 28, 691–715.
Dispute Resolution in Cooperative and Hierarchical Worksites
65
Hoffmann, E. A. (2004). Selective sexual harassment: How the labeling of token workers can produce different workplace environments for similar groups of women. Law and Human Behavior, 28, 29–45. Hoffmann, E. A. (2005). Gender differences in dispute resolution strategies: Procedural justice at a worker cooperative. Law & Society Review, 39, 51–82. Hoffmann, E. A. (2006). Exit and voice: Organizational loyalty and dispute resolution strategies. Social Forces, 84, 2313–2330. Hoffmann, E. A. (submitted). Workplace dispute resolution and the impact of official and unofficial power. American Sociological Review. 1–42. Honigsberg, P. J., Kamoroff, B., & Beatty, J. (1982). We own it: Starting and managing coops, collectives, and employee owned ventures. Laytonville, CA: Bell Springs Publishing. Iannello, K. P. (1992). Decisions without authority: Feminist interventions in organization theory and practice. New York: Routledge. Kanter, R. M. (1972). Commitement and community. Cambridge: Harvard University Press. Langway, L. (1997). The best cities for women. Ladies Home Journal, November, 199–204. Linehan, M., Tucker, V. (Eds.), (1983). Workers’ cooperatives: Potential and problems. County Cork, Ireland: Litho Press Co: Midleton. Lipset, S. M., Trow, M. A., & Coleman, J. S. (1956). Union democracy. New York: Free Press. Mackin, C. (1997). Employee ownership and industrial relations. Perspectives on Work, 1, 23–31. Marx, K., & Engels, F. (1986). Selected works: Their essential thinking in philosophy, political science, history, social change and communism. New York: International Publishers. Michels, R. (1981). Oligarchy. In: O. Grusky & G. A. Miller (Eds), The sociology of organizations. New York: The Free Press. Miller, R. E., & Sarat, A. (1981). Grievances, claims, and disputes: Assessing the adversary culture. Law and Society Review, 15, 525–561. Morrill, C. (1995). The executive way. Chicago: University of Chicago Press. Ouchi, W. G. (1980). Markets, bureaucracies and clans. Administrative Science Quarterly, 25. Pateman, C. (1970). Participation and democratic theory. London: Cambridge University Press. Powell, W. W., & DiMaggio, P. J. (1991). The new institutionalism in organizational analyis. Chicago: The University of Chicago Press. Putterman, L. (1982). Some behavioral perspectives on the dominance of hierarchical over democratic forms of enterprise. Journal of Economic Behavior and Organization, 3, 139–160. Ragin, C. C. (1987). The comparative method: Moving beyond qualitative and quantitative strategies. Berkeley: University of California Press. Rock, C. (1991). Workplace democracy in the United States. In: J. D. Wisman (Ed.), Worker empowerment: The struggle for workplace democracy (pp. 37–58). New York: The Bootstrap Press. Rothschild, J., & Whitt, J. A. (1986). The cooperative workplace: Potentials and dilemmas of organizational democracy and participation. Cambridge: Cambridge University Press. Scott, W. R. (1987). The adolescence of insitutional theory. Administrative Science Quarterly, 32. Scott, W. R. (1992). Organizations: Rational, natural and open systems. Prentice-Hall: Englewood Cliffs, NJ. Silbey, S., & Sarat, A. (1989). Dispute processing in law and legal scholarship: From institutional critique to the reconstruction of the judicial subject. Denver University Law Review, 66, 437–487. Swidler, A. (1979). Organizations without authority: Dilemmas of social control in free schools. Cambridge, Mass: Harvard University.
66
ELIZABETH A. HOFFMANN
Thornley, J. (1981). Workers’ cooperatives: Jobs and dreams. London: Heinemann Educational Books. Tjosvold, D., Motohiro, M., & Belsheim, J. A. (1999). Complaint handling on the shop floor: Cooperative relationships and open-minded strategies. The International Journal of Conflict Management, 10, 45–68. Tucker, J. (1999). The therapeutic corporation. New York: Oxford University Press. Weber, M. (1946). From Max Weber: Essays in sociology. New York: Oxford University Press. Wheeler, H. N., Klaas, B. S., & Rojot, J. (1994). Justice at work: An international comparison. The Annals of the American Academy of Political and Social Science, 536, 148–161. Whyte, W. F., Tove, H. H., Christopher, B. M., Reed, N., & Stern, R. N. (1983). Worker participation and ownership. Ithica, NY: ILR Press. Zucker, L. G. (1977). The role of institutionalization in cultural persistence. American Sociological Review, 42, 726–743.
HUMAN RESOURCE PRACTICES, UNIONIZATION AND THE ORGANIZATIONAL EFFICIENCY OF FRENCH INDUSTRY Hristos Doucouliagos and Patrice Laroche ABSTRACT Organizational performance improves through several channels, including changes in efficiency, innovation and technological change. Most of the extant research has focused on overall performance, often measured by partial measures of productivity, with little attention given to the components of performance. The aim of this paper is to analyze the impact of HR practices and unionization on one important channel – organization efficiency - as measured by technical and scale efficiency. Using French industry survey data, the paper shows that HR practices do influence efficiency, but this is moderated by the existence of unions. The results show a rather complex set of associations. We find robust results that show that in France, HR practices have a positive effect on scale efficiency but this effect is dampened in the presence of unions. On their own, HR practices have no effect on technical efficiency. However, some of the results suggest that HR practices can exert a positive influence when combined with unions.
Advances in Industrial and Labor Relations, Volume 15, 67–102 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15001-5
67
68
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
1. INTRODUCTION Recently, attention has turned to human resource (HR) practices that emphasize job enrichment, training and development, performance-based pay and employee empowerment, in order to enhance employee skills, participation and motivation, as well as to maximize employees’ commitment to the organization (Godard, 2004). Although different terms have been used such as ‘‘high-involvement’’ (Lawler, 1991), ‘‘high-commitment’’ (Arthur, 1994), ‘‘high-performance’’ (Huselid, 1995), or simply innovative HR practices (MacDuffie, 1995), they all aim to promote efficiency and productivity by creating incentives and by stimulating creativity. A large body of literature deals with the impact of high-involvement work practices on firm performance, especially in the United States (for recent reviews, see Wood, 1999; Boselie, Dietz, & Boon, 2005). The existing literature consists of two broad groups of studies: (a) studies that focus on the effects of specific work practices, such as training, compensation, employee selection and information sharing; and (b) research that examines the influence of systems or ‘‘bundles’’ of such practices on firm performance (for a recent review, see Godard & Delaney, 2000). Most of these studies have found empirical links between the use of HR practices and productivity. A recurrent result is that single or isolated HR practices have a limited effect on productivity, and that a combination of practices has a larger positive impact when they operate in a complementary way (see MacDuffie, 1995; Ichniowski, Shaw, & Prennushi, 1997; Appelbaum, Bailey, Berg, & Kalleberg, 2000). Despite a voluminous literature, there are still many gaps in our understanding of the process through which HR practices impact on firm performance. Existing studies suffer from three important limitations. First, studies tend to focus on the overall effects of HR practices. In many studies, the term ‘performance’ is used loosely to summarize a complex set of associations and to infer the net effect of several dimensions of performance. While interest usually lies on the net overall effect, analyses of individual components can better inform management and policy making. One of the more important channels is efficiency – the ability to produce the maximum attainable output given the available resources. Second, only a limited range of countries have been investigated (mainly the US and UK).1 Third, important interactions are often ignored, especially those between unions and HR practices. The aim of this paper is to address these noticeable lacunas in the literature with respect to efficiency. While several studies have been conducted on the impact of unions on efficiency (e.g. Byrnes, Fare, Grosskopf, &
Human Resource Practices, Unionization and the Organizational Efficiency
69
Lovell, 1988; Bronars, Deere, & Tracy, 1994; Cavalluzzo & Baldwin, 1993; Dickerson, Geroski, & Knight, 1997; Torii, 1992; Torii & Caves, 1992), we are unaware of prior research into the effects of HR practices on efficiency. Accordingly, in this paper we explore the impact of HR practices on technical as well as scale efficiency. This sheds light on one of the important channels through which HR practices influence performance. By using French data we are able to see the impact of innovative-work practices in other parts of the developed world. Additionally, our particular focus is on the interaction between unions and HR practices in French industry. Unions and HR practices can moderate the influence each has on performance and it is thus important to investigate this moderating effect. The paper is structured as follows. Section 2 presents a brief review of the theoretical arguments linking HR practices, unions and firm performance. Section 3 discusses the data and empirical methodology. The main results are presented and interpreted in Section 4, with sensitivity analysis reported in Section 5. Section 6 concludes the paper.
2. THEORETICAL ARGUMENTS Productivity and efficiency have a clear and precise meaning in economics. However, these terms are used often very loosely, especially in the media and by non-economists. Productivity is the relationship between outputs and inputs, and is measured as the ratio of output to inputs. Productivity improvements occur when this ratio rises. This can occur through technological change and change in technical and scale efficiency. The degree of technical efficiency refers to actual output levels in relation to what is achievable. Technical efficiency improvements occur when output levels more closely approximate what is known as the best practice frontier. The frontier is the maximum output of a given quality that can be produced given the available input levels, skills and technology.2 When technical efficiency improves, an organization is producing more output with the same level of inputs. Hence, ceteris paribus, its productivity rises. Technological change occurs when the frontier expands outwards. Technological change enables firms to produce more output with the same level of inputs, even if that firm is fully technically efficient. Hence, technological change also improves productivity. Technological change can, of course, occur even when firms are not fully efficient. It is possible for firms to find that over time technological change increases their productivity, while their efficiency is deteriorating and hence having a depressing effect on
70
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
productivity. These forces can move in opposite directions or in the same direction. Their net effect determines productivity growth. Scale efficiency refers to the scale of a firm’s operations or the scale of the factors employed. Firms can be 100% technically efficient (they produce the maximum output given their inputs) but the output level can still be low – they are scale inefficient. Optimal scale occurs with constant returns to scale (see Coelli, Rao, & Battese, 1998). At any given point of time, productivity (the ratio of output to inputs) is maximized when both technical and scale efficiency are at their maximum. Where there is either technical or scale inefficiency, then productivity is lower than it could potentially be. HR practices may impact on technical efficiency, technological change, as well as scale efficiency. HR practices may contribute to an outward shift in the best practice frontier. They can also impact on a firm’s distance from any given frontier (technical efficiency). In our analysis we are limited by the cross-sectional nature of the data (discussed below). Hence, we are unable to explore shifts in the frontier over time and whether HR practices influence these. The available data does, however, enable the exploration of both technical and scale efficiency. If HR practices improve efficiency, the firm is producing more output with the same inputs, and hence we can conclude that innovative-work practices led to higher productivity levels. If firms operate with increasing returns to scale, it is desirable for them to expand their scale of operations.3 Doing so means that output rises at a faster rate than inputs and hence productivity rises. It is possible that unions and HR practices exert their impact on scale and technical efficiency in different directions – increasing one, while decreasing the other. It may, for example, be the case that HR practices improve technical efficiency, but at the same time HR practices may be associated with sub-optimal manning levels and hence lead to a deterioration in scale efficiency, or they may induce sub-optimal investment in capital. It follows that an investigation of employee relations on performance should consider technical efficiency, as well as scale efficiency (and technological change if the data are available). Hence, it is important to explore also whether HR practices and unions affect productivity through scale efficiency effects. 2.1. Economic Impact of HR Practices In the literature, it is argued generally that high-involvement work practices have a positive impact on performance. However, the way in which HR practices affect productivity remains the object of much speculation in
Human Resource Practices, Unionization and the Organizational Efficiency
71
economics and (HR) management. According to the existing literature, three approaches can explain the impact of HR practices on productivity. First, it is stated that firms can increase their performance if they use the best HR practices (Wright & Boswell, 2002). This universalistic perspective claims that some HR practices are always better than others. Within this school of thought, several theories have been proposed to explain the link between individual HR practices and firm performance. For example, Kato and Morishima (1995) argue that human capital effects can explain the relationship between HR practices and productivity. Drawing upon human capital theory, Ichniowski (1990) shows that a HR practice such as training, led to greater employee competence, and hence, improved productivity. The best HR practice approach has been criticized on several grounds. One of these reasons is that there is a much more complex relationship between HR practices and organizational performance. For example, some HR practices are more suitable under certain conditions and less suitable under others (Delery & Doty, 1996). According to the contingency approach, firms can perform better if HR practices are consistent with other aspects of the organization (Arthur, 1992). For example, a firm’s strategy necessitates certain behavioral requirements for success and the use of HR practices can reward or control employee behavior. A third approach states that firms can perform better if there are complementarities both among HR practices and the firm’s strategy. For example, Kato and Morishima (1995, p. 19) note that: ‘‘information sharing, which induces employees’ cooperative behavior is not likely to be effective over a long haul in the absence of tangible rewards, since employees may lose interest in being cooperative and reduce their loyalty.’’ This configurational perspective is concerned with how patterns of multiple HR practices improve firm performance. The implication is that work practices should be thought of as a part of a system of HR practices or a ‘bundle’ that fit together in a coherent manner. This notion of bundles does not mean that a firm simply has to increase the number of innovative practices. The change process is much more complex. The idea is to generate a synergistic effect, which in turn depends on the internal fit (or horizontal fit) between the various HR practices and on the external fit (or vertical fit) between HR practices and other organizational characteristics such as firm’s strategy (Becker & Gerhart, 1996). Most of the US research has found a positive relationship between HR practices and productivity (see, for example, MacDuffie, 1995; Huselid et al., 1997; Ichniowski et al., 1997; Becker & Huselid, 1998; Ichniowski & Shaw, 1999; Guthrie, 2001). In France, relatively little empirical research
72
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
has been undertaken on the links between HR practices and productivity. Allouche’s (1984) results suggest that it is possible to identify a generally positive relationship between firm performance and certain HR practices. Bayad and Liouville (1998) show that HR practices such as compensation, training expenses or employment policies are positively associated with productivity in small and medium French manufacturing companies. Using the French Company Personnel Report, d’Arcimoles (1997) found significant correlations between HR indicators and productivity. For example, d’Arcimoles indicates that social expenditures and high training expenses have a positive impact on productivity in French firms. However, most of these studies do not test if bundles have a stronger impact on productivity than isolated practices. Recently, Guerrero and Barraud-Didier (2004) deal with the impact of a bundle in comparison with isolated HR practices in France. The authors find that: ‘‘when high-involvement practices are combined through a latent variable, they have stronger impact on performance than when they are studied in isolation’’ (Guerrero & Barraud-Didier, 2004, p. 1423). This result provides strong evidence in favor of a complementary system of work practices in the French context. 2.2. The Economic Impact of Unions The relationship between unions and productivity has attracted considerable attention from scholars in industrial relations and economics. There is a very large theoretical and empirical literature on unions and productivity. Turnbull (2003), Addison and Belfield (2004) and Hirsch (2004) are just some of the excellent recent reviews of this area. Traditional economic analysis states that unions distort relative wages, impose employment restrictions, and contribute to aggregate as well as sectoral unemployment. In addition, unions can have adverse effects on R&D, innovation and technological change and, hence, they can depress total factor productivity growth. Freeman and Medoff (1984) argue that in addition to these effects, unions can raise productivity, by opening up communication channels between workers and management, and by inducing managers to alter methods of production and to adopt more efficient policies. The net effect may be that unions increase productivity. Critics respond to this by pointing out that even if unions do increase productivity levels, the factor price effects can still reduce profits. Empirical findings are divided between positive and negative union-productivity effects, and many studies cannot reject the hypothesis of a zero effect. A recent meta-analysis of 73 studies of productivity effects of unions found a near zero effect (Doucouliagos & Laroche, 2003a).4
Human Resource Practices, Unionization and the Organizational Efficiency
73
The evidence on the economic effect of unions in France is not as developed as it is in the US. In considering the impact of unions in France, one must bear in mind that the effectiveness of voice provision by French unions is weak. First, since French unions have a small number of members, union activity within the workplace is likely to be weak. Second, union voice is fragmented at the French workplace because of multiple union representation on site and because each union competes for the same employees with the same preferences and job attributes. Third, a further concern in France is the incidence of labor law and public policy within the industrial relations system. Incentives for unions to act in close accord with the needs and preferences of the workers they represent are weakened further by laws that allow collective agreements in the workplace even if only one union representing a minority of workers signs the agreement. Two researchers have explored the effects of unions on labor productivity in the French setting (Coutrot, 1996; Laroche, 2004b). Their empirical results suggest that the impact of French unions on productivity is positive but weak and that the union wage effect is difficult to assess. 2.3. Unions and HR Practices Interactions In addition to their separate effects upon performance, there may also exist important interactions between unions and HR practices. The role of unions in supporting or inhibiting the introduction of HR practices has been the subject of much debate in the US context (Verma & McKersie, 1987; Eaton & Voos, 1989, 1992; Kizilos & Reshef, 1997; Reshef, Bemmels, & Wolfe, 1993; Reshef, Kizilos, Ledford, & Cohen, 1999). Some authors have suggested that the presence of unions in a workplace constrains the ability of management to develop HR innovation. In particular, when innovation is perceived as a threat to existing union influence, indirect pressure may be placed on the employer to abort the planned innovation (Kizilos & Reshef, 1997). In contrast, many experts have concluded that unions participate constructively in the introduction of new organizational systems. Indeed, Eaton and Voos (1992) see unions as playing a positive role in the successful introduction of new work practices in US establishments. If unions perceive workplace innovation as favoring their influence, they will likely support and facilitate it.5 Although there has been much empirical research on the effects of HR practices on productivity and on the effects of unions on productivity, there has been relatively little empirical work on the effects of both HR practices and unions on workers’ productivity (see Delaney, 1991).6 One of the first empirical studies on the combined effect of HR practices and unions on
74
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
productivity was by Cooke (1994). Cooke examined the impact of unions on the effectiveness of employee-participation programs and group-based incentives in US manufacturing firms. He found that, in general, unionized firms provide a better environment for realizing the benefits of employeeparticipation programs than do non-unionized firms. However, according to Cooke, incentive pay schemes such as profit-sharing contribute substantially more to performance in non-unionized firms than in unionized firms. Using US data from a nationally representative sample of businesses, Black and Lynch (2001) also examined the combined effect of workplace practices and unions on productivity. The authors estimated an augmented Cobb–Douglas production function with both cross-section and panel data covering the period 1987–1993. Their results indicate that, among highperformance work systems, only benchmarking is positively and statistically associated with labor productivity. Regarding employee voice practices, only the proportion of workers meeting regularly in groups is positively and statistically related to productivity. Unions are not associated with labor productivity in cross-sectional estimations and are negatively and statistically associated with productivity in the panel estimation. Contrary to Huselid (1995) and Ichniowski et al. (1997), the authors report no evidence of a synergistic bundling of workplace practices, although there is a positive and statistically significant relationship between unions and non-managerial profit-sharing and productivity. Finally, Black and Lynch (2001, p. 434) find that: ‘‘it is not whether an employer adopts a particular work practice but rather how that work practice is actually implemented within the establishment that is associated with higher productivity’’. MacNabb and Whitfield (1997) estimate the relationship between unions, HR practices and firm performance in the British context. However, they measured financial performance rather than productivity or efficiency. Using data from the 3rd Workplace Industrial Relations Survey held in Great Britain, their results indicated that the presence of a recognized union is beneficial to the introduction of innovative work practices. They also showed that the joint effect of union presence and new workplace practices on financial performance is positive, even if the single effect of union presence is negative. According to MacNabb and Whitfield (1997, p. 826), it is possible: ‘‘that an analysis examining just one of the factors could be biased due to the omission of the other. It is possible, for instance, that the coefficient on the union variable in union-only equation is capturing the joint influence of union presence and, say, the lower propensity of unionized firms to introduce new work practices. Second, there may be a joint effect of union presence and new work practices over and above their individual
Human Resource Practices, Unionization and the Organizational Efficiency
75
influences’’. The union-HR interactive effects are likely to depend on whether unions simply coexist or whether HR practices can be an anti-union strategy (Wood, 1999). Fiorito (2001) argues that HR practices can displace unions by offering a competing form of representation. In such cases, we would not expect a positive interaction effect. Machin and Wood (2005) review previous studies for the UK as well as offering their own empirical analysis and find no evidence of a union-HR substitution process in the UK.7 In such cases, unions and HR practices may work together. In summary, the empirical literature supports the view that the impact of unions on performance is mixed, and that HR practices typically have a positive effect on productivity. There are obviously many research issues that warrant additional investigation. The focus in the remainder of this paper will be to explore the impact of unions, HR practices and the interaction between unions and HR practices on the efficiency aspects of organizational performance in French industry.
3. DATA AND EMPIRICAL METHODOLOGY 3.1. Data The data used in our analysis are derived from the second REPONSE Survey (Relations Professionnelles et Ne´gociations d’Entreprises). REPONSE was conducted in 1998 and was based on a nationally representative sample of 2,978 French establishments with more than 20 employees.8 The REPONSE survey collected information from managers and union representatives. This survey contains detailed information on the establishments, their organizations, workplace practices and the environment in which they operate. The REPONSE survey was matched with the financial accounts of the same establishments obtained from another database (DIANE).9 From the 2,978 establishments, we used a sample of 2,831 establishments for which all the necessary data were available to calculate efficiency. Our dataset includes 9% from the food sector, 6% from health, 13% from consumer goods, 25% from commerce, 17% from equipment manufacturing, with the remainder from various other industries. Several studies have already used this database. This dataset has been used previously by Coutrot (2000) to examine the determinants of work innovations in French workplaces. Doucouliagos and Laroche (2003b) used the same dataset to investigate the impact of unions on capital investment. Their results indicate that French unions in general do not have a negative
76
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
impact on investment behavior, which is in contrast to findings for the US. In a set of recent studies, Laroche (2004a, 2004b) also used this database to examine the relationship between union presence and corporate financial performance in France. His results suggest that French unions have no impact on firm profits. This reflects the particular French industrial relations context. Taken together, these studies indicate that unions in France have a positive effect on productivity, and no adverse effects on either capital formation or financial performance. 3.2. HR Practices in France In the early 1980s, French management implemented several components of high-involvement HR practices associated with the new American managerialism. Some firms modernized and introduced workplace practices such as formal participation processes, seeking to win the cooperation of their employees. Nevertheless, in 1999, more than half of all workplaces with 10 or more employees (employing nearly one in five French workers) had no employee representatives (Ruelland, 2001). In France’s large firms, unions remain well entrenched and managers have to accept the political and legal obligations of a pluralistic work environment (such as union representatives and works councils) that can restrict their room to maneuver. Around 40% of French firms continue to work along traditional hierarchical lines and exercise control over their workers through authoritarian methods. They are largely small-sized firms (Coutrot, 1998). A further third – largely medium-sized – have embraced new technology to control the work process without any resistance. Among the larger firms, we can distinguish two main work regimes: about a fifth of all workers were subjected to work process where their degree of autonomy is structured by the production system. Within the HR literature these firms would be considered as implementing ‘hard’ HR strategies, focusing essentially on numeric flexibility. Only around one in six French workers is offered a work environment with some real degree of autonomy, that is a ‘soft’ HR strategy (Coutrot, 2000). 3.3. Measurement of HR Practices In this study, we used 17 HR practices which are commonly included in prior empirical research (MacDuffie, 1995; Ichniowski et al., 1995; Osterman, 1994). The respondents were asked to report what percentage of non-managerial and managerial employees have jobs that: (a) are subject to a formal job analysis process; (b) received formal performance appraisal;
Human Resource Practices, Unionization and the Organizational Efficiency
77
(c) received individual wage increases; (d) received performance bonus; (e) participate in total quality management; (f) participate in quality circles; (g) participate in job rotation programs; (h) participate in autonomous work teams; (i) are included regularly in information sharing programs (e.g. regular meetings, newsletter, etc.); (j) participate in co-decision (3 different measures); and (k) have access to a formal complaint resolution system (employee suggestion boxes). Descriptive statistics for each of these variables are presented in Table 1. In order to group the practices more comprehensively, we applied factoranalysis to each item’s standard score, using principal-component extraction with varimax rotation. Several studies have used this approach (Huselid, 1995; Katz et al., 1987).10 Table 2 reports the five factors that emerged from the factor analyses.11 The first factor was labeled ‘‘HR Development’’. It regroups practices aimed at improving and maintaining the knowledge, competence and skills of the workforce. Training plan for non-managers and for managers are geared towards this goal. Performance appraisal could serve to update and maintain competence and skills, providing employees regular information on their performance and on the changing work conditions. Table 1.
Descriptive Statistics for High-Involvement HRM Practices (n ¼ 2,831).
High-Involvement HRM Practices Performance appraisal for non-managers (0, 1) Performance appraisal for managers (0, 1) Training plan for non-managers (0, 1) Training plan for managers (0, 1) Individual wage increases for non-managers (0, 1) Individual wage increases for managers (0, 1) Performance bonuses for non-managers (0, 1) Performance bonuses for managers (0, 1) Profit-sharing (0, 1) Labor-management committee (0, 1) Information sharing about market position (0, 1) Information sharing about employment growth (0, 1) Information sharing about competitive strategy (0, 1) Autonomous work teams (0, 1) Project teams (0, 1) Employee suggestion boxes (0, 1) Quality circles (0, 1) Source: REPONSE database.
Mean
Standard Deviation
0.47 0.55 0.45 0.59 0.76 0.66 0.51 0.56 0.48 0.23 0.57 0.44 0.48 0.19 0.24 0.27 0.57
0.50 0.50 0.50 0.49 0.43 0.47 0.50 0.50 0.50 0.42 0.49 0.50 0.50 0.39 0.43 0.45 0.49
78
Table 2.
HR Practices Factor Analysis Results: Principal Component, Varimax Rotationa.
High-Involvement Work Practices
Eigen value Proportion of variance accounted for
Factor 2
Factor 3
Factor 4
Factor 5
HR development
Performance-based pay
Information sharing
HR participation
Team-working
0.77 0.74 0.79 0.74 0.12 0.20 0.07 0.17 0.11 0.12 0.05 0.13 0.11 0.06 0.02 0.00 0.12
0.13 0.31 0.03 0.25 0.68 0.73 0.61 0.73 0.46 0.11 0.02 0.05 0.05 0.19 0.17 0.10 0.08
0.08 0.05 0.14 0.09 0.02 0.05 0.02 0.02 0.16 0.79 0.77 0.74 0.01 0.02 0.16 0.00 0.16
0.01 0.09 0.05 0.11 0.02 0.02 0.03 0.03 0.23 0.08 0.08 0.02 0.67 0.67 0.55 0.07 0.10
0.03 0.08 0.00 0.06 0.12 0.15 0.10 0.00 0.21 0.05 0.04 0.07 0.11 0.23 0.14 0.75 0.66
1.84 10.8%
1.33 7.9%
3.82 22.4%
1.25 7.4%
1.00 5.8%
a Practices are classified according to their link factor. Bold type coefficients indicate a practice that is conceptually linkable to one of the factors. The associated practices load at 0.40 or greater on one of the factors.
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Performance appraisal for non-managers Performance appraisal for managers Training plan for non-managers Training plan for managers Individual wage increases for non-managers Individual wage increases for managers Performance bonuses for non-managers Performance bonuses for managers Profit-sharing Information sharing about market position Information sharing about employment growth Information sharing about competitive strategy Employee suggestion boxes Quality circles Labor-management committee Autonomous work teams Project teams
Factor 1
Human Resource Practices, Unionization and the Organizational Efficiency
79
The second factor can be summarized under the term ‘‘performance-based compensations’’ because the practices correlated with this factor regroup compensation policies such as profit-sharing, performance bonuses for non-managers and managers, and individual wage increases for both nonmanagers and managers. The third factor regroups HR practices aimed at ‘‘sharing information’’ with employees. It regroups information sharing about market position, employment growth and competitive strategy. The fourth factor is related to what we called ‘‘HR Participation’’. It regroups practices aimed at encouraging employee participation and involvement in the attainment of the objectives of the establishment, as well as the work group. Suggestion boxes and team-working are forms of participation that require employee input to improve the performance of the establishment. The last factor is related to ‘‘team working’’ (or teams for short), and includes practices such as autonomous work teams and project teams. In addition to the five factors, we constructed an aggregate HR practices index (HRINDEX1) by simply counting the number of HR practices existing in a firm. This consists of the 17 variables, so that values for this variable range from 0 to 17. This is a rather naive index and is included to test the sensitivity of the results. A second aggregate measure was constructed as a binary variable that takes the value of 1 if the firm has more than 10 HR practices and zero otherwise (HRINDEX2). This variable identifies firms where there is a greater incidence of HR practices. A third aggregate measure was constructed by combining the HR practices into one index reflecting a bundle of high-involvement HR practices (HRINDEX3). We obtained this index by summing the five factor scores derived from the exploratory factor analysis. As Youndt, Snell, Dean, and Lepak (1996, p. 849) noted: ‘‘this approach is conceptually and empirically better than a multiplicative approach to creating HR systems because it does not reduce the index value to zero if a single HR practice is absent from a system’’. Our focus in this paper is to explore whether HR practices impact on organizational efficiency. In order to partial out the contributions of HR to efficiency, it is necessary to control for other factors that may also impact on efficiency. Like all empirical studies, we are constrained by the availability of data on the choice of control variables. We are constrained also by lack of theory as to what the determinants of efficiency might be. Descriptive statistics for the control variables used in this paper are listed in Table 3. The number of employees is included as a proxy for firm size. We have no prior expectations regarding this variable. While on the one hand, smaller firms may lack the managerial skills and resources to operate efficiently, on the
80
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Table 3.
Descriptive Statistics for Control Variables.
Variable Size/100: The number of employees. New technology: A dummy variable with a value of 1 if a firm has introduced a new technology in the last 3 years and 0 otherwise. Organizational change: A dummy variable with a value of 1 if a firm has introduced organizational change in the last 3 years and 0 otherwise. Union: A dummy variable with a value of 1 if a union representative is present and 0 otherwise. Strategy: A dummy variable with a value of 1 if the firm has a differentiation strategy and 0 otherwise (i.e., a cost reduction strategy).
Mean
Standard Deviation
1.54 0.21
4.58 0.43
0.37
0.50
0.71
0.46
0.84
0.37
other they may not be as prone to coordination problems and the complexity of decisions faced by larger firms. The technical and organizational change variables are included because they influence the likelihood that HR practices are adopted in any particular firm. Moreover, both technical and organizational change can also impact on productivity (Osterman, 1994; Michie & Sheehan, 1999). A measure of the business strategy adopted by the firms was also included as a control variable. The typology developed by Porter (1980) is well known and has been widely used by previous studies examining the influence of HR practices on firm performance (e.g. Fombrun et al., 1984; Schuler & Jackson, 1987; Arthur, 1992; Delery & Doty, 1996). This includes the cost reduction strategy and the differentiation strategy. The link between these strategies and HR practices stems from differences in the uncertainty of the production process. For example, the characteristics of a cost reduction strategy are tight controls and pursuit of economies of scale, whereas a differentiation strategy may be considered as a strategy with more flexible technology to produce a broader range of specialized products. The implications for HR practices is that it involves selecting high skilled workers, giving employees more discretion and using minimal managerial control. Firms with a differentiation strategy are more likely to implement high-involvement HR practices. Consequently, it can be hypothesized that firms pursuing a differentiation strategy are more likely to obtain higher performance.
Human Resource Practices, Unionization and the Organizational Efficiency
81
The union variable is defined as a dummy variable with the value of 1 if a union representative (de´le´gue´ syndical) is present at the workplace and 0 otherwise. In France, trade union representatives have a legal right to conduct collective bargaining with the employer at the workplace level. The Auroux laws (1982) and later the Aubry laws of 1998 and 2000 – on the 35-hour week – strengthened the role of trade unions at the workplace level. Thus, in this new context, it is interesting to examine how union representatives influence efficiency in the light of their new role at the workplace level. Evidence in support of their impact on employment relations may be drawn from the huge increase in the volume of collective bargaining at the firm level, up from 6,479 agreements reached in 1990 rising to 30,434 in 2000 (DARES, 2004). 3.4. Methodology Unlike financial measures of performance (such as ROA) or partial measures of productivity (such as output-to-labor ratios), efficiency is not observed and has to be inferred from the data. Consequently, statistical techniques need to be employed to reveal the underlying economic performance, using data on inputs and outputs. In this paper, we use data envelopment analysis (DEA) to calculate technical and scale efficiency for each individual French establishment. DEA is used to estimate the degree of technical efficiency among a group of organizations. DEA is a set of non-parametric programming techniques that can be used to rank the efficiency levels of like institutions. DEA compares outputs to inputs and orders the institutions in terms of their relationship to a best practice standard, where the best practice standard is the most efficient institution(s) in the group (or establishments in our case). DEA has now become a standard tool with numerous applications in a wide range of fields. For a full explanation of the methodology, see Coelli et al. (1998). The benefit of using DEA is that it is a non-parametric technique. Hence, no assumptions need to be made regarding the functional form of the underlying production function (e.g. there is no need to worry about whether a Cobb-Douglas or a Translog specification should be used).12 Our first step is to estimate the technical and scale efficiency scores. These are derived by using DEA to explore the association between the outputs and inputs for each firm, separately for each industry. We then calculate the distance between the maximum attainable output and the actual output produced – this is technical inefficiency. Once the DEA technical efficiency measures are calculated, the next step is to identify the factors that influence the degree of technical efficiency. To do this, we use Tobit regression
82
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
analysis to identify the determinants of both technical and scale efficiency.13 The dependent variable in this analysis is the efficiency score (technical or scale) for each firm, regressed against a set of potential explanatory variables, the main focus of which is the HR and HR–union interaction variables. This process effectively means that we are able to separate the contributions of HR practices from the contributions of labor and capital. This is important, as HR practices are not introduced in isolation. Both labor and capital are needed to implement them. Our first step (the use of DEA) controls for the effects of labor and capital. The second step (the Tobit regression analysis) explores the contributions of HR practices net of the contributions of labor and capital. It is important to note that the efficiency scores were derived separately for each industry. That is, DEA was applied to data for each of the industries separately, rather than pooling all data together, thus avoiding comparing apples to oranges. Accordingly, in the Tobit regression analysis, we include dummies for each of the main industry groups, such as transport, consumer goods, construction, health and food. Our DEA analysis was based on an output orientated approach with output measured as value added and labor and capital as the two inputs. Labor input is measured by the number of employees and the capital input is proxied by the value of fixed assets. The use of an input orientated approach resulted in largely similar results. An output orientated approach means that we estimate by how much output can be expanded given existing inputs.14 Our data is cross-sectional. Hence, we are unable to explore shifts in the production possibilities frontier. That is, the nature of data does not allow us to explore the impact of HR practices on both movements towards the frontier and shifts in the frontier. Time series data is needed to explore shifts in the frontier. Accordingly, our analysis is limited to explore the impact of HR practices on technical inefficiency (the distance from the frontier). It should be noted also that there is a difference between technical efficiency and profitability. The vast literature on the economic impact of unions (and HR practices) can be categorized into studies that explore the impact of unions on output levels and those studies that explore the impact of unions on profits. Unions can increase productivity, but if wages paid increase at a faster rate, profits will fall. Ceteris paribus, improvements in efficiency should increase profits. However, the introduction of HR practices may be associated with wage effects that impact on profits. Our analysis focuses on technical and scale efficiency and hence it explores only the output side of the ledger. We do not explore the impact of efficiency changes on profits.
Human Resource Practices, Unionization and the Organizational Efficiency
83
4. RESULTS The first set of results is presented in Table 4 and relate to the association between HR practices and technical efficiency. These inform on whether HR practices assist firms in achieving their full potential level of output, given the available pool of resources and technological know-how. Column 1 presents the results when HR is measured by HRINDEX1 (the number of HR practices). In column 2 we use HRINDEX2 (the binary variable). In column 3 we use HRINDEX3 (the sum of the factors). In column 4 we use individual factors: HR development; information sharing; participation; performance based pay; and teams. These were defined earlier. Note that due to lack of observations in some of the explanatory variables, the number of observations (firms) is reduced to 1,486. That is, the efficiency scores are obtained from a wider pool of firms from which best practice is identified, but a smaller number of firms is used to explore the impact of HR practices on efficiency.15 Column 1 presents the results using the simple count of HR practices and shows that HR practices on their own have a statistically significant detrimental effect on technical efficiency. However, when HR practices are interacted with trade union presence a positive effect emerges.16 This result is significant and important. On their own, unions have a negative effect on efficiency. This confirms what has been found in previous studies for other countries (e.g. Baldwin, 1992; Torii, 1992). However, those firms that have unions and which have implemented HR practices have higher levels of technical efficiency, than those firms that are unionized but do not have HR practices. HR practices interacted with strategy have no impact on technical efficiency. When Index 2 is used, there is no association between HR practices and technical efficiency (the coefficient is still negative but it is no longer statistically significant). However, as with Index 1, when HR practices (as measured by HRINDEX2) are interacted with union presence there is a statistically significant positive effect on technical efficiency. When the more informative and reliable HRINDEX3 is used, there is again a detrimental effect on technical efficiency, and once again when HR practices are interacted with union presence there is a statistically significant positive effect on technical efficiency. The final column of Table 4 reports the results of separating the HR practices into smaller groupings. Once again, trade union presence is detrimental to technical efficiency. With the exception of information sharing, none of the individual HR variables has an impact on technical efficiency. Information sharing apparently has a negative impact on efficiency in
84
HR Practices and Technical Efficiency, French Industry, 1998.
Explanatory Variables HR practices HR HR – development HR – information sharing HR – participation HR – performance-based pay HR – teams Interactions Union HR Union HR – development Union HR – information sharing Union HR – participation Union HR – performance-based pay Union HR – teams Strategy HR Strategy HR – development Strategy HR – info sharing Strategy HR – participation Strategy HR – performance-based pay Strategy HR – teams
HRINDEX1 (1) 0.013 (2.38) – – – – – 0.016 (3.74) – – – – – 0.003 (0.49) – – – – –
HRINDEX2 (2)
0.068 (1.62) – – – – – 0.094 (2.94) – – – – – 0.008 (0.20) – – – – –
HRINDEX3 (3) 0.024 (2.55) – – – – – 0.020 (3.00) – – – – – 0.015 (1.66) – – – – –
Factors (4)
0.026 0.037 0.022 0.008 0.030
0.036 0.044 0.013 0.011 0.001 0.002 0.001 0.001 0.009 0.056
– (1.27) (1.66) (1.19) (0.39) (1.63) – (2.43) (2.64) (0.87) (0.76) (0.09) – (0.12) (0.02) (0.33) (0.45) (3.36)
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Table 4.
0.003 (11.23) 0.011 (0.62) 0.009 (0.60) 0.274 (6.76) 0.019 (0.36) 0.63 (12.18) 149.04 88.60 15.17 1,486 0.19
0.003 (11.24) 0.013 (0.72) 0.008 (0.55) 0.171 (9.01) 0.001 (0.03) 0.542 (22.03) 147.19 89.06 8.99 1,486 0.18
Note: Dependent variable ¼ technical efficiency score, Tobit estimations. All Tobit regressions include unreported industry dummies. Statistically significant at the 10% level; Statistically significant at the 5% level; Statistically significant at the 1% level.
0.003 (11.37) 0.013 (0.74) 0.007 (0.47) 0.134 (8.62) 0.004 (0.21) 0.515 (23.93) 150.67 89.10 13.18 1,486 0.19
0.003 (11.29) 0.011 (0.60) 0.009 (0.63) 0.132 (8.18) 0.010 (0.51) 0.511 (23.63) 142.23 84.31 36.83 1,486 0.19
Human Resource Practices, Unionization and the Organizational Efficiency
Controls Size/100 Introduced new technology Introduced organizational change Union Strategy Constant Industry dummies (Wald test) w2 Union variables (Wald test) w2 HR variables (Wald test) w2 Sample size Adjusted R2
85
86
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
French industry (however, this is only of marginal statistical significance). The union interaction effects however tell a different story. Union presence interacted with information sharing and union presence interacted with HR development both have a positive impact on technical efficiency. While the strategy variables are not statistically significant, strategy interacted with HR teams has a positive effect. The results suggest that simply introducing HR practices is not enough to improve efficiency. HR practices are not universally relevant. There is a much more complex relationship between HR practices and technical efficiency. Our results indicate that some HR practices are more suitable under certain conditions and less suitable under others. The implication is that HR practices should be thought of as a part of a system of workplace practices that fit together in a coherent manner and fit with the overall strategy of the firm. Furthermore, HR practices should be implemented in a high-trust environment, since it appears that their effect is dependent upon a union’s willingness to share information. As the context in which HR practices operates is important to its success, we suggest that a contingency approach is certainly more appropriate to describe the effects of HR practices on efficiency. Table 1 reports also the Wald test results for the statistical significance of groups of variables. In all cases, the industry dummies are statistically significant.17 The Wald tests for both the unionization and HR practices groups of variables confirm the joint statistical significance of these groups of variables. That is, both unionization and HR practices are important explanatory variables of technical efficiency. Table 5 repeats the Tobit analysis for scale efficiency.18 All three aggregate HR indices are positively associated with scale efficiency – HR practices enable firms to move closer to the optimal scale. The results for the disaggregate HR practices are interesting (column 4). Development, information sharing and performance based pay all have a positive and direct effect on scale efficiency, while participation and teams have no effect at all. Interestingly, unions are also associated with higher levels of scale efficiency. The union-HR interactions are mostly negative. Unions have a negative impact on scale efficiency when interacted with HR practices, particularly with information sharing. The explanation for this finding is beyond the scope of this paper. However, it is possible that the use of information sharing practices is considered to be a threat by unions – a rival voice – and they do not support such practices. Consequently, the interaction term between unions and information sharing practices results in a significant negative relation with scale efficiency.
HR Practices and Scale Efficiency, French Industry, 1998.
Explanatory Variables
HRINDEX2 (2)
HRINDEX3 (3)
0.034 (7.43) – – – – –
0.208 (6.03) – – – – –
0.043 (5.53) – – – – –
Interactions Union * HR Union * HR – development Union * HR – information sharing Union * HR – participation Union * HR – performance–based pay Union * HR – teams HR * Strategy Strategy * HR – development Strategy * HR – information sharing Strategy * HR – participation Strategy * HR – performance-based pay Strategy * HR – teams
0.014 (3.49) – – – – – 0.011 (2.44) – – – – –
0.111 (3.70) – – – – – 0.065 (3.70) – – – – –
0.024 (3.82) – – – – – 0.014 (1.98) – – – – –
0.031 0.011 0.006 0.021 0.007
Controls Introduced new technology Introduced organizational change Union Strategy Constant Industry dummies (Wald test) w2
0.005 (0.35) 0.007 (0.57) 0.316 (8.71) 0.128 (2.95) 0.290 (7.03) 151.41
0.010 (0.64) 0.012 (1.02) 0.252 (13.82) 0.062 (2.92) 0.488 (23.37) 151.45
0.011 (0.72) 0.013 (1.03) 0.210 (14.17) 0.037 (2.40) 0.57 (30.02) 308.00
0.009 (0.61) 0.006 (0.51) 0.203 (15.28) 0.036 (2.29) 0.568 (32.43) 193.51
HR practices HR HR – development HR – information sharing HR – participation HR – performance-based pay HR – teams
Factors (4)
0.083 0.084 0.018 0.043 0.001
0.021 0.059 0.006 0.010 0.015
– (4.77) (4.97) (1.10) (2.69) (0.02) – (1.65) (4.32) (0.51) (0.85) (1.13) – (1.81) (0.62) (0.37) (1.40) (0.42)
87
HRINDEX1 (1)
Human Resource Practices, Unionization and the Organizational Efficiency
Table 5.
88
Table 5. (Continued ) Explanatory Variables Union variables (Wald test) w2 HR variables (Wald test) w2 Sample size Adjusted R2
HRINDEX1 (1) 196.74 87.34 1,486 0.35
222.85 50.62 1,486 0.33
HRINDEX3 (3) 31.51 25.95 1,486 0.33
Factors (4) 257.27 121.18 1,486 0.36
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Note: Dependent variable ¼ Scale efficiency score, Tobit estimations. All Tobit regressions include unreported industry dummies. Statistically significant at the 10% level; Statistically significant at the 5% level; Statistically significant at the 1% level.
HRINDEX2 (2)
Human Resource Practices, Unionization and the Organizational Efficiency
89
The results are consistent with prior research in that unions have a detrimental effect on technical efficiency. The scale efficiency results are, however, novel. Prior studies have largely ignored the scale efficiency issue. The results presented here indicate that French unions appear to raise scale efficiency even though they reduce technical efficiency.19 If the scale efficiency effects dominate technical efficiency effects, the net effect would be an increase in productivity (something that prior studies using the same dataset found, e.g. Coutrot, 1996). Our analysis suggests that the positive unionproductivity effect is driven by scale efficiency effects, as well as favorable union-HR practice interactive effects on technical efficiency. Interestingly, the results also show that the introduction of new technology and the introduction of organizational change has no effect on technical efficiency, once unions and HR practices are controlled for. This could, of course, be an artefact of the data, as information on technological and organizational change is available only as a binary variable. Strategy has no effect on technical efficiency but does have a favorable effect on productivity through scale efficiency improvements.
5. SENSITIVITY ANALYSIS The results presented in Tables 4 and 5 are derived from using the largest possible number of firms. There are two issues that need to be explored with respect to this dataset. First, there is the issue of aggregation bias. Second, it is pertinent to explore the impact of changing samples on the results. 5.1. Industry Aggregation Tables 4 and 5 were constructed using fairly broad industry categories; health; food; intermediate goods; equipment; consumer; construction; commerce; transport; services; and personal services. Within each of these categories, we assume that the underlying technology and product market conditions are similar. Hence, while we assume that consumer goods manufacturers are different to food processors, we assume also that all firms categorized as consumer goods manufactures have the same underlying technology. This assumption remains unproven for our dataset. Tables 6 and 7 repeat the analysis using more detailed sub-industry groupings: food; textiles; leather and garments; wood and paper; printing; chemicals and plastics; mineral products; metal transformations; mechanical equipment; family equipment; electric equipment and electronics; electronic
90
HR Practices and Technical Efficiency, Disaggregate French Industries, 1998.
Explanatory Variables HR practices HR HR – development HR – information sharing HR – participation HR – performance-based pay HR – teams Interactions Union * HR Union * HR – development Union * HR – information sharing Union * HR – participation Union * HR – performance–based pay Union * HR – teams Strategy * HR Strategy * HR – development Strategy * HR – information sharing Strategy * HR – participation Strategy * HR – performance-based pay Strategy * HR – teams
HRINDEX1 (1)
HRINDEX2 (2)
HRINDEX3 (3)
0.002 (0.23) – – – – –
0.025 (0.51) – – – – –
0.010 (0.96) – – – – –
0.008 (1.50) – – – – – 0.003 (0.35) – – – – –
0.055 (1.49) – – – – – 0.012 (0.27) – – – – –
0.008 (1.04) – – – – – 0.017 (1.63) – – – – –
Factors (4)
0.001 0.024 0.039 0.040 0.024
0.020 0.012 0.023 0.001 0.006 0.001 0.022 0.019 0.015 0.053
(0.02) (1.00) (1.82) (1.70) (1.12) (1.11) (0.62) (1.36) (0.55) (0.36) – (0.03) (0.95) (0.88) (0.66) (2.73)
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Table 6.
0.005 (10.45) 0.001 (0.01) 0.004 (0.22) 0.127 (2.74) 0.106 (1.64) 0.678 (9.97) 203.65 13.42 7.79 1,472 0.24
0.005 (10.52) 0.001 (0.04) 0.006 (0.38) 0.080 (3.48) 0.090 (3.18) 0.672 (17.70) 200.38 12.50 4.71 1,472 0.24
Note: Dependent variable ¼ technical efficiency score, Tobit estimations. All Tobit regressions include unreported industry dummies. Statistically significant at the 10% level; Statistically significant at the 5% level; Statistically significant at the 1% level.
0.005 (10.59) 0.001 (0.01) 0.003 (0.18) 0.060 (3.19) 0.086 (3.57) 0.666 (19.35) 206.39 11.67 11.15 1,472 0.24
0.005 (10.55) 0.001 (0.01) 0.004 (0.25) 0.061 (3.08) 0.083 (3.50) 0.665 (19.44) 208.97 14.05 29.44 1,472 0.24
Human Resource Practices, Unionization and the Organizational Efficiency
Controls Size/100 Introduced new technology Introduced organizational change Union Strategy Constant Industry dummies (Wald test) (w2) Union variables (Wald test) (w2) HR variables (Wald test) (w2) Sample size Adjusted R2
91
92
HR Practices and Scale Efficiency, Disaggregate French Industries, 1998.
Explanatory Variables HR practices HR HR – development HR – information sharing HR – participation HR – performance-based pay HR – teams Interactions Union * HR Union * HR – development Union * HR – information sharing Union * HR – participation Union * HR – performance–based pay Union * HR – teams HR * Strategy Strategy * HR – development Strategy * HR – information sharing Strategy * HR – participation Strategy * HR – performance-based pay Strategy * HR – teams
HRINDEX1 (1)
HRINDEX2 (2)
HRINDEX3 (3)
0.026 (5.15) – – – – –
0.145 (4.46) – – – – –
0.032 (3.95) – – – – –
0.019 (5.03) – – – – – 0.012 (2.30) – – – – –
0.124 (4.96) – – – – – 0.060 (1.96) – – – – –
0.024 (4.24) – – – – – 0.018 (2.21) – – – – –
Factors (4)
0.067 0.059 0.020 0.039 0.014
0.033 0.062 0.027 0.009 0.009 0.035 0.006 0.010 0.036 0.004
– (3.78) (3.52) (1.28) (2.48) (0.81) – (2.59) (4.36) (2.09) (0.71) (0.72) – (2.06) (0.39) (0.69) (2.43) (0.24)
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Table 7.
0.006 (0.45) 0.002 (0.15) 0.202 (5.85) 0.093 (2.00) 0.610 (13.29) 307.29 35.14 38.07 1,472 0.27
Note: Dependent variable ¼ Scale efficiency score, Tobit estimations. All Tobit regressions include unreported industry dummies. Statistically significant at the 10% level; Statistically significant at the 5% level; Statistically significant at the 1% level.
0.009 (0.64) 0.001 (0.03) 0.098 (5.81) 0.019 (0.95) 0.765 (36.20) 317.92 36.51 30.36 1,472 0.26
0.010 (0.68) 0.001 (0.0) 0.050 (3.75) 0.005 (0.33) 0.821 (51.07) 308.00 31.51 25.95 1,472 0.26
0.006 (0.38) 0.002 (0.17) 0.049 (3.61) 0.009 (0.57) 0.823 (51.74) 208.97 14.05 29.44 1,472 0.27
Human Resource Practices, Unionization and the Organizational Efficiency
Controls Introduced new technology Introduced organizational change Union Strategy Constant Industry dummies (Wald test) (w2) Union variables (Wald test) (w2) HR variables (Wald test) (w2) Sample size Adjusted R2
93
94
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
components; transport; aeronautical and naval; construction; car repairs and sales; wholesale; detail; and hotels and restaurants. On the one hand using more detailed classifications is desirable as it enables more refined groupings for comparison purposes. However, this also means that we are forced to use fewer observations when identifying efficiency. For example, the smallest number of observations for Tables 4 and 5 were 63 (Health industry) compared to 21 for Tables 6 and 7 (aeronautical and naval industry). For this reason, we have a preference for the results presented in Tables 4 and 5. The results for technical efficiency are noticeably different. While the coefficients for both the HR practices and the union–HR interaction variables are the same as in Table 4, these associations are estimated with less precision and consequently these variables are no longer statistically significant in Table 6. However, unionization continues to have a negative and statistically significant coefficient. Hence, the technical efficiency results are not robust. On the other hand, all the results for scale efficiency from Table 5 are confirmed by Table 7. HR practices and unionization both have a positive effect on scale efficiency and the interaction between these two variables has a depressing effect on scale efficiency. 5.2. Impact of Samples Table 8 presents the results of a sensitivity analysis derived from varying the samples. Columns 1 and 3 present the technical and scale efficiency results, respectively, when HRINDEX1 is used. The associated results for both technical and scale efficiency when HRINDEX3 is used are presented in columns 2 and 4. The first row in each panel reproduces the results from Tables 4 and 5. There are various problems with the measurement of output in service industries. Hence, the second row uses only manufacturing firms. The third row excludes the poorest 20% performing firms. Extending this notion, the fourth row considers only those firms whose efficiency score was at least 70%. The results are informative. The determinants of technical efficiency are not robust to the use of different samples. However, the determinants of scale efficiency are. In the case of scale efficiency scores, the magnitude of the coefficients does vary but the sign is unchanged regardless of the sample used. Importantly, the three variables of interest-HR practices, unionization and the interaction between HR and unions – are always statistically significant. This is strong evidence of a robust association between scale efficiency and HR practices and unionization. Both HR practices and
Human Resource Practices, Unionization and the Organizational Efficiency
Table 8.
Unionization All firms Manufacturing only Without worst performers EfficiencyZ0.7
Sensitivity Analysis. Scale Efficiency HRINDEX1 (3)
Scale Efficiency HRINDEX3 (4)
0.013 (2.38) 0.024 (2.55) 0.006 (0.63) 0.015 (1.04)
0.034 (7.43) 0.031 (3.93)
0.043 (5.53) 0.030 (2.42)
0.009 (1.54)
0.014 (1.54)
0.018 (3.73)
0.020 (2.91)
0.001 (0.15)
0.001 (0.11)
0.008 (2.36)
0.009 (1.97)
Technical Efficiency HRINDEX1 (1) HR practices All firms Manufacturing only Without worst performers EfficiencyZ0.7
95
Technical Efficiency HRINDEX3 (2)
0.274 (6.76) 0.134 (8.62) 0.316 (8.71) 0.331 (5.31) 0.156 (6.80) 0.392 (6.88)
0.210 (14.17) 0.227 (9.81)
0.224 (5.66) 0.095 (6.05) 0.175 (5.37)
0.098 (8.14)
0.083 (1.50)
Union * HR practices All firms 0.016 Manufacturing 0.019 only Without worst 0.016 performers EfficiencyZ0.7 0.005
0.014 (0.65)
0.070 (2.68)
0.035 (4.05)
(3.74) (2.98)
0.020 (3.00) 0.014 (3.49) 0.024 (3.82) 0.024 (2.55) 0.021 (3.09) 0.032 (3.18)
(3.48)
0.019 (2.84) 0.010 (2.77) 0.016 (3.24)
(0.82)
0.016 (1.83)
0.004 (1.60)
0.007 (1.84)
Note: Dependent variable ¼ Technical efficiency or scale efficiency score, Tobit estimations. All Tobit regressions include unreported industry dummies. Statistically significant at the 10% level; Statistically significant at the 5% level; Statistically significant at the 1% level.
unionization are individually associated with higher levels of scale efficiency, although their interaction tends to depress scale efficiency. The net effect is, however, that HR practices and unionization improve scale efficiency. Our estimated model predicts an average scale efficiency of 0.76 for a unionized firm compared to an average scale efficiency score of 0.56 for an nonunionized firm (using the average value of the HRINDEX3 and other variables). The regression analysis of technical efficiency was not, unfortunately, robust. HR practices have a negative association with technical efficiency only when the entire dataset is used, without any modifications for potential aggregation bias. All other samples suggest that HR has no effect on
96
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
technical efficiency in French firms. Unionization has a negative and statistically significant coefficient when the entire sample is used, when only manufacturing firms or when the worst performing firms are omitted. Unionization does not appear to have any effect on technical efficiency among the more efficient firms. While Tables 4 and 5 present our preferred results, Table 8 does suggest caution. Three robust conclusions emerge from Tables 4 to 8. First, unions on their own have a detrimental effect on technical efficiency. Second, there is a positive effect on scale efficiency from HR practices and unionization. To our knowledge the links between HR and scale efficiency have not been explored adequately in the literature. The theoretical underpinnings of this association are beyond the scope of this paper, and further empirical research is needed to explore whether our findings apply to other nations. We do however hypothesize two channels through which this effect may work. Scale efficiency is achieved by increasing the scale of production. One way through which this may be achieved is through mergers and acquisitions. It is possible that HR practices and unionization may influence the success or otherwise of these activities. A more likely channel, we believe, is through coordination and the efficient management of expanded operations. Technological characteristics are necessary but not sufficient for the exploitation of economies of scale. Technology may, for example, enable an 8% increase in output from a 5% increase in all inputs (increasing returns to scale). Whether this is actually realized however depends on the ability of an organization to co-ordinate and manage the expanded use of inputs (especially labor). It is our view that HR practices and unions can play a key role in this regard. Our results suggest that for French industry, both HR practices and unions enable firms to realize economies of scale.20 Tables 5 and 7 relate to scale efficiency scores. Also of interest, however, is the nature of the returns to scale. The association between returns to scale and the number of HR practices is presented in Table 9. Table 9 lists the proportion of firms that are operating with increasing, decreasing and constant returns to scale. Only 4% of the firms in the sample operated at the optimal level (constant returns to scale), and none of these were firms that had less than 5 HR practices. There is a clear association between the number of HR practices and decreasing returns to scale. As the number of HR practices increases, the proportion of firms operating with decreasing returns to scale rises. The reverse holds for increasing returns to scale. The more HR practices, the smaller the fraction of firms operating with increasing returns to scale. Taken together, the results presented in columns 2
Human Resource Practices, Unionization and the Organizational Efficiency
Table 9.
97
Returns to Scale, HR Practices and Unionization.
Number of HR Practices
Increasing Returns to Scale (1)
Constant Returns to Scale (2)
Decreasing Returns to Scale (3)
HR practices 0o5 5o10 10 plus
11% 25% 10%
0% 2% 2%
2% 13% 25%
Unionization Unionized Non-Unionized
43% 86%
5% 2%
53% 11%
Average efficiency score Unionized Non-Unionized
0.75 0.51
1.00 1.00
0.72 0.90
and 3 indicate that as firms adopt more HR practices, they are more likely to expand the size of their operations and operate with either constant returns to scale or decreasing returns to scale. The results presented in Table 9 do not imply causation. It may be that larger firms implement HR practices because they have more capacity to bear the costs of doing so. It may also be the case that firms facing DRS implement HR practices as a way of improving efficiency. As noted earlier, HR practices may enable better coordination and management of larger operations. We need time series data to explore these matters further. It is not clear why the HR–union interaction has a negative effect on scale efficiency. One explanation could be that unions feel threatened by HR practices and consequently negotiate suboptimal manning levels. The second panel of Table 9 compares unionized firms to non-unionized. The average scale efficiency for all unionized firms was 0.75, compared to 0.57 for all non-unionized firms. Non-unionized firms that operate with decreasing returns to scale do so at higher levels of efficiency than do corresponding unionized firms (0.90 compared to 0.72). However, if we treat any score of 0.95 or over as effectively on the frontier, then we find that 13% of non-unionized firms can be considered as fully scale efficient, compared to 25% of unionized firms. Among those firms operating with increasing returns to scale (86% of non-unionized firms), the average efficiency score is 0.51 for non-unionized compared to 0.75 for unionized. That is, the distribution of scale efficiency is such that unionized firms fare better in terms of this measure of efficiency.
98
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
6. SUMMARY While there has been much research on the impact of HR practices on overall performance, little is known about the impact of HR practices on economic efficiency. The aim of this paper was to investigate the links between HR practices and efficiency, as well as between unions and efficiency and the impact of HR-union interaction effects. DEA was used to calculate technical and scale efficiency scores for a large sample of French firms. The determinants of efficiency scores were then explored through Tobit analysis. In assessing the impact of HR on efficiency, the results show a complex set of associations and that context matters. When HR practices are introduced into a French unionized workplace, the interaction results in a less unfavorable effect on technical efficiency, while at the same time resulting in a negative interaction effect on scale efficiency. On the other hand, HR practices on their own have no affect on technical efficiency, but they do have a large positive effect on scale efficiency. French unions have a detrimental effect on technical efficiency, as well as a positive effect on scale efficiency. On balance, HR practices appear to have a positive effect on productivity in French industry. The analysis needs to be extended in several ways. For example, the impact of HR practices on profitability through their effects on efficiency is a key research question. The use of panel data will inform on the impact of HR practices on technological change and productivity growth.
NOTES 1. Recent studies include Bartel (2004) for the US and Guest (2003) for the UK. 2. Producing the maximal level of outputs given inputs, or using the minimal level of inputs given output. 3. Similarly, if firms operate with decreasing returns to scale, it is optimal for them to contract their scale of operations. 4. Importantly, this conclusion is robust and is not affected by publication bias (see Doucouliagos, Laroche, & Stanley, 2005). 5. The impact of HR practices on performance can be moderated by other factors. For example, Dawson, Knight-Turvey, Neal, and West (2004) found that financing strategy influences the HR-performance association. 6. Delaney (1991, p. 39) argues that the lack of collaboration between industrial relations and HR management researchers is one barrier to such an assessment of these subjects. 7. A related issue is worker attitudes, including work motivation, job satisfaction and morale. In an interesting study, Guest and Conway (1999) find that the poorest
Human Resource Practices, Unionization and the Organizational Efficiency
99
work experience was among employees from unionized workplaces with little HR practices, and that unionized places with no HR practices faired better. 8. Owing to the cross-sectional nature of the data, it is not possible to explore the impact of HR practices on technological change, but we can explore technical and scale efficiency. 9. Even though researchers have access to REPONSE, because of confidentiality it is not possible to identify the firms covered in the REPONSE database. Hence, it is not possible for researchers themselves to match performance data with the employee relations data. The matching of the REPONSE and DIANE databases was kindly carried out by Anne Saint-Martin, from the French Ministry of Labor. 10. An alternative approach involves using cluster analysis to identify groups of homogeneous cases (MacDuffie, 1995; Arthur, 1994; Ichniowski, 1990). 11. The scores for each group or factor were calculated in the form of linear combinations of all the practices, using the SAS r Software. 12. The main alternative to DEA is to Stochastic Frontier Analysis, but the efficiency scores derived from this can be sensitive to the specification of the production function and assumptions regarding the distribution of the error terms. 13. DEA generates efficiency scores in the range of 0–1 (1 is the most efficient). Hence, the results are truncated at 1, and OLS is not an appropriate estimator to use. 14. The input orientated approach asks by how much can inputs be reduced without affecting output. 15. The larger pool is recommended to get more accurate efficiency scores. 16. In prior analysis, it was found that interactions between HR practices and technology and organizational change were not statistically significant. 17. The results confirm also that there are statistically significant differences in the average levels of efficiency across industries. 18. We do not include establishment size in this analysis. 19. The average scale efficiency score for unionized firms is 0.75 and is 0.57 for non-unionized firms. 20. The concept of returns to scale is related directly to long-run cost curves. We leave the issue of the actual shape of French industry cost-curves to further research. Our DEA analysis was based on production associations and is, hence, informing only on the dual to the cost side.
REFERENCES Addison, J., & Belfield, C. (2004). Union voice. Journal of Labor Research, 25(4), 563–596. Allouche, J. (1984), Performances e´conomiques et performances sociales, cohe´rence ou contradiction. Institut ADIA. Appelbaum, E., Bailey, T., Berg, P., & Kalleberg, A. L. (2000). Manufacturing advantage: Why high-performance work systems pay off. Ithaca and London: Economic Policy Institute, Cornell University Press. Arthur, J. (1992). The link between business strategy and industrial relations systems in American steel minimills. Industrial and Labor Relations Review, 45, 488–506. Arthur, J. (1994). Effects of human resource systems on manufacturing performance and turnover. Academy of Management Journal, 37(3), 670–687.
100
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Baldwin, J. (1992). Industrial efficiency and plant turnover in the Canadian manufacturing sector. In: R. E. Caves (Ed.), Industrial efficiency in six nations (pp. 273–310). Cambridge, MA: MIT Press. Bartel, A. (2004). Human resource management and organizational performance: Evidence from retail banking. Industrial and Labor Relations Review, 57(2), 181–203. Bayad, M., & Liouville, J. (1998). Human resource management and performances: Proposition and test of a causal model. Human System Management, 17(3), 18–38. Becker, B., & Gerhart, B. (1996). The impact of human resource management on organizational performance: Progress and prospects. Academy of Management Journal, 39, 779–801. Becker, B. E., & Huselid, M. A. (1998). High performance work system and firm performance: A synthesis of research and managerial implications. Research in Personnel and Human Resources Management, 16, 53–101. Black, S., & et Lynch, L. (2001). How to compete: The impact of workplace practices and information technology on productivity. Review of Economics and Statistics, 83(3), 434–445. Boselie, P., Dietz, G., & Boon, C. (2005). Commonalities and contradictions in HRM and performance research. Human Resource Management Journal, 15(3), 67–94. Bronars, S., Deere, D., & Tracy, J. (1994). The effect of unions on firm behavior: An empirical analysis using firm-level data. Industrial Relations, 33(4), 426–451. Byrnes, P., Fare, R., Grosskopf, S., & Lovell, C. A. (1988). The effect of unions on productivity: US surface mining of coal. Management Science, 34(September), 1037–1053. Cavalluzzo, L., & Baldwin, D. (1993). Unionization and productive efficiency. In: F. Harold, C. A. Knox Lovell & S. Schmidt (Eds), The measurement of productive efficiency (pp. 210–220). Oxford University Press. Coelli, T., Rao, P., & Battese, G. E. (1998). An introduction to efficiency and productivity analysis. Boston: Kluwer. Cooke, W. (1994). Employee participation programs, group-based incentives and company performance: A union-nonunion comparison. Industrial and Labor Relation Review, 47(4), 594–609. Coutrot, T. (1996). Relations sociales et performance e´conomique: une premie`re analyse du cas franc- ais. Travail et Emploi, 66, 39–66. Coutrot, T. (1998). L’entreprise ne´o-libe´rale, nouvelle utopie capitaliste? Paris, La De´couverte. Coutrot, T. (2000), Innovation dans le travail: La pression de la concurrence internationale, l’atout des qualifications. Premie`res Synthe`ses, MES-DARES, 09.2, Mars. d’Arcimoles, C. H. (1997). Human resource policies and company performance: A quantitative approach using longitudinal data. Organization Studies, 18(5), 857–874. DARES. (2004). Bilan de la ne´gociation collective en 2003. La tendance, les dossiers, les chiffres. Paris: Editions Le´gislatives. Dawson, J., Knight-Turvey, N., Neal, A., & West, M. (2004). The impact of an innovative human resource function on firm performance: The moderating role of financing strategy. Centre for Economic Performance, LSE, CEP Discussion Papers. Delaney, J. T. (1991). Unions and human resource policies. Research in Personnel and Human Resources Management, 9, 39–71. Delery, J. E., & Doty, D. H. (1996). Modes of theorizing in strategic human resource management: Tests of universalitic, contingency and configurational performance predictions. Academy of Management Journal, 39(4), 802–835. Dickerson, A. P., Geroski, P. A., & Knight, K. G. (1997). Productivity, efficiency and strike activity. International Review of Applied Economics, 11(January), 119–134.
Human Resource Practices, Unionization and the Organizational Efficiency
101
Doucouliagos, C., & Laroche, P. (2003a). What do unions do to productivity? A meta-analysis. Industrial Relations, 42(4), 650–691. Doucouliagos, C., & Laroche, P. (2003b). Unions and tangible investments: A review and new evidence in France. Relations Industrielles, 58(2), 314–338. Doucouliagos, C., Laroche, P., & Stanley, T. D. (2005). Publication bias in industrial relations research? Relations Industrielles, 60(2), 320–346. Dickerson, A. P., Geroski, P. A., & Knight, K. G. (1997). Productivity, efficiency and strike activity. International Review of Applied Economics, 11(1), 119–134. Eaton, A., & Voos, P. (1992). Unions and contemporary innovations in work organization, compensation, and employee participation. In: L. Mishel & P. B. Voos (Eds), Unions and economic competitiveness. Armonk, NY: M.E. Sharpe. Fiorito, J. (2001). Human resource management practices and worker desires for representation. Journal of Labor Research, 33(2), 335–354. Fombrun, C., Tichy, N. M., & Devanna, M. A. (1984). Strategic human resource management. New York: John Wiley & Sons. Freeman, R., & Medoff, J. (1984). What do unions do? New York: Basic Books. Godard, J. (2004). A critical assessment of the high-performance paradigm. British Journal of Industrial Relations, 42(2), 349–378. Godard, J., & Delaney, J. T. (2000). Reflection on the ‘‘High Performance’’ paradigm’s implications for industrial relations as a field. Industrial and Labor Relations Review, 53(3), 482–502. Guest, D. E., & Conway, N. (1999). Peering into the black hole: The downside of the new employment relations in the UK. British Journal of Industrial Relations, 37(3), 367–389. Guest, D. E. (2003). Human resource management and corporate performance in the UK. British Journal of Industrial Relations, 41(2), 291–314. Guerrero, S., & Barraud-Didier, V. (2004). High-involvement practices and performance of French firms. International Journal of Human Resource Management, 15(8), 1410–1425. Guthrie, J. P. (2001). High-involvement work practices, turnover and productivity: Evidence from New Zealand. Academy of Management Journal, 44(1), 180–190. Hirsch, B. (2004). What do unions do for economic performance? Journal of Labor Research, 15, 415–455. Huselid, M. A. (1995). The impact of human resource management practices on turnover, productivity, and corporate financial performance. Academy of Management Journal, 38(3), 635–672. Huselid, M. A., Jackson, S. E., & Schuler, R. S. (1997). Technical and strategic human resource management effectiveness as determinants of firm performance. Academy of Management Executives, 40(1), 171–188. Ichniowski, C. (1990). Human resource management systems and the performance of US manufacturing businesses, Cambridge, Mass: National Bureau of Economic Research, Working Paper No. 3449. Ichniowski, C., & Shaw, K. (1999). The effects of human resource management systems on economic performance: An international comparison of US and Japanese plants. Management Science, 45(5), 704–721. Ichniowski, C., Shaw, K., & Prennushi, G. (1997). The effects of human resource management practices on productivity: A study of steel finishing lines. American Economic Review, 87, 291–313.
102
HRISTOS DOUCOULIAGOS AND PATRICE LAROCHE
Kato, T., & Morishima, M. (1995). The productivity effects of human resource management practices: Evidence from new Japanese panel data. Working Paper No. 143, Colgate University, June. Kizilos, M., & Reshef, Y. (1997). The effect of workplace unionization on worker responses to HRM innovation. Journal of Labor Research, XVIII(4), 641–656. Laroche, P. (2004a). Pre´sence syndicale et performance financie`re des entreprises: une analyse statistique sur le cas franc- ais. Finance Controˆle Strate´gie, 7(3), 117–145. Laroche, P. (2004b). The impact of unions on workplace financial performance: An empirical study in the French context. In: P. B. Voos (Ed.), Proceedings of the industrial relations research association (IRRA) 56th Annual Meeting, IRRA Series, pp. 154–169. Lawler, E. E. (1991). High-involvement management: Participative strategies for improving organizational performance. San Francisco, CA: Jossey-Bass. MacDuffie, J. P. (1995). Human resource bundles and manufacturing performance: Organizational logic and flexible production systems in the world auto industry. Industrial and Labor Relations Review, 48(2), 197–221. MacNabb, R., & Whitfield, K. (1997). Unions, flexibility, team working and financial performance. Organization Studies, 18(5), 821–838. Machin, S., & Wood, S. (2005). Human resource management as a substitute for trade unions in British workplaces. Industrial & Labor Relations Review, 58. Michie, J., & Sheehan, M. (1999). Human resource management and corporate performance in the UK. British Journal of Industrial Relations, 41(2), 291–314. Osterman, P. (1994). How common is workplace transformation and who adopts it? Industrial and Labor Relations Review, 47(2), 173–188. Porter, M. E. (1980). Competitive advantage: Creating and sustaining superior performance. New York: Free Press. Reshef, Y., Bemmels, B., & Wolfe, R. (1993). The effect of unionization on workplace innovation. Economic and Industrial Democracy, 14(February), 109–131. Reshef, Y., Kizilos, M., Ledford, G., & Cohen, S. (1999). Employee involvement programs: Should unions get involved? Journal of Labor Research, XX(4), 557–569. Ruelland, N. (2001). Les institutions repre´sentatives du personnel en 1999. Premie`res Synthe`ses, MES-DARES, 48 (1). Schuler, R. S., & Jackson, S. E. (1987). Linking competitive strategies and human resource management practices. Academy of Management Executives, 3(1), 207–219. Torii, A. (1992). Technical efficiency in Japanese industries. In: R. E. Caves (Ed.), Industrial efficiency in six nations (pp. 31–120). Cambridge, MA: MIT Press. Torii, A., & Caves, R. E. (1992). Technical efficiency in Japanese and US manufacturing industries. In: R. E. Caves (Ed.), Industrial efficiency in six nations (pp. 423–457). Cambridge, MA: MIT Press. Turnbull, P. (2003). What do unions do now? Journal of Labor Research, 14, 491–527. Verma, A., & McKersie, R. B. (1987). Employee involvement: The implications of noninvolvement by unions. Industrial and Labor Relations Review, 40(July), 556–568. Wood, S. (1999). Human resource management and performance. International Journal of Management Reviews, 1(4), 367–413. Wright, P. M., & Boswell, W. (2002). Desegregating HRM: A review and synthesis of micro and macro human resource management research. Journal of Management, 28, 247–276. Youndt, M. A., Snell, S. A., Dean, J. W., & Lepak, D. P. (1996). Human resource management, manufacturing strategy, and firm performance. Academy of Management Journal, 39, 836–866.
AN EMERGENT THEORY OF HRM: A THEORETICAL AND EMPIRICAL EXPLORATION OF DETERMINANTS OF HRM AMONG IRISH SMALL- TO MEDIUM-SIZED ENTERPRISES (SMES) Brian Harney and Tony Dundon ABSTRACT Utilizing data drawn from 18 in-depth case studies the authors explore in detail the factors shaping employment in a diverse range of Irish smalland medium-sized enterprises. Existing theory in HRM is deemed inadequate in capturing the complexity of HRM in SMEs especially as it treats organizations as hermetically sealed entities. In an effort to animate the criticism directed at normative models of HRM the authors use a conceptual framework with an emergent, open systems theoretical proposition to examine the parameters, dynamics and determining factors of HRM at each of the case study companies. The results show that the notion of a normative HRM model was not coherent in terms of actual practices but rather reactive, and emergent HRM-related processes were often imposed to meet legislative requirements or to reinforce owner– manager legitimacy and control. The authors conclude that an Advances in Industrial and Labor Relations, Volume 15, 103–153 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15002-7
103
104
BRIAN HARNEY AND TONY DUNDON
appreciation of the interaction between structural factors both inside and outside the immediate work milieu is crucial if the heterogeneity of HRM in SMEs is to be adequately accommodated and understood.
INTRODUCTION Recognition of the positive role that HRM plays in facilitating the attainment of competitive advantage is becoming ubiquitous (Boxall & Purcell, 2003; Tyson, 1997). Research contributions continuously extol the virtues of HRM and stress the imperative of introducing formal, sophisticated HRM to increase organizational performance (Huselid, 1995; Pfeffer, 1998). The theoretical debate around this nexus has been consumed by a contest between two normative models of HRM, namely ‘best practice’ and ‘best fit’ (Boxall & Purcell, 2000). Yet traditional accounts suffer from an obsessive concentration on atypical (i.e. large) firms, taking as their referent HRM interests as they have been articulated in these firms (Hendry, Arthur, & Jones, 1995; Wilkinson, 1999). Research in HRM is therefore extremely skewed to the periphery in terms of firm size as it provides little information as to the nature and form of practices, whether labeled HRM or not, adopted in small to medium sized enterprises (SMEs). In this paper, we address this imbalance by examining in detail the parameters, dynamics, and determining factors of HRM among a sample of Irish SMEs. The espoused universalism which characterizes the HRM project rests on a homogenous projection of HRM across all settings. This assumed universal relevance of the findings derived from large firms to SMEs has been termed ‘little big business syndrome’ (Cassell, Nadin, Gray, & Clegg, 2002; Welsh & White, 1981). Very rarely, however, has the applicability of normative frameworks been assessed, or an awareness of HRM in SMEs been considered, to inform broader debates (Baron, 2003). Through examining the determinants of HRM in SMEs we therefore hope to contribute in an ‘‘area where the debate has stalled and little advancement has occurred’’ (Barrett & Rainnie, 2002, p. 426). Our approach focuses on exploring the influences that shape and condition HRM in the sample of SMEs as opposed to assessing the business performance effects of the HRM practices in place. In so doing our ultimate objective is to move beyond mere recognition of the heterogeneity of HRM in SMEs toward understanding, accommodation, and explanation (Rainnie, 1989; Wilkinson, 1999, p. 214). Arguably, until the type and form of HRM adopted by SMEs is recognized and explored more fully, and the complexity used to shape current debates,
An Emergent Theory of HRM
105
then theoretical models of HRM can only ever be partial (Dundon et al., 2001, p. 449).
CONTEXTUALIZING THE SIGNIFICANCE OF SMES The marginalization of SME from mainstream HRM research is both ‘disheartening’ (Heneman, Tansky, & Camp, 2000) and intriguing as it comes at a time when managerial discourse espouses the virtue of the small firm format, as evidenced by literature on core competencies, delayering, and industrial clusters (Redman & Wilkinson, 2001, p. 5). The numerical, social, and economic importance of SMEs cannot be overstated (Storey, 1994). In Ireland and the UK it is estimated that 99% of firms have less than 250 employees (Cully, Woodland, O’Reilly, & Dix, 1999; DTI, 2001; Forfas, 1999). Recent empirical studies show that SMEs contribute to over 55% of GDP and 65% of total employment in high-income countries (ILO, 2004; OECD, 2005). In the European Union (EU) it is estimated that SMEs represent two-thirds of total employment, accounting for over 65% of EU financial turnover (OES, 2002). Comparative figures relating specifically to an Irish context are difficult to source (Lynch & Roche, 1995). In the UK, SMEs are said to account for 37% of financial turnover and 44% of nongovernmental employment. In contrast, the 7,000 largest businesses account for 45% of non-governmental employment and 49% of turnover (DTI, 2001). Yet rigorously defining SMEs and measuring their relative economic contribution has always been difficult, even controversial (d’Amboise & Muldowney, 1988, p. 226, our emphasis). Moreover there are dangers in using figures in a homogenous and deterministic way. In reality SMEs are remarkably diverse in terms of size, sector, activity, ownership, location and the markets they serve (Fuller & Moran, 2001, p. 18; Churchill & Lewis, 1983, p. 30). Accounts which might provide insights into HRM in SMEs have been plagued by size determinism and crude stereotypes (Barrett & Rainnie, 2002). Typically, these have involved conflating the characteristics of SMEs along opposite ends of a continuum of practices and their associated effects either toward a ‘small is beautiful perspective’ or a ‘bleak house perspective’ (Sisson, 1993; Wilkinson, 1999). Although conceptually elegant such polarizations gloss over the complexity and dynamic nature of work and work relations in small firms. Surface level characteristics of SMEs, such as the extent of informality, cannot be judged to be indicative of the substance and effectiveness of HRM (Gunnigle & Brady, 1984). Informality is dynamic
106
BRIAN HARNEY AND TONY DUNDON
and often co-exists with more formal practices (Bacon, Ackers, Storey, & Coates, 1996; Ram, Edwards, & Gilman, 2001). Moreover, even from the ‘small is beautiful’ perspective the perceived surface level ‘good practice’ may not reflect the reality as experienced by employees. Control, for example, can be achieved through the simultaneous use of both paternalistic and authoritarian managerial styles (Dundon, Grugulis, & Wilkinson, 1999). In practice the form HRM takes in SMEs is mediated through a web of economic and social relations and so HRM in this context is best noted for its ‘high unevenness’, ‘marked heterogeneity’ and complexity (Carroll, Marchington, Earnshaw, & Taylor, 1999; Duberley & Walley, 1995; Edwards, Gilman, Ram, & Arrowsmith, 2003; Storey, 1994). Based on these considerations and in an effort to advance our understating of this complexity the next section questions the ‘best practice’ and ‘best fit’ perspectives in much of the managerialist literature, assessing their utility in capturing the dynamic and contradictory nature of HRM in an SME context.
NORMATIVE MODELS IN AN SME CONTEXT Applicability of Best Practice in an SME Context The best practice argument is best epitomized in Pfeffer’s assertion that ‘‘the effects of high-performance management practices are real, economically significant, and general, and thus should be adopted by your organization’’ (1998, pp. 33–34). Yet while the universal appeal of best practice is intuitively attractive a number of conceptual and methodological issues draw caution to overemphasizing its promise (Legge, 2001; Sparrow, 1999; Wall & Wood, 2005). Specifically, neither conceptual/prescriptive (e.g. Lawler, 1992; Pfeffer, 1998) nor empirical work (e.g. Arthur, 1994; Huselid, 1995) yields agreement as to what actually constitutes best practice HRM or how it should be measured (e.g. indexes, scales, clusters). Few studies provide details as to the precise mechanism(s) by which HRM-performance take effect. Yet in the absence of a specific theory as well as extensive consideration of employee attitudes and behaviors (for exceptions on both counts see Applebaum, Bailey, Berg, & Kalleberg, 2000) alternative interpretations indicating negative implications of best practice HRM for employees and trade unions cannot be dismissed (Ramsay, Scholarios, & Harley, 2000; Delbridge & Whifield, 2001). Some have also questioned the very premise of
An Emergent Theory of HRM
107
identifying best practice bundles as one that compromises the scope for building competitive advantage as a result of firm-specific HRM (Duberley & Walley, 1995, p. 892; Purcell, 1999). Research in this area has also been subject to a number of methodological criticisms in particular because it has pursued largely a cross-sectional survey-based agenda directed largely at manufacturing establishments (e.g. Arthur, 1992, 1994; Applebaum et al., 2000). This raises difficulties in establishing the direction of causality and questions the relative generalizability of these studies. Moreover varying levels of analysis, e.g. plant/workplace studies (Arthur, 1994) or firm level studies (Huselid, 1995; Lee & Johnson, 1998), and differing performance measures from direct productivity considerations (e.g. adjusted line up time Ichniowski et al., 1997) to more removed firm-level performance data (e.g. Tobins Q Huselid, 1995) make comparison and reaching conclusive assertions difficult. Some have also questioned the reliability of single respondent self-reported performance data. Notably only a few studies (Ichniowski et al., 1997; MacDuffie, 1995) have utilized multiple respondents and relied on objective productivity measures, and even in these cases it is unclear as to how the information was combined, and the extent to which it was consistent across sources. Yet even in cases where best practice effects can be demonstrated, financial constraints and market changes may mean that they are simply overtaken by events (Stace & Dunphy, 1991; Mabey, Salaman, & Storey, 1998). Thus, while intra-level studies and sophisticated statistical techniques might control or allow for exogenous variables, in general, the impact of organizational size, technology, and market conditions and the way that they may condition the adoption and sustainability of practices has been downplayed (Becker & Gerhart, 1996). Clearly, however, the magnitude of the economic benefits from adopting new systems will be qualified by the vagaries of market forces, contextual idiosyncrasies, and institutional factors (Godard, 2004; Pil & MacDuffie, 1996, p. 450; Purcell, 1999). In an effort to explain the sporadic diffusion and sustainability of best practices some work has directed more attention toward the role of sectoral and organizational variables (e.g. Ichniowski, Kochan, Levine, Olson, & Strauss, 1996, p. 327; Osterman, 1994; Pil & MacDuffie, 1996). Nonetheless contributions in this area tend to focus exclusively on factors conditioning how rather than whether best practice should be implemented, and by so doing are largely ignorant of the idea that what constitutes best practice may vary across time and place (Delaney & Godard, 2001). These criticisms of best practice are ones which carry particular weight when the attributes of factors influencing HRM in an SME context are
108
BRIAN HARNEY AND TONY DUNDON
considered. Critically, best practice assumes the ability and resource base for long-term investment in HRM practices (Kerr & McDougall, 1999, p. 66; Marchington & Grugulis, 2000, p. 918). Research has shown, however, that in many SMEs performance can be achieved with modest levels of training and wages (Brand & Bax, 2002; Hendry et al., 1995). In terms of recruitment Carroll et al. (1999) found that SMEs effectively used ‘tried and trust of methods’ such as word of mouth as opposed to sophisticated techniques. Economies of scale clearly create unique challenges for SMEs and influence HRM investment decisions (Klass, McClendon, & Gainey, 2002; Sels et al., 2006). Recently, Cardon and Stevens (2004, p. 297) noted reluctance among SMEs to engage in costly and restrictive practices, therefore questioning the viability, or even necessity, of best practice prescriptions. Casual factors for the slow adaptation of best practice in SMEs may include labor market conditions and product market contingencies (d’Amboise & Muldowney, 1988, p. 237; Jackson Schuler, 1995; McMahon, 1996). The ‘hierarchically contracted’ nature of SMEs also mitigates against the internal promotion prescribed by best practice authors such as Pfeffer (1998). Truss’ (2001) criticism of HRM survey research for its biased imposition of preconceived ideals of HR practices therefore seems to find added weight in an SME context. Marchington, Carrol, and Boxall (2003, p. 20) note that firms may utilize a range of ‘intelligent’ formal and informal methods specific to their context which should not be necessarily castigated as unsophisticated or less effective. Moreover, where more sophisticated practices are in existence in SMEs these may be directed solely at attracting and retaining a selected few or group of core employees (Cardon, 2003; Matlay, 2002). Evidently, there may be variances in the type of HRM applied within as well as across firms (Lewin, 2001, p. 277; Osterman, 1994). The best practice literature also ignores broader societal and legislative issues, thereby neglecting factors that may condition the choice and implementation of practices (Godard, 2004). Such factors are critical in an SME context as it has been noted that SMEs are uniquely positioned in the face of environmental changes and have a consequent requirement for adaptability (Hendry et al., 1995). Furthermore in an SME context more immediate, short ranged, and pragmatic goals linked with issues of survivability and sustainability may carry more weight than the quest for competitive advantage (d’Amboise & Muldowney, 1988). Boxall and Purcell (2003, p. 238), for example, suggest that there is merely a minimum HRM ‘table stake’ required to compete in each industry. By glossing over such factors the best practice argument presents a biased and over simplistic depiction of the realities of HRM (Marchington & Grugulis, 2000, p. 922). Such
An Emergent Theory of HRM
109
accounts are inadequate in capturing the dynamic and diverse nature of HRM in SMEs. Given these limitations the next section reviews an alternative normative approach under the guise of ‘best fit’ which may better capture the importance of internal and external contingencies in the SME context. Applicability of Best Fit in an SME Context Best fit argues that different HRM responses may be appropriate in light of the strategy being pursued and the organization’s environmental context (Baird & Mesaulham, 1988; Paauwe & Boselie, 2005). Much of this literature has focused on the vertical linkage between HRM and corporate strategy (e.g. Schuler & Jackson, 1987) or alternatively advocated matching HRM responses to the stage of development of the organization (Baird & Meshoulam, 1988; Greiner, 1988). Youndt, Dean, and Lepak (1996) highlight the importance of looking at organizational phenomena to derive more situationally specific theories and prescriptions for management. Yet most criticisms of best fit argue that it overestimates the clarity and rationality of this process. The approach seems to be imbued with a level of determinism with the notion that HRM can simply be ‘read off’ strategy or stage of development, and indeed that suitable HR interventions can be found to ‘fit’ in the first place (Mabey & Salaman, 1995, p. 7). While some extensive research claims to have demonstrated the mediating role of strategy in enhancing the HRM performance linkage (e.g. Youndt et al., 1996; Arthur, 1992) results have been subject to varying interpretations. Huselid and Rau (1997), for example, contend that Arthur’s (1992) work indicates a significant association between HRM systems and business strategy choices, yet 40% of mills in Arthur’s study with a strategy of differentiation did not have the commitment HRM system (Arthur, 1992, p. 502). Wood (1999) argues that the performance effects of a number of studies (e.g. Arthur, 1992, 1994; MacDuffie, 1995) may be attributable to horizontal linkages between HRM systems and production strategies or ‘organization logic’ as opposed to emanating exclusively from HRM systems or their linkages with business strategy. More critically accounts may founder conceptually on debates over the mutual exclusivity of Porter’s generic strategies; quality and cost may not be necessarily antithetical (see Murray, 1988). Further within diversified firms it is questionable whether one specific strategy is pursued, so that level of analysis becomes crucial. The best fit literature therefore presents a very traditional, rational conception of strategy, often emphasizing vertical fit
110
BRIAN HARNEY AND TONY DUNDON
and the expense of horizontal interactions (Brewster, 1999, p. 366). Further little research is directed at understanding how (HRM) planning really works in practice. Clearly, HR initiatives and the management of employees in SMEs may be crafted rather than designed (Mintzberg & Waters, 1985; Wilkinson, 1999). Strategy may therefore not precede action but may only emerge retrospectively once action has taken, the classical sequence of form and then implementation reversed (Whittington, 2001). Thus, HRM can be considered as an emergent, stepwise, iterative approach just as much as a series of grand rational leaps forward (Mintzberg, 1994). This is said to be particularly the case for SMEs as their environment is often too complex or unstable to comprehend or too imposing to defy (Mintzberg & Waters, 1985, p. 271). Processes of strategy formulation are therefore not exercises in applied logic but rather are shaped by cognitive abilities and informational limitations (Mabey & Salaman, 1995, p. 56). By failing to allow for the crucial processes of learning and adaptation best fit prescriptions seem to be wrapped in the straight jacket of classical orthodoxy (Deakins & Freel, 1998; Whittington, 2001). Ultimately in some cases it has been argued that SMEs have no strategy at all and therefore ‘‘if there is no coherent or unified strategy then no HR strategy can be integrated with business strategy’’ (Mabey et al., 1998, p. 509). Some have argued that notion of fit promoted by best fit contributions is a rather static and inappropriate metaphor (Redman & Wilkinson, 2001, p. 12). Situational contingency theorists argue that change is the intervening variable and that one can predict more about HRM by ‘‘knowing the degree of change, than one can by knowing the corporate/business strategy’’ (Stace & Dunphy, 1991, p. 71). In this respect the other popular branch of best fit writings go way some toward accommodating change, in the form of life cycle models (e.g. Baird & Meshoulam, 1988; Greiner, 1988). Yet while serving as useful templates these models largely depict change as a one way sequential process synonymous with growth. The age of a firm however is not necessarily an indicator of growth stage. In practice drivers of change and change efforts will be uneven and complex so that organizations do not follow smooth growth trajectories and so cannot be forced into predetermined stages (Rutherford, Buller, & McMullen, 2003). Indeed a recent empirical investigation of HRM in 2,903 family-owned SMEs indicated vast diversity at perceived different stages therefore concluding that a traditional life cycle was not evident (Rutherford et al., 2003). In a similar vein Baron & Hannon’s (2002) longitudinal research on high-technology start-ups draws attention to the impact of founders’ expectations and ‘mental models’ of proper HR practices (termed ‘organizational blueprints) as impacting the
An Emergent Theory of HRM
111
type, form and sequence of the adoption of HRM. This impact was said to be enduring even after taking account of age, size, access to venture capital, and the economic environment (2002, p. 19, our emphasis). Moreover even firms within industries and pursuing similar strategies were said to have ‘striking differences’ in organizational blueprints (2002, p. 13). This path dependency and potential diversity of approaches animates the potential limitations of over relying on simple ‘matching models’ of HRM. The best fit literature also tends to ignore social and legal norms by depicting a very simplistic and narrow notion of fit (Truss, 2001). Research by McMahon (1996), however, draws attention to how legislation in the form of the minimum wage, EU directives, and taxation can impinge HRM decisions. Clearly, HRM is heavily shaped by contextual contingencies (Boxall & Purcell, 2000). While best fit does acknowledge contingencies it does so in a deterministic way which may not fully capture SMEs proximity to the environment and resultant insecurity and vulnerability to external changes (Bacon et al., 1996; d’Amboise & Muldowney, 1988). Overall dominant models of HRM offer little promise in fully capturing the varied and complex nature of HRM in SMEs. Extant approaches tend to view organizations from a largely closed, rational perspective which captures only part of how HRM systems operate (Ferris et al., 1998, p. 239). Moreover there is an inherent assumption that formal policies must prevail. Consequently, investigation in SMEs presents dilemmas for people using traditional HRM paradigms (Katz, Aldrich, Welbourne, & Williams, 2000). Clearly there is a requirement for accounts which capture emergent processes and more fully embrace broader market forces, societal norms, and institutional settings (Lee, 1997). In an effort to animate the criticisms directed at dominant models of HRM and better accommodate the contextual determinants of HRM in SMEs, an alternative theoretical lens in the form of an emergent, open systems theoretical proposition is presented next.
EMERGENT HRM: A THEORETICAL PROPOSITION Open systems theory emphasizes two important features of organizations: their system characteristics, and their openness to environmental influences (Wright & Snell, 1991, p. 208). Of particular significance is the interdependence between existing internal structures of an organization and the conditions of the environment in which it operates and competes (Jaffe, 2001, p. 209). Recent research has suggested that a complex interaction of
112
BRIAN HARNEY AND TONY DUNDON
internal and external factors shape the parameters of HRM decisions in SMEs (Dundon et al., 1999; McMahon, 1996). An approach under the guise of an emergent open systems proposition may therefore be more appropriate for examining SMEs as they epitomize organizations which rely on flows of personnel, resources, and information from their external environment and are often embedded in dependency relationships with larger capital (Scott, 2001, p. 27). Moreover an emergent, open system approach does not presuppose or assume alignment, but rather by broadening the emphasis from goal achievement per se to survival acknowledges that the ‘‘relationships between organization and environment are variable’’ (Child, 1972, p. 10). In a rationale supportive of the application of an open systems approach to HRM Williamson and Cable (2003, p. 349) have argued that ‘‘social environmental factors should be incorporated into human resource management theory’’. A more macro level perspective captures the determinants of HRM as it recognizes that firms are embedded in a web of social, political, and economic relationships (Edwards et al., 2003; Ferris et al., 1998). Appreciation of environmental interdependency allows for consideration of the numerous factors simultaneously at play in organizations. Given that small firms have less control over their environment than larger organizations, research which considers SMEs in isolation is ultimately misleading (Barrett & Rainnie, 2002; Cassell et al., 2002). Indeed, it has been argued that HR issues are part of open systems, and consideration is theoretically bankrupt unless placed in the broader context of organizations (Zedeck & Cascio, 1984, p. 463). Similarly, others argue the importance of locating analysis in relation to wider macro factors that may shape managerial action in a micro-context, thereby acknowledging the tensions between external pressures and internal HRM structures, policies, and practices (Kinnie et al., 1999; Dundon et al., 2001). Furthermore, strategic exploitation of external structural conditions may form the basis of managerial levers of control when managing paid employment in the SME context (Barrett & Rainnie, 2002). Previous literature has cited the ‘exciting prospects’ that a perspective grounded in organizational theory offers to further advance our understanding of HRM (Wright & Snell, 1991; Katz et al., 2000; Williamson & Cable, 2003). In order to accommodate the influence of contextual variables it has been suggested that drawing on an open systems approach in the form of both institutional and resource dependency theories helps to facilitate the linkage between external influences and their associated HRM effects as well as capturing the dynamics of the process by which organizations adapt (DiMaggio & Powell, 1983; Paauwe & Boselie, 2003).
An Emergent Theory of HRM
113
Institutional Theory and Resource Dependency Theory Institutional theory depicts normative and isomorphic pressures that arise from social and economic inter-relations among firms and in so doing provides a thorough base for inclusion of the context of HRM (Paauwe, 2004, p. 5). Specifically, institutional and political forces coupled with the ‘table stake’ requirements necessary to compete in an industry mean that specific HRM practices are introduced or imposed not as a direct result of market forces but rather as legitimacy enhancing actions (Di Maggio & Powell, 1983; Kinnie et al., 1999; Wright & McMahon, 1992). Thus, HRM may be introduced in an effort to seek approval from a subordinate entity or alternatively HRM practices maybe imposed coercively as mandated by legislation, or as a consequence of pressures exerted by other organizations. This provides explanation for practices that are not the result of predetermined strategies but rather derive from institutional and political influences (Wright & McMahon, 1992; Jaffe, 2001, p. 227; Godard, 2004). Complimentary to this approach a resource dependency perspective captures the dynamics of power relations inherent within ‘the political economy’ of SMEs as they experience pressures exerted by larger suppliers or dominant customers (Imrie, 1986; Katz et al., 2000; Rainnie, 1989). By focusing on the nature of research exchange it is evident that HRM initiatives can reflect the distribution of power and dynamics of the system within which the organization operates (Fuller & Moran, 2001; Jackson & Schuler, 1995, p. 10). Research in SMEs has invariably hinted at the power that can be exerted by various stakeholders: suppliers, large customers, employees, and owner–managers themselves (see, for example, typologies offered by Rainnie, 1989; Goss, 1991). In extreme cases the adoption of new practices can be imposed from outside the immediate work environment (Cassell et al., 2002). Appreciation of such issues negates the criticism directed at HRM frameworks whose failings often derive from exaggerated conceptions of strategic choice (Purcell, 1993). To date, however, inter-firm relationship effects and blurred organizational boundaries have been largely excluded from explorations of HRM issues (Rubery, Earnshaw, Marchington, & Vincent, 2002). Evidently ‘‘we need theoretical models and accompanying research design that take into account the institutional setting and allow reality to emerge and enable us to analyze the underlying processes’’ (Paauwe, 2004, p. 69). In an effort to animate the criticism directed at normative models of HRM the following section presents a conceptual framework as a contextual lens which attempts to accommodate the heterogeneity of HRM by detailing the determinants of HRM in an SME context.
114
BRIAN HARNEY AND TONY DUNDON
EMERGENT HRM: A CONCEPTUAL FRAMEWORK While research has indicated that a plethora of variables that may shape HRM in an SME context such arguments have not been underpinned by any sort of integrative theoretical framework (Barrett & Rainnie, 2002, p. 416; Fuller & Moran, 2001; Wilkinson, 1999). Scase argues ‘‘evidence is so fragmentary and theories so empirically specific that there are severe limitations to the development of cumulative theory’’ (Scase, 1996, p. 580). In terms of HRM in SMEs some have looked at individual practices; recruitment (Carroll et al., 1999); training (Brand & Bax, 2002); new management techniques (Bacon et al., 1996); employee relations (Matlay, 1999); or specific contextual features such as non-union establishments (Dundon et al., 1999); relations with larger firms (Kinnie et al., 1999); owner manager characteristics (Entrialgo, 2002) and social networks (Ram, 1991).1 Research attempts have focused on the determinants and aspects of HRM in SMEs in isolation at the expense of a more holistic perspective which addresses the whole domain of HRM in context (notable exceptions include Arthur & Hendry, 1992; McMahon, 1996). The conceptual framework presented in Fig. 1 is an effort to provide a more holistic perspective for considering the determinants of HRM in SMEs. The framework accommodates the complex interaction of variables that may serve to influence HRM in an SME context (Duberley & Walley, 1995). The emergent theoretical underpinning animates the criticisms directed at dominant models of HRM, while at the same time provides a solid base for understanding how and why organizations have adopted various aspects of HRM. Such an approach is by no means novel but rather it draws on historic perspectives from strategy and industrial relations (e.g. Dunlop, 1958; Kochan, Katz, & McKersie, 1986; Pettigrew, 1985). These approaches similarly emphasize the importance of institutional and environmental forces, the role of strategic choice, and the role of values and ideologies in shaping HRM policies. This form of modeling also echoes similar work by Lewin, Feuille, Kochan, and Delaney (1988), MacDuffie (1995), and more recently Paauwe (2004). The framework facilitates a move beyond rational and planned approaches to HRM toward emergent processes which embrace both market and institutional factors. An inherent part of an emergent open systems approach is an appreciation of the dynamics of change and the neglected processes through which HRM outcomes emerge (Arthur & Hendry, 1992). This provides a mechanism to capture issues of survivability and adaptability and the more immediate objectives of managing paid employment in the form of labor
FEEDBACK
PRODUCT/ MARKET STRUCTURE Marchington and Parker, 1990; McMahon, 1996
REALISED HRM PRACTICES
HR SUPPLY McMahon 1996;Wilkinson 1999; Carroll 1999
HRM INDUSTRY SECTOR
Emergent HRM
SME
Curran and Stanworth, 1979,1981
An Emergent Theory of HRM
EXTERNAL INFLUENCES (DEPENDENCE)
Labor Productivity Flexibility Social Legitimacy (Boxall and Purcell 2003)
VALUE CHAIN
INTERNAL INFLUENCES
McMahon,1996; Rainnie 1989; Kinnie et al., 1999
TECHNOLOGY
-
Gunnigle and Moore, 1992
OWNERSHIP MANAGEMENT STYLE/IDEOLOGY
Informality/Formality of Relations (Bacon et al., 1996; Carroll et al., 1999; Ram et al., 2001)
(Baron & Hannon, 2002; Cassell et al., 2002; Dundon et al., 2001; Entrialgo 2002; Goss, 1991; Matlay, 1999; Rainnie 1989)
LEGISLATION McMahon, 1996; Osterman1994; Hortsman 1999
- TRADE UNION PRESENCE (Dundon et al 2001; Gunnigle and Brady, 1984; Rainnie 1989; Wager 1998)
- SIZE (Kaman et al., 2001; Little, 1986; Robinson and Pearce, 1984)
-
EMPLOYEEs (Guest, 1999, 2001; Nadler and Tushman, 1997)
115
Fig. 1. Conceptual Framework of Factors Determining HRM in SMEs. Source: The skeletal frame is based on Arthur and Hendry’s (1992) model, supplemented from similar propositions from McMahon (1996) and more recently Cassell et al.(2002).
116
BRIAN HARNEY AND TONY DUNDON
productivity, organizational flexibility and societal legitimacy (Boxall & Purcell, 2003). It is important to note that while the framework emphasizes that external structural factors shape the parameters of HRM it suggests the actual form HRM takes is likely to be contingent on idiosyncratic firm responses. This is an important qualification against the criticisms of environmental determinism sometimes directed at open systems theory, as it draws attention toward Child’s (1972) notion of strategic choice and more recently Paauwe’s (2004) concept of ‘leeway’. Overall, the purpose therefore is not to debate the existence of choice but rather the conditions that enlarge or restrict its breadth (Mintzberg, 1994). In this way the analysis should aid in capturing complexities while developing a more integrated approach to understanding why HRM in SMEs has been noted for its ‘marked heterogeneity, complexity and high unevenness’ (Duberley & Walley, 1995; McMahon, 1996; Cassell et al., 2002). The validity of the emergent conceptual framework stems from the fact that it graphically depicts and surmises existing research contributions in this area. Specifically, the conceptual framework facilitates moving beyond size determinism by indicating how firm size (Kaman et al., 2001; Little, 1986; Robinson & Pearce, 1984) interacts with other factors such as labor and product market influences (Curran & Stanworth, 1979, 1981), technology (Gunnigle & Moore, 1992), legislation (Hortsman, 1999) ownership and managerial style (Dundon et al., 2001), trade union presence (Wagar, 1998), employee attributes (Guest, 2001; Nadler & Tushman, 1997) and dependency and relations with customers and suppliers (Marchington & Parker, 1990) to shape HRM. It is important to acknowledge that while the internal and external factors identified are identical to those affecting large firms, given SMEs proximity to the external environment, it is ‘‘the way in which these factors impact on small firms that makes the situation for small firms different from that of large firms’’ (McMahon, 1995, p. 199, original emphasis). Research on HRM in SMEs has shown that typically people issues have top priority in SMEs as they play a vital role in sustaining their competitive advantage (Brand & Bax, 2002). Yet, in general, SMEs face resource deficiencies of time, finance, and expertise which is thought to inhibit the use of sophisticated management strategies, appointment of HR specialists, and the development of unions (Storey, 1994). Resource constraints often mean SMEs spend more time adjusting to turbulence than predicting or controlling it (d’Amboise & Muldowney, 1988, p. 227). Consequently, SMEs are often noted for their direct and informal approach to people management. Even where formalization does occur this is often alongside the informal organic nature of management in small organizations (Bacon et al., 1996).
An Emergent Theory of HRM
117
Methodologically, it has been argued ‘‘in order to formulate meaningful theories y common conceptual frameworks must be established and researchers must examine total organisations as opposed to parts’’ (d’Amboise & Muldowney, 1988, p. 237 our emphasis). This presents the rationale for the current study. To advance our understanding we therefore need to understand not the form HRM takes in SMEs per se, but rather why it takes that particular form. In applying the conceptual framework it will therefore be necessary to first detail the nature and form of HR practices adopted in the sample of SMEs. Following on from this depth of analysis will stem from considering the role of external structural factors in shaping the parameters of HRM decisions prior to considering the role internal influences in determining the actual form HRM takes. Crucially, in operationalizing the framework the focus is on understanding and explaining rather than prediction, thereby appreciating that the nature of the interactions between open systems can create novel and unpredictable patterns (Fuller & Moran, 2001; Truss, 2002, p. 20).
RESEARCH METHODOLOGY Given the inherent limitations of HRMs dominant research paradigm a case study approach was chosen as the most appropriate method for exploring what was contextually unique in SMEs. The objective of the case study method was to enable a more holistic understanding of the critical contingencies influencing the nature of work and employment in a smaller social setting, whether labeled HRM or not. Our emphasis was on understanding behavior from participants’ own frame of reference thereby seeking insight and meaning rather than patterned regularities of behavior (Legge, 2001).The research was exploratory in nature given the proposed challenge to existing theory and the dearth of research in an SME context. The conceptual framework (Fig. 1) with its emergent, open systems theoretical proposition served as a sensitizing framework in capturing the parameters, dynamics, and determining factors of employment and work in SMEs. Critically, the purpose of exploratory research informed by this framework was not to derive formal hypothesis for future testing but rather to garnish an understanding of what was contextually unique and why it was so (Brewster, 1999). Guided by the conceptual framework research instruments focused on the extent to which structural factors and internal influences characterized how and why SMEs have adopted aspects of HRM. To this end the conceptual
118
BRIAN HARNEY AND TONY DUNDON
framework was operationalized by considering four key areas The type of HRM practices being used and their degree of sophistication. This was done by means of a ‘twenty point profile of HRM in SMEs’. The nature and extent of strategic, planned and emergent HRM decisions influencing the development, and adoption of aspects of HRM. ‘Dependence’–determinants of HRM in terms of environmental factors; and, ‘Internal Influences’–determinants of HRM in terms of selected organizational characteristics. Careful attention was given to operationalizing variables in an ‘SME friendly’ way (Cassell et al., 2002). In order to facilitate this we conducted a series of exploratory interviews with two owners of SMEs to explore key trends and issues impacting upon their management of HRM. Informed by these considerations we attempted to build on and extend research instruments utilized in previous research on SMEs (e.g. Bacon et al., 1996; Cassells et al., 2002; Dundon et al., 2001; McMahon, 1996). The twenty point profile is based on the 11 HRM practices in SMEs researched by Bacon et al. (1996) amended by Dundon et al. (2001) for their research into the use of ‘new management techniques’ in SMEs, with the addition of outsourcing given its contemporary significance. In examining HRM grouped functionally as recruitment and selection, training and development, performance appraisal and reward mechanisms we follow a rationale akin to that of Cassell et al. (2002) and Hornsby and Kurakto (1990). Reflecting the emergent, theoretical underpinning multiple embedded cases were utilized in order to provide variety so that the nature of work, and in particular normative models of HRM, could be examined in a range of contexts. Definitions of SMEs was taken from the current European definition (DTI, 2001): micro (less than 10 employees), small (10–49 employees), and medium-sized enterprises (50–249 employees). The logic of case selection related to the open systems theoretical underpinning of the research in that the sample of case studies was purposefully random and dissimilar. It is important to note however, that the majority of the cases fall under the medium-sized firm category. Thus while there is clearly a need for research into SMEs, the use of this term as an aggregate category may in itself be problematic (Cardon & Stevens, 2004, p. 299). Reflecting our critique and theoretical approach HRM was defined broadly as the management of employment, subject to a number of context-specific, contingent choices (Boxall & Purcell, 2000).
An Emergent Theory of HRM
119
Data was collected from 18 case studies in the Republic of Ireland between August 2002 and November 2003. Detailed interviews were conducted with managers with responsibility for HRM in all of the companies. As a result of resource and access constraints the time spent and numbers of people interviewed naturally varied. At 11 of the companies (Nursing Co, Manu Co, Packaging Co, Property Co, Ent Co, Law Inc, Hotel Co, Airline, Clinic Co, Recycle Co, and Sports Center) multiple interviews took place. In seven of the case studies (Property Co, Sports Center, Manu Co, Hotel Co, Packaging Co, Nursing Co, and Airline) multiple site visits were conducted and owners/managers and employees interviewed on more than one occasion. Workers were interviewed in only 10 companies owing to access difficulties. In total, 41 people were interviewed at various stages in the research project. In addition documentary evidence (where made available or in existence) such as mission statements, personnel handbooks/policies were examined. Details and contextual information on the 18 case study companies as well as interview subjects is provided in Table 1.
FINDINGS 20 Point Profile of HRM in Use Initial analysis of the range of practices reported by managerial respondents at the case study companies immediately suggests heterogeneity in terms of the HRM practices in use (see Table 2). Ten of the companies had less than half of the 20 practices in use and among these companies many practices were implemented on a relatively informal basis (Clinic Co, ICT Co, Manu Co, Packaging Co). Clinic Co, for example, had only six of the practices on the 20-point profile and most of these were implemented informally. This evidence does not lend support to Bacon’s ‘beyond bleak house’ argument (Bacon et al., 1996). Of those companies exhibiting a larger range and sophistication of practices (i.e. devolved management, team working, performance appraisal, psychometric testing) these seemed to reflect the service nature of their interactions and the imperative of quality control for customer (client) satisfaction (Hotel Co, Airline Co, Law Inc, Leisure Co, Property Co). Nature and Extent of HRM Planning: Emergent HRM Each of the case study companies had a unique way of planning for HRMrelated issues. In none of the cases did this involve an explicit HR strategy.
Organization
Airline
Ent Co
Background Information on the Case Study Companies. Locus of HR Decisions
Unionized
Passenger and Freight Travel
HR Manager and GM
Yes
27
Medical and Professional Services
Owner– Manager
Yes
190
Retail Services
HR Manager
No
135
Industrial Sector
Background/Market Context
Interview Respondents
Founded in the early 1970s as a HR Manager Clerical Operative one plane entity serving a Pilot limited market under its owner manager Airline has expanded rapidly into European routes. Reflecting this a HR Department was set up in 2001 to formalize policies Owner Manager Clinic Co is run by one owner Nurse manager who established the Receptionist company in 1992. The organization focuses on a niche market. Turnover is very low. There is no HR department or HR policies. People management activities are dealt with by the owner manager Established in 2000 Ent Co is a HR Manager Bar Staff state of the art leisure center, providing conference facilities as well as two restaurants, a bar and a nightclub. It is a non-unionized firm with relatively sophisticated HRM policies although the informal
BRIAN HARNEY AND TONY DUNDON
Clinic Co
No of Employees
120
Table 1.
185
Hotel/Retail Services
HR Manager
No
ICT Co
20
Technology Solutions
Financial Manager
No
Law Inc
210
Professional Services HR Manager
No
An Emergent Theory of HRM
Hotel Co
nature of relations is emphasized. Ent Co employs a large number of part time staff on a seasonal basis HR Manager Front Hotel Co is a luxurious three Desk Staff star hotel re-opened in the Waitress Bar Man mid 1990s by two owners. A HR department was recently established. Hotel Co exhibits relatively sophisticated HRM policies but suffers from high labor turnover (ca. 15%) ICT Co is privately owned and Financial Manager commenced operations in 1995 providing Technology Solutions to multinational clients. It is currently experiencing intense competition. HR issues are dealt with by the Financial Manager. The company has no formally documented HR policies, even in areas such as health and safety Law Inc is a partnership of HR Manager Senior lawyers established in the Partner 1980s by two owners. The company has a HR department and the HR manager is a member of the firm’s management committee
121
122
Table 1. (Continued ) Organization
Manu Co
Industrial Sector
Locus of HR Decisions
Unionized
72
Engineering Manufacturing
General Manager
Yes
106
Medical Device Manufacturing
HR Manager
No
Background/Market Context
Interview Respondents
General Manager Manu Co which produces Machine engineering products such as Operative Clerical gas cutting torches initiated a Staff turnaround strategy in 1999 as the owner manager stepped down and a new GM was hired from outside. The change program has involved job losses and the introduction of automated technology, as well as attempts to reduce union power HR Manager Founded in 1966 Medical Co started as a research and design medical devices provider. The company has a HR department that is staffed by a HR manager and an assistant. HR policy is developed with the board of management at meetings every Friday. The company has relatively formalized HR policies and practices
BRIAN HARNEY AND TONY DUNDON
Medical Co
No of Employees
210
HR Manager
Medical and Professional Services
HR Manager
Yes
Set up in the 1970s Nursing Co is privately owned limited company caring for elderly patients. The company is unionized and has a large number of relief staff and increasingly employs nonnationals on a fixed term contract basis
Owner Manager Packaging Co was founded by Machine its present owner in 1975. Operator(x2) Despite progression recent R&D Expert times have seen intense Secretary competition resulting in layoffs and attempts at diversification. The owner manager has retained familial control over all issues and this has resulted in loyalty exhibited by the low labor turnover General Manager PC Co remanufactures and remarkets computer products to sell in the European market place. The company employs a part-time HR manager and has formal policies as well as an explicit policy of nonunion recognition
Packaging Co
65
Print Supplies and Manufacturing
Owner– Manager
No
PC Co
80
Technology Manufacturing
HR Manager and General Manager
No
An Emergent Theory of HRM
Nursing Co
123
124
Table 1. (Continued ) No of Employees
Property Co
160
QualTech
Recycle Co
Industrial Sector
Locus of HR Decisions
Unionized
Property Management/ Consultancy
HR Manager
No
120
Service Provider
Director
No
203
Waste Management Solutions
HR Manager
No
Background/Market Context
Interview Respondents
HR Manager Property Co has 50 years Property experience in dealing in Consultant Office commercial, home and Staff professional services. It experienced dramatic growth in late 1990s, and is now market leader. A new HR department has recently been established Established in 1988 QualTech is Director a value added service provider for the telecommunications and high technology industries. The company serves a range of customer sectors. The company has no HR department and very few documented policies HR Manager Truck Established in 1995 by four Driver entrepreneurs Recycle co has Maintenance Staff grown through acquisition and now has eight recycling facilities around the country serving both domestic and commercial customers. The company has an explicit nonunion policy and is in the process of centralizing and formalizing its HR policies
BRIAN HARNEY AND TONY DUNDON
Organization
200
Sports/Leisure
HR Manager
No
Temp Co
185
Manufacturing Temperature Solutions Provider
CEO/Line Managers
No
Telecom Co
143
Telecommunications
Financial Controller
No
HR Manager Originally established as one Purchasing Staff shop by two brothers in 1983 Sales Staff Leisure Co has grown rapidly by acquisition and now has 20 shops across a large geographical spread. Leisure Co is attempting to formalize and develop HR practices CEO Line Manager The company was founded in 1983 by an Irish American and deals with temperature control measurement specifically in the medical, aerospace and communication industries. The company has entirely devolved HRM to line managers although policy decisions are made by the CEO Telecom Co is a privately owned Financial Controller Irish company providing cable assembly, box build and end-to-end telecommunication solutions. Subcontracting is a growth area. The company has limited HR policies with the financial controller managing HR issues
An Emergent Theory of HRM
Sports Center
125
126
Table 1. (Continued ) Organization
Waste Co
No of Employees
Industrial Sector
Locus of HR Decisions
Unionized
150
Waste Management Services
HR Manager
Yes
Background/Market Context
Interview Respondents
HR Manager Founded in 1978 as family venture Waste Co has evolved into an expert in waste management and renewable energy. The company has benefited from government policy and growth in environmental concerns and has recently hired a HR Manager
BRIAN HARNEY AND TONY DUNDON
Twenty Point Profile of HRM in the Case Study Companies.
Airline
Clinic Co
Ent Co
Hotel Co
ICT Co
Law Inc
Manu Co
Medical Nursing Packag- PC Co Co Co ing Co
Property Co
Qual Tech
Recycle Co
Sports Tele Centre Com
Temp Co
Waste Co
(1) Cultural change program (2) Devolved management (3) Teamworking (4) Performance appraisal (5) Mission statement (6) Team briefing (7) Quality circles (8) Harmonized terms and conditions (9) Psychometric tests (10) De-layering (11) Increased job flexibility (12) Customer quality schemes (13) Training programs for all employees (14) Staff suggestion schemes (15) Company wide meetings (16) Staff attitude surveys (17) Employee welfare (18) Family friendly working
O
X
X
O
X
X
INF
INF
X
X
X
O
X
X
INF
X
X
INF
O
X
X
O
X
O
X
X
X
X
X
O
X
O
X
O
O
X
O O
X INF
O O
O O
O INF
O O
O INF
O O
O O
X INF
O INF
O O
O INF
X INF
X O
X O
X O
X O
O
O
O
O
X
O
X
O
O
O
O
O
O
O
O
X
O
O
O X X
X X X
X X O
X O X
O O X
O X X
X X X
O O X
X X O
X X X
O O X
X X O
X X O
X X X
INF O O
X X O
X X X
X X O
O
X
X
O
X
X
X
X
X
X
X
X
X
X
X
X
X
O
X O
X INF
X O
X O
X X
O X
X X
X X
X O
X O
X O
X X
X O
X O
X O
X O
O O
X X
O
X
O
O
X
O
O
X
X
O
X
O
O
O
O
O
O
O
O
X
O
O
O
O
O
O
O
O
O
O
O
X
O
O
O
O
X
INF
INF
O
INF
INF
INF
INF
X
INF
INF
X
INF
INF
X
O
O
O
X
INF
X
X
O
X
X
O
X
X
O
X
X
X
INF
X
X
X
O
X
O
O
X
X
X
X
X
O
X
X
X
X
X
X
X
O
X
X
O
O
X
X
X
X
X
X
X
O
X
X
X
X
X
X
O
X
X
O
X
O
X
X
X
O
X
X
X
X
INF
X
X
O
127
Practice
An Emergent Theory of HRM
Table 2.
128
Table 2. (Continued ) Practice
Airline
Clinic Co
Ent Co
Hotel Co
ICT Co
Law Inc
Manu Co
Medical Nursing Packag- PC Co Co Co ing Co
(19) Social events (20) Outsourced practices
INF O
INF X
INF X
O O
INF X
O O
INF O
INF X
O O
INF X
O O
Property Co
Qual Tech
Recycle Co
Sports Tele Centre Com
Temp Co
Waste Co
INF O
INF X
O X
INF X
INF X
O O
INF O
Note: O ¼ Practice in Use, X ¼ Practice not in Use, INF ¼ Implemented on an informal basis.
BRIAN HARNEY AND TONY DUNDON
An Emergent Theory of HRM
129
Indeed the findings indicate that relatively little formal strategic planning took place. At Manu Co, PC Co, and QualTech, for example, strategic planning consisted of brief and infrequent chats between the GM/Director and the Financial Controller. Packaging Co did not undertake any formal planning. The owner manager attributed this to pressing nature of immediate business concerns; ‘‘I’m too busy trying to keep our heads above water’’. Similarly at Recycle Co the HR manager commented: I suppose we’re fire fighting at the moment, we’re not getting the chance to look forward to see what we want. For the moment our policies are quite disjointed (HR Manager – Recycle Co).
The reactive nature of HRM attempts was captured by the Director of QualTech who noted HR issues are dealt with as they arise, this was also said to be the case at TeleCom. This unstructured, ad hoc and reactive nature of planning seriously calls into question the best fit arguments. Put simply, there is no significant evidence of a coherent strategic approach upon which HR can be integrated. There were exceptions to this pragmatic approach however. At Medical Co the HR manager attended weekly management meetings while at Law Inc and Airline Co HR managers had seats on management committees albeit in a ‘co-opted capacity’ with no direct impact on decisions. More integrated decision-making occurred at Hotel Co as HRM policy decisions were said to reflect the direct input from line managers. At ICT Co and Clinic Co strategic planning initiatives were minimal to the extent that they were seen as inherent part of the owner managers’ day-to-day efforts to manage the business. A common feature was the terminology employed by the managers which suggested a focus on survivability and dependability in terms of HRM decisions. The owner manager of Packaging Co used the analogy of ‘plodding along’ noting we might start off pro-active but always end up reactive. Despite Property Co’s recent efforts at consolidating and formalizing procedures for HRM practices recruitment was still described as ‘haphazard’. Likewise although training was increasingly essential at Sports Center the HR Manager commented that training schemes were ‘fragmented and uneven between stores’. At Manu Co HRM issues were dealt with as they emerged or warranted attention. Such connotations of survival, adaptation, and subsequent minimal HRM efforts seem to replicate findings in other areas (Hill & Wright, 2001). SMEs therefore seem to spend more time adjusting to turbulence than trying to predict it or control it (d’Amboise & Muldowney, 1988). This was seen to be the case not merely for those firms exhibiting relatively little strategic planning but also for those with more structured and formal
130
BRIAN HARNEY AND TONY DUNDON
strategic planning initiatives. This is exemplified at Hotel Co where it was found that HR planning was more complex and contradictory, at times forward looking and strategic in nature (e.g. performance appraisal) although also reactive as issues emerged or gaps were identified (recruitment and training). Documenting the contradictory and dynamic nature HRM planning evident at the case study companies sets the context for a more informed consideration of the heterogeneity of specific HRM practices. Recruitment and Selection In terms of recruitment the findings suggested that most of the companies utilized informal practices. Informal word of mouth recruitment was seen as ‘reliable’ at Packaging Co and cost effective at ICT Co: the first port of call for a company of this size is always informal contacts. This radically reduces our recruitment costs (Financial Manager, ICT Co). While some companies had adopted a more systematic approach to recruitment, including psychometric testing (Hotel Co, Ent Co, Waste Co) these methods were complimented by more informal approaches for lower level staff, for example, the HR Manager at Hotel Co noted employees also bring their friends with them. Other approaches can only be understood with respect to specific labor market contingencies. Those companies requiring a more specific skill set tended to use specialist agencies (PC Co, Medical Co, and Property Co). At Law Inc the more formalized and structured approach to recruitment was a reflection of the tight labor market for specialized tax expertise to represent client companies competing in the UK and US markets. Other companies noted that they also received ‘CV’s on spec’ as a result of the seasonal nature of the industry (Hotel Co, Ent Co, Sports Center) or as a consequence of an established reputation (Airline). Training and Development While all but two (Clinic Co, Recycle Co) of the companies indicated that they had training schemes for all staff, the level and extent of such training was highly uneven across the case study companies. At Airline, Hotel Co, and Property Co significant resources had been devoted to new areas of quality and customer service training while extensive ‘in house’ training programs were available for graduates. Similarly PC Co had a detailed induction program for new recruits while Law Inc offered monthly ‘soft skills development’ programs and was devoted to facilitating Continuous Professional Development. At the other end of the spectrum at Manu Co, Packaging Co, and Recycle Co training was less structured determined ‘on the job’ as skills were generally easy to pick up yourself (General Manager (GM) – Manu Co).
An Emergent Theory of HRM
131
At Clinic Co there was no training provided what so ever as the owner manager relied on external agencies such as the Irish Nurses Organization (INO) to ensure that nurses were fully qualified and trained. Performance Appraisal and Reward Mechanisms At Airline, Hotel Co, Property Co, Law Inc, and Telecom Co performance appraisals were conducted annually, with employees eligible for rewards and bonuses based on their performances. Reflecting a more informal approach toward HRM at Manu Co, Packaging Co, PC Co, and Temp Co there was a tendency for ‘on-going systems of appraisals’ (PC Co) or what the GM of Manu Co referred to as ‘continuous conversations’. Such approaches involved appraisal as a regulative mechanism rather than as a developmental tool explicitly evident in the discourse of management as appraisal at Packaging Co took the form of ‘mid course corrections’ (Owner – Packaging Co). Similarly at Temp Co the aim of appraisal was to eliminate the fat and waste and become more productive (Line Manager – Temp Co). In the more extreme cases such as Clinic Co the owner manager enthusiastically noted that there was ‘no need’ for appraisal systems as this was an implicit part of his day to day activities while at Recycle Co the only notion of performance appraisal evident was managers come to me because someone is not working out, and we can get rid of them or whatever (HR Manager – Recycle Co). Yet although in many cases formal appraisals were largely absent that is not to say employees who worked particularly hard or were in need of motivation were not rewarded. Informal methods of linking individual remuneration to performance included Christmas Bonuses (Packaging Co, PC Co) special staff discounts (Ent Co) or being taken out for dinner (Manu Co). The rationale for this undercurrent of informality was provided by the GM of Manu Co who noted: informality makes things feasible and gives you options. Furthermore the HR Manager of Ent Co commented often formal methods make employees uncomfortable. Overall it was evident that certain HRM practices were deemed as simply not viable or necessary at some of the case study companies (Clinic Co, Ent Co, Hotel Co, ICT Co, Manu Co, Nursing Co, Packaging Co, Recycle Co, Sports Center, and Temp Co). These arguments surfaced in particular with respect to attitude surveys, sophisticated recruitment methods and formal company wide meetings. These sentiments were probably best captured by the GM of Manu Co you have to keep HRM simple, if I was to follow what was prescribed I would be out of Business. Although useful the inadequacies of such surface level reporting by managers must be acknowledged. Taking into account more qualitative
132
BRIAN HARNEY AND TONY DUNDON
dimensions the way such policies were implemented was at times found wanting. At both Hotel Co, and Waste Co, for example, psychometric testing had been utilized on only one occasion for a managerial level position. At Law Inc, Property Co, PC Co, and Airline Co teams had very little autonomy in terms of deciding how to perform a given task or input into the appointment a team leader. Furthermore, there was evidence of work intensification at Airline and Property Co as a result of dramatic growth and extended work rotas. Dissatisfaction was becoming manifest at Hotel Co and Sports Center through high levels of employee turnover (ca. 15%) as employees often moved to rival firms. In contrast despite the dearth of formalized HRM practices in use at Clinic Co and Packaging Co, and to a lesser extent Manu Co, staff turnover was minimal and employees reported that they were loyal to their respective companies. This serves to undermine simplistic categorizations which attempt to predict the nature and impact of HRM in SMEs based on checklist profiles of HRM practices. In order to appreciate such dynamics and contradictions it is necessary to appreciate how structural factors both directly outside and inside of the work milieu have shaped the parameters of HRM decisions. Dependence (External Factors) Product/Market Structure The competitive market forces faced by the case study companies heavily informed the extent and type of HRM adopted. Those companies operating at the more mature stage of their life cycle were seen to have a dedicated emphasis on controlling costs and increasing productivity (Manu Co, Packaging Co, PC Co). Packaging Co, for example, experienced extreme cost pressures in its main ‘bread and butter low margin’ market and had been forced to initiate layoffs for the first time. At Manu Co intensified competition had sparked investment in technology and the introduction of strict job descriptions. Many of the companies experiencing rapid growth had attempted, with varying degrees of success, to consolidate and formalize HR policies (Airline Co, Law Inc, Sports Center, Waste Co). The HR Manager at Airline Co noted growth and expansion demanded full time HR positions while Sports Center began to develop a clear HR agenda as a consequence of growth by acquisition and the subsequent need to ‘amalgamate various micro businesses’. Growth trajectories were not necessarily smooth, however, as Airline suffered from global economic pressures in air travel, particularly the proliferation of ‘budget airlines and on-line booking’ technologies. Furthermore the experience of rapid growth at Recycle Co
An Emergent Theory of HRM
133
and ICT Co involved extreme competitive pressures in their relatively new markets which meant that capital and infrastructural investments took precedence; It’s more important that we spend money on the operations level at the moment than spend money on training or HR policies, because we need to update our equipment (HR Manager – Recycle Co).
Industrial Sector Those companies operating in a service capacity tended to take a more proactive approach to meeting market pressures through emphasizing the ‘people side’ as evidenced through staff and customer quality initiatives at Hotel Co, Sports Center, Law Inc, and Property Co. Some of the explanations for discrepancies in terms of the HRM practices developed by each of the companies are therefore attributable to industrial sector. For example, Hotel Co, Property Co and Airline Co, and to a lesser extent Ent Co and Sports Center, are illustrative of high contact services whose attempts at delivering service are dependent on relatively sophisticated and pro-active application of HRM practices. In contrast at manufacturing companies such as ICT Co, Manu Co, Packaging Co, TeleCom, Temp Co the impact of competitive pressures mitigated against investment in HRM, evidenced by the fact that none of these companies had a dedicated HR specialist. Evidently at the case study companies placement along industrial trajectories and market forces were seen to shape the broad parameters of HRM decisions (Barrett & Rainnie, 2002). Closely interrelated to industrial sector and more explicitly leveraged by management, however, was the nature of the labor market as an active determinant of HRM investment decisions. Labor Markets (HR Supply) Many of the case study companies relied primarily on local labor markets where there was a supply of labor readily available (Manu Co, Packaging Co, PC Co, QualTech, Recycle Co, TeleCom, Temp Co, and Waste Co). Such loose and static labor markets served to limit the constraints on managerial autonomy in terms of HRM-related decisions and may aid in understanding the limited investment in HRM practices at these companies. At Manu Co these conditions were explicitly leveraged by the GM who emphasized the availability of replacement labor through deskilled ‘easy to pick jobs’ which kept employees ‘on their toes’. Similarly at Recycle Co the HR Manager pointed to the ‘limited skills set required’ due to the simplistic nature of the business model. Despite the fact that Hotel Co and Airline Co also seemed to have a readily available supply of national and non-national
134
BRIAN HARNEY AND TONY DUNDON
labor, they did exhibit more investment in sophisticated HRM. This may be attributable to the high contact nature of the service offered which meant that at Airline cabin crew employees required more training in terms of customer service skills and quality initiatives, as was the case for staff at Hotel Co. Those companies operating in specific segments of the services sector were subject to, and dependent on, the irregularities of transient labor market and product markets (Ent Co, Sports Center, and Hotel Co). The impact of this became manifest in a large staff churn. The HR Manager of Sports Center, particularly reliant on a student workforce noted how it was difficult to retain employees; they use us for two years and then leave. Hotel Co had other difficulties particular to the sector, with trainee managers being poached by rival hotels. The HR Manager at Hotel Co summed-up the general situation: we do suffer, they go to other hotels, mainly for money. More specialized companies such as Law Inc, Medical Co, Property Co had occasional difficulties attracting skilled labor amidst buoyant markets given the small pool from which to choose from and so utilized external agencies. At PC Co and Medical Co given the more technical and specialist nature of jobs recruitment of engineering and appropriately qualified graduates proved an onerous task as the skills they desired were in short supply. This was also said to the case for Clinic Co and Nursing Co and they turned their focus on foreign labor markets (through recruitment agencies) to supply adequately qualified nurses. Clearly therefore, labor markets were not undifferentiated (Jackson & Schuler, 1995, p. 15). All of the case study companies had experienced difficulties in recruiting and retaining certain staff. This was particularly the case for professional and more specialized, higher salary occupations such as lawyers (Law Inc), pilots (Airline), I.T. specialists (Manu Co, Hotel Co, Medical Co), and nurses (Clinic Co, Nursing Co). Value Chain and Technology None of the case study companies were operating at the leading edge in terms of technology. Thus, although the extent to which firms had invested in technology varied somewhat, technological developments did not directly impact the HRM employed at the case study companies (see also Duberley & Walley, 1995, p. 901). For the most part, the technology employed by the case study companies (and associated HRM effects) related to pressures emanating more indirectly from the firm’s value chain. Specifically, at Medical Co, Packaging Co, and Manu Co dependency on a small number of (large firm) customers serves to render defunct a notion of over exaggerated strategic choice. The GM of Manu Co noted of their dependency on a
An Emergent Theory of HRM
135
particular customer: [large company customer] came and viewed the plant, checking our quality procedures, we had to have everything documented
In response to this Manu Co had introduced new automated machinery in an effort to boost productivity and accountability. Yet it transpired that pressures emanating from the ‘political economy’ of SMEs to implement HRM came not only from specific larger companies but also through managerial attempts to gain legitimacy and consolidate control. This is illustrated in the comment by the HR manager at Hotel Co about its HRM audit in order to obtain the Irish Hotel Federation (IHF) Quality Employer Award. The HR manager explained that this award was ‘‘a have to have to be luxury hotel in an industry driven by image, even though it’s not actually required’’. In some cases HRM practices were seen to be more directly imposed by the immediate business environment. The HR Manager of Recycling Co noted how ‘‘business is dictating to the company what policies are needed’’. At Airline, Manu Co, Property Co, Packaging Co, QualTech, Temp Co, and Waste Co training schemes were devised for employees in order to meet standards for ISO accreditation. The HR manager of Waste Co noted of the ISO accreditation procedure it involved a lot of work in terms of quality management, there were a number of policies, procedures and standards to implement in order to comply. At Airline and Property Co ISO accreditation attempts neatly overlapped with pro-active attempts at consolidating and formalizing HRM. At other companies such as Manu Co, Packaging Co, QualTech, Temp Co, and Waste Co such initiatives were seen to be more imposed and implemented in an ad hoc fashion, not as a strategic measure, but rather tactically as a survival mechanism to meet expectations for quality initiatives as dictated by specific industrial sectors. In this sense, the elements of HRM at each of the case studies were introduced to reflect the dynamics of the system within which each organization operated, akin to the suggestions of Fuller and Moran (2001). Legislation In an Irish context there are approximately 40 pieces of primary legislation relating to employment matters, that must be dealt with, irrespective of whether a company employs 1 or 1,000 people (SFA, 2003). At most of the case studies such legislation was seen as a necessary burden. When questioned about upcoming acts and amendments with specific implications for HRM (e.g. Data Protection Amendment Bill, 2002) HR managers at Property Co, Law Inc, Airline Co and Medical Co exhibited extensive expertise. This
136
BRIAN HARNEY AND TONY DUNDON
knowledge of obligations/rights under various acts, however, was not matched elsewhere as none of the other case study companies were aware of the ramifications of this specific Amendment Bill. Such knowledge deficiencies were attributed to administrative and timing constraints. These sentiments are best reflected by the owner–manager at Packaging Co who expressed a difficulty in keeping track of legislation and a skepticism as to the benefits of doing so: it could all be very positive, but I don’t have time to scratch my ass, never mind all those procedures and paperwork, if it’s not broke why should I fix it.
Similarly, those companies with a reliance on a non-national workforce (Hotel Co, Ent Co, Nursing Co, and Clinic Co) referred to the work permits as a ‘legislative nightmare’. At ICT Co and Waste Co the cost implications of a rise in the national minimum wage was a particular concern. Perhaps more alarming were hints at non-compliance, for example, at TeleCom Co the Financial controller noted that the company had no generic policies – not even in terms of health and safety while at a number of the companies there was no anti-harassment/bullying policy (Clinic Co, ICT Co, Recycle Co, TeleCom, and Temp Co). Further, there was evidence that in some cases taxation and social insurance costs influenced the propensity to recruit in a formal manner. At those companies dealing in low-skilled operations with fluctuating demand, dependent on a local labor market, freelance workers or ‘old reliables’ were drafted in and paid in a ‘tax efficient manner’. This practice was evident at Manu Co, Packaging Co, QualTech, and Recycle Co and to a lesser extent, and usually for more qualified staff, at Hotel Co and Ent Co. At Airline Co employees were ‘encouraged’ to opt out of the working time regulations owing to pressures for extended work rotas. The most extreme stance on the role of legislation was taken by GM at Manu Co who remarked legislation is there to be abused. Clearly external structural factors at the case study companies have had major effects in terms of shaping the parameters of HRM decisions. This highlights the inadequacies of closed system approaches which focus exclusively within SMEs in order to explain HRM (e.g. Goss, 1991) and supports those who advocate a more holistic approach (Duberly & Walley, 1995; Fuller & Moran, 2001; McMahon, 1996). Indeed, Barrett and Rainnie note that the inherent limitation in research on SMEs is ‘‘the partial dislocation of the small firm from its totality as approaches taken focus attention within the small firm’’ (2003, p. 416). Moreover this highlights the problem of divorcing debates about the nature of HRM in SMEs (i.e. small is beautiful vs. bleak house) from the structural conditions that condition and set the parameters of HRM decisions. More attention should therefore be given to the institutional and
An Emergent Theory of HRM
137
political dynamics of the system in which SMEs operate or what Wood (1999) terms ‘environmental fit’ (Wood, 1999) (see, e.g., Gooderham, Nordhaug, & Ringdal, 1999; Lee & Johnson, 1998; Sparrow, 1999). Significantly, this view resonates with the emergent, open systems theoretical proposition. Ultimately, however, while external factors shape the broad parameters for action the actual form HRM will take depends on the ‘internal dynamics’ which determine the unique organization responses to structural factors. Internal Influences (Selected Organizational Characteristics of SMEs) Ownership Arthur and Hendry (1992) contend that varying SME ownership arrangements will yield alternative HRM implications. As one moves from a spectrum from family-ownership, to private and then publicly owned organizations one might expect increasingly sophisticated HRM (Reid & Adams, 2001). De Kok and Uhlaner (2001), however, caution this determinism and suggest far from being homogenous, there is diversity in terms of the HRM practices in use under each ownership arrangement. At the case study companies various ownership structures and management styles clearly influenced the HRM responses to intensified competition. At Packaging Co, Clinic Co, Nursing Co and ICT Co the owner–managers engendered loyalty from employees by maintaining a high degree of informality. Employees, in turn, were offered flexible working terms in the form of ‘banked hours’ (Packaging Co), part-time attendance (Clinic Co) or teambased projects (ICT Co). The banked hours system at Packaging Co was fundamental in meeting customer demands and also served the owner managers’ interests as when they were working late employees often opted for banking hours rather than over time pay. At Clinic Co part-time employees provided greater control over evening attendances while at ICT Co although the informality of team-based project working was stressed this was increasingly being subjected to tighter monitoring. More severe reaction to intensified competition was evident at Airline Co, Manu Co, Hotel Co, and Property Co centering on attempts to develop a more formally regulated workplace. At Manu Co this took the form of automation of processes through technology while at Airline there was a distinct move toward shiftworking. At Property Co HRM practices had been formalized as a mechanism to control both the performance and development of staff. The direct influence of ownership on HRM-related decisions was readily apparent at Clinic Co and Nursing Co. Despite the semi-retired status of the owner manager of Clinic Co staff noted that he was heavily involved in the
138
BRIAN HARNEY AND TONY DUNDON
day-to-day running of the organization. The director at Nursing Co played a similar role as the locus of control in determining and ‘dictating’ policy. In such cases the informality of relations was emphasized. In contrast at the other case study companies the direct influence ownership was less prominent. At Hotel Co, for instance, almost all decision-making powers rested with the HR manager rather than directly with the owners while line managers had responsibility for HRM at Temp Co. The HRM set up at Law Inc and Property Co also allowed for autonomy in terms of HRM decisions, although there were hints that the historical legacy of Partners at Law Inc and founding Directors at Property Co exerted some degree of influence on policy-related decisions. Hotel Co, Property Co, Airline Co, and Law Inc displayed a sophisticated paternalism and a unitarist outlook evident in references to ‘working together’ and creating of a ‘strong happy culture’. How this more sophisticated approach to HRM operated in practice, however, remained questionable as evidence from employees suggested practices were either loosely applied (teamwork – Hotel Co), were seen as ‘token gestures’ (employee representatives structures – Hotel Co, Law Inc) or led to work intensification (Airline, Property Co). At Manu Co the General Manger exercised autonomy on all decision matters it has to be one person in charge otherwise control is an illusion. Managerial Style A number of the companies emphasized the paternalistic nature of their managerial style, albeit in different ways. This paternalism was often associated with informal relations and was explicit in the discourse deployed by managers. At Ent Co, for example, the HR Manager noted that informal communication is crucial in maintaining the family style atmosphere and harmony of the organisation. This ‘family’ analogy was also deployed at PC Co and Recycling Co where the HR Manager referred to the atmosphere as ‘very easy going’. Similarly at Sports Center the HR Manager commented: We are like a family here. The retail sector is not a high stress business. I think this is apparent in the relaxed management style and subsequently this impacts on the atmosphere.
Despite such assertions however, there was an inherent tension at Sports Center between efforts at maintaining this informal management style and attempts to consolidate practices in order to effectively integrate new acquisitions and address the relatively large staff turnover (ca. 15%). At Packaging Co informality and paternalism generated a sense of loyalty and was utilized to ameliorate the harshness of working conditions and long working hours. Yet this surface level harmony of a relaxed social atmosphere and ‘craic’ identified by the employees at Packaging Co was at times juxtaposed
An Emergent Theory of HRM
139
by an authoritarian streak evident in particular from the nickname ascribed to the owner manager as ‘The General’. Thus, informality did not disguise or smooth away the power relations that shape the management process (Ram et al., 2001, p. 846). The owner–manager was clearly the locus of control and decision making, evident in his assertion that ‘the buck stops with me’. Likewise despite the semblance of harmony at Recycle Co and ICT Co competitive pressures had served to alter the dynamics of informality. At Recycle Co managers had become ‘very hands on’ closely monitoring performance and ‘ensuring that things were getting done’. While at ICT Co monitoring of attendance and contribution had moved from a standard clocking card system to being measured in more specific detail by hours worked on particular projects. Informality, therefore, is a matter of degree rather than kind, and operates in a dynamic and context-specific manner (Ram et al., 2001). At Manu Co there was a more explicit intention to reassert managerial prerogative as the GM introduced new automated machines and tight job descriptions to exert closer control and increase productivity. The GM explained that the purpose of these ‘levers of consistency and predictability’ was ‘to take the witchcraft out y stop people hiding behind jobs that weren’t there.’ Less extreme but similar authoritative managerial styles were evident at Nursing Co, QualTech, and Waste Co. At Nursing Co it was noted that the Director ‘dictated all major policies’ and the general ethos of the organization was driven around the dictum ‘stick by the rules and you can’t go wrong.’ At Waste Co a focus on operational efficiency and cost minimization had led to a particularly narrow and regulative view of the people management role. Unionization The manifestation of ownership type and managerial style was clearly apparent in managerial attitudes toward trade unionism and the role of employee voice at the case study companies. At Manu Co ‘dealing with the unions’ was an explicit part of the GM’s attempt to reassert managerial prerogative. Indeed the GM continuously demonized trade unions by attributing the company’s current financial predicament to the negative influence of trade unions commenting: they hold up the place with their rights, what about management rights and responsibilities. This explicit opposition of trade union presence was softened at the edges for employees, however, through use of team working and a staff suggestion schemes which had resulted in a TV and coffee machine being introduced in the canteen. Similar explicit non-union sentiments and ‘policies’ were evident at Ent Co, ICT Co, PC Co, and QualTech. At Clinic Co the owner openly expressed his
140
BRIAN HARNEY AND TONY DUNDON
disapproval for collective representation even though he relied on the nurses union, the Irish Nursing Organization (INO) to supply qualified nurses in times of labor shortage. At Nursing Co and Airline unions were begrudgingly accepted as something that reflected the professional status of nurses and pilots rather than something that directly undermined managerial prerogative. Thus the HR Manager of Airline noted of the pilot’s membership of IMPACT (Irish Municipal, Public, and Civil Trade Union) it’s less about industrial relations but more about the perks of membership such as issues to do with licences. Overall both Airline Co and Nursing Co still maintained an ‘official’ nonunion policy and regarded these specific incidents of union membership as ‘unfortunate features’ of their respective industries. These approaches to unionization resonate with Gall’s (2004) revision of Roy’s (1980) typologies, as indicative of a combination of ‘sweet and awkward stuff’. Clinic Co and Law Inc were more reminiscent of Guest and Hoque’s (1994) ‘lucky’ non-union employer, in that direct representation and recognition had never emerged as an issue among employees. In contrast, at Packaging Co and Recycling Co the evidence points more toward crude ‘fear and evil stuff’ as management were openly anti-union in that collective representation was viewed as an affront to the owner’s prerogative. Justification for this stance was made through a combination of economic arguments; if we were unionized we couldn’t be competitive in the market place (Owner Manager – Packaging Co) and also deeper ideological sentiments; We’re a non-unionized company and we would like to stay that way. I come from a background where it was unionized and it’s more difficult to cope, unions will stick their nose in y restrict freedom and frustrate processes (HR Manager – Recycle Co).
At Hotel Co, Law Inc, and Medical Co, Property Co there was also a disdain for union activity although this became manifest through HRM practices as substitution mechanisms rather than a suppression of collective organization among workers. These ‘substitution mechanisms’ included employee of the month schemes (Hotel Co), extensive training schemes (Law Inc, Hotel Co, Property Co), 3601 feedback appraisal (Property Co), family friendly working (Law Inc, Medical Co, Hotel Co) and employee representative structures (Law Inc, Hotel Co). At Hotel Co the HR Manager explicitly emphasized the need for ‘proactive’ attempts to alleviate the ‘risk of union activity’. Furthermore such approaches were underpinned by a unitarist conception of the organization which emphasized a non-conflictual, familial atmosphere. For example, the HR Manager of
An Emergent Theory of HRM
141
Property Co noted of any attempts at unionization We would not accept it, professionals are treated exceptionally well and receive a huge amount of recognition. We deal with problems and listen to staff; there would be no role for a union.
Attitudes to trade unions were therefore underpinned by various levels of ideological opposition to trade unions as a constraint on managerial legitimacy or control (Airline, Clinic Co, Ent Co, Manu Co, Nursing Co, Packaging Co, PC Co, QualTech, Recycle Co) (Chamberlain, 1948) and/or deemed unnecessary as a result of a unitarist conception of the organization (Clinic Co, Hotel Co, Law Inc, Medical Co, Packaging Co, and Property Co). How these attitudes became manifest in specific approaches and efforts ‘to deal’ with unions, however, can only be made sense of with respect to the organizational context and resource constraints at each company. This issue is elaborated upon below in the context of a consideration of the role of organizational size and the characteristics of size in informing and shaping HRM. Characteristics of Size The findings clearly indicate the heterogeneous nature of HRM across case study companies. Akin to suggestions by Osterman (1994) size therefore cannot be considered as an isolated determinant of the adoption of practices. Looking at the 20-point profile of HRM in SMEs, for example, indicates that two of the firms with the lowest number of HRM initiatives, Nursing Co (8) and Recycle (7) are among the three largest firms, each with over 200 employees. Nonetheless there are a number of characteristics stemming from size which might be seen to play a mediating role in shaping the interplay between internal and external variables and therefore informing the actual form of HRM adopted. Economies of scale and resource constraints meant that in some cases practices were not viable, or even necessary. The Director of QualTech explained the company had few documented HRM policies because of the fundamental nature of the business which does not have critical mass. Further, time constraints and lack of expertise made dealing with administrative issues, particularly legislation, extremely burdensome (this was especially the case at Clinic Co, Nursing Co, Packaging Co, and TeleCom Co). In terms of the proceeding discussion size and resource constraints meant that Manu Co and Packaging Co, and PC Co were limited in that they could not afford the sophisticated union substitution strategies (e.g. employee representative structures) evident at Hotel Co, Property Co, and Medical Co. This draws attention, however, to the ability of these firms to leverage loyalty from employees through informal policies and more direct, flexible
142
BRIAN HARNEY AND TONY DUNDON
management. In this sense, as a result of hierarchically constrained organizational structure, management had more of a direct role in shaping HRM. This proximity to employees is best captured by the GM at Manu Co who commented about attitude surveys there are 70 people working here if I don’t know what the attitude is and need a survey to tell me I have no chance. Packaging Co is also exemplar in this respect, as the authoritative ‘buck stops with me’ stance of the owner manager and poor working conditions were in someway ameliorated by the informal friendly relations. The owner manager noted of relations with employees we roar at each other, slag each other, and tell each other to get lost y after working together for so long is probably natural but not polite. Moreover tinkering around the edges of the flexible ‘banked hours’ work system was tolerated, as the system still served managerial prerogative in meeting fluctuating customer demands, paying less overtime, while at the same time leveraged loyalty and motivation from employees. Informality was therefore utilized to provide some scope in the application of practices. In a similar vein flexible, informal reward systems used on a ‘quiet basis’ including cash bonuses at PC Co, employee discounts schemes at Ent Co or being taken out for dinner at Manu Co served this function. As the GM of Manu Co noted practices were flexible and informal y if you formalise something you are dead. Thus where management is more direct and flexible working rules can be varied to suit the individual this may go some way in diminishing employee demands for more sophisticated practices or union representation (Bacon et al., 1996). This may explain the loyalty as exhibited by low turnover at companies such as ICT Co, Manu Co, Packaging Co, PC Co, and Recycle Co despite the dearth of formal practices in place, thereby serving to undermine any simplistic surface level categorizations of HRM in SMEs on this basis. Employees2 Evidently the type of workers employed and their reaction to practices will also inform and shape the nature and type of HRM adopted. Thus at Manu Co, for example, high absenteeism had resulted in poor productivity which prompted the GM to introduce tight job descriptions and new technology as a mechanism of control. The return on this investment, however, was constrained by a skill shortage in IT and hints of worker dissatisfaction manifested through a ‘go slow’. Employee efforts to operate new machinery therefore seemed to counter the GM’s notion of automatically increasing productivity, a machine operative noted now it’s computerized and all y ,it might make things better if we really bothered. Employee discretionary behavior in relation to ‘managing’ working hours to their benefit was also
An Emergent Theory of HRM
143
evident at Clinic Co (Nurses), Recycle Co (Truck Drivers), and at Packaging Co (Machine Operatives). Similarly at Hotel Co and Sports Co, resistance to work intensification may have been in the form of covert action manifest in the relatively high turnover (Gunnigle & Brady, 1984). Furthermore in many cases there was evidence that employees may be making a conscious trade off accepting the limitations of their current employment as a trade off in order to ‘stay local’ (Nurse at Clinic Co), because there was a lack of alternatives (Pilot at Airline), or because of the informal atmosphere (Clinic Co, Packaging Co) (see McMahon & Murphy, 1999). These issues point to the danger of treating labor simply as a residue in studies of HRM in SMEs (see, for example, Goss’s typology of managerial control in SMEs). Interestingly, the informal relations evident at many of the case study companies did not necessarily imply that information was shared freely with all employees. Application of HRM and practices can therefore vary within as well as between firms (Lewin, 2001). At a number of the case studies there was a reluctance to involve employees in decisions fundamental to the strategic thrust of operations, save for the exception of a ‘trusted elite’ (e.g. the Financial Controller at Packaging Co, operational managers at Airline and ‘two long term aids’ at QualTech). There were notable exceptions to this at companies exhibiting more sophisticated employee representative structures (e.g. Airline, Hotel Co, Law Inc). How these voice mechanisms operated in practice however was questionable. Overall therefore characteristics of SMEs are helpful in facilitating understanding of HRM although not in a deterministic manner. This draws attention to the importance of a holistic perspective for examining HRM in SMEs and the utility of the emergent open systems theoretical proposition in this respect.
DISCUSSION AND CONCLUSION Previous research and literature on HRM has been extremely skewed to the periphery in terms of firm size. Clearly the role of SMEs on the world economic stage stands in stark contrast to our limited knowledge of HRM within them (Arthur & Hendry, 1992, p. 246). The current research has strived to address this imbalance while at the same time facilitate a move away from the crude stereotypes which assume that HRM in SMEs takes a particular form be it either ‘small is beautiful’ or ‘bleak house’. Utilizing a conceptual framework with an emergent open systems theoretical proposition we have attempted to accommodate the issue of small firm heterogeneity by capturing the determinants of HRM in an SME context.
144
BRIAN HARNEY AND TONY DUNDON
Evidence from across the case study companies indicates a remarkable diversity in terms of the nature and form of HRM adopted. More critically the findings suggest that while size constraints and resource limitations clearly had a role in delineating feasible parameters in terms of HRM decisions, size per se did not determine HRM. Rather, HRM was determined through a complex interplay of structural forces and power relations both outside and inside the immediate work milieu. The sophistication and relative formality of HRM at Airline, Ent Co, Law Inc, Hotel Co, and Property Co, for example, can only be adequately appreciated by reference to their competitive environments, the nature of their service offering, as well as managerial structure and power relations. Furthermore each company utilized a specific approach to ‘deal’ with unions, which only made sense with respect to their organizational context and in comparison to how these efforts differed to the approaches of other companies. Similarly the nature of informality evident at Clinic Co, Packaging Co, PC Co, ICT Co, and Recycle Co was not indicative of the substance of HRM but rather informality was dynamic, reflecting unique responses to the structural forces that shaped the parameters of feasible HRM options. This dialectical relationship between agency and broader structural factors exposes the limitation of considering SMEs in isolation from their totality (Barrett & Rainnie, 2002). Reflecting the multiplicity of intervening variables and their vulnerability to environmental forces the approach to HRM by the case study companies was at times piecemeal and reactive. This is best summed up in the terminology utilized by management of dealing with HRM issues ‘as they emerge or warrant attention’ (Manu Co), ‘plodding along’ (Packaging Co), and ‘fire fighting’ (Recycle Co). This uneven approach was also evident at those firms exhibiting more structured and formal strategic planning. At Property Co recruitment was noted as being ‘haphazard’ while at Sport Center training was ‘fragmented and uneven’. Thus the notion of a normative HRM model was not coherent in terms of actual practices, as is often portrayed in best practice and best fit literatures (Cardon & Stevens, 2004; Duberley & Walley, 1995). Linked to this HRM at the case study companies was heavily shaped by the institutional and political contexts in which the firms operated. Reactive and emergent processes were imposed as a consequence of relations with larger capital (e.g. Medical Co, Packaging Co) legislative requirements or to reinforce owner–manager legitimacy and control (Manu Co). Similarly the findings point to the importance of ‘non-strategic determinants’ of HRM such as ISO accreditation (Airline, Property Co, Manu Co, QualTech, Temp Co, and Packaging Co), quality employer awards
An Emergent Theory of HRM
145
(as at Hotel Co), and external agency influences/accreditations (e.g. Irish Nurses Organization at Clinic Co). As noted by the HR Manager of Recycling Co business is dictating to the company what policies are needed. This highlights the limitations of over exaggerated conceptions of strategic choice and the points to potential of institutional and resource dependency approaches in understanding the type and form of HRM adopted. Such factors also suggest that more attention should be given to issues of survivability and ‘table stake HRM’ practices required to compete rather than the literature simply being consumed by the rhetoric of sustainable competitive advantage (Boxall & Purcell, 2003). The emergent open systems theoretical framework therefore draws attention to the external structural factors that shape the parameters of HRM decisions. However, as the case studies illustrate, the actual form HRM takes will reflect a complex interaction of internal dynamics and external determinants. Specifically at Manu Co the introduction of a new GM coupled with intense competitive pressure had served to alter the character of informality and create a more formally regulated workplace as evidenced by the introduction of tight job descriptions and new technology. At Airline Co and Property Co the impact of completive pressures involved work intensification through shift work and extended rotas. In contrast, despite similar competitive pressures, at other companies efforts were made to keep the informal nature of relations intact (Clinic Co, ICT Co, Nursing Co, Packaging Co, PC Co, and Recycle Co). The system of ‘banked hours’ at Packaging Co and part-time and flexible attendance at Clinic Co, for example, helped leverage loyalty and motivation from employees while at the same time served managerial prerogative meeting fluctuating customer demands. This approach of generally investing relatively little in HRM reflected the autonomy provided by a reliance on loose and local labor markets. At companies such as Airline Co, Ent Co, Hotel Co, Law Inc, Medical Co, Property Co, and Sports Center, however, a more sophisticated approach to HRM was evident reflecting the imperative of quality control for customer (client) satisfaction. Sophistication and formalization of HRM, however, did not necessarily imply good working conditions as there was evidence of work intensification and stress among employees. The complexity of HRM was further advanced through varying placement along industrial trajectories as less established companies, such as Recycle Co and ICT Co, faced intensive competitive pressures in their relatively new markets which meant that capital and infrastructural investments took precedence over HRM. This mutual inclusiveness in terms of determining factors lends weight to the emergent, open systems theoretical perspective.
146
BRIAN HARNEY AND TONY DUNDON
Overall the findings highlight the validity of a holistic approach in accommodating the contextual determinants of HRM in SMEs. Moreover, the evidence animates the criticisms directed at best practice and best-fit theories of HRM. In pursuing a largely rational agenda, emphasizing formal practices, with a narrow appreciation of environmental forces such approaches do not adequately accommodate the complex nature of HRM in SMEs. Further, dominant HRM theories impose a biased preconceived ideal of HRM in a context where a range of ‘intelligent’ methods, a lot of the time grounded in informality, may suffice (Marchington et al., 2003). Consequently, the universal HRM project is somewhat misplaced as the homogeneity so desperately sought does not exist (Purcell, 1993; Rainnie, 1989, p. 52). Evidently SMEs operate in a contextually plural world and there is a necessity to develop holistic approaches which accommodate this heterogeneity. A recent review of HRM in SMEs argued that ‘‘not only are more studies needed but there is also a need for developing theory that better applies to SMEs’’ (Tansky & Heneman, 2003, p. 299). This research was an exploratory step along this path indicating that critical to these efforts is an adequate appreciation of the dialectical relationship between structure and agency in shaping HRM. Research into SMEs should therefore be forthcoming, not merely because of the economic importance of SMEs, but critically because without detailed consideration of SMEs theoretical models of HRM can only ever remain partial.
NOTES 1. This list is only indicative and by no means exhaustive. 2. As noted in the methodology section access issues meant that we were only able to interview employees in 10 of the 18 firms. Nonetheless a number of issues can be raised which point to the importance of gaining employee perspectives, the dearth of which is often lamented in HRM research (see Guest, 1999).
ACKNOWLEDGMENTS The authors wish to acknowledge research funding provided by Enterprise Ireland, the Irish Higher Education Authority (HEA), and the Centre for Innovation and Structural Change, NUI Galway. The authors would also like to thank David Lewin and William Brown for comments on an earlier draft of this paper.
An Emergent Theory of HRM
147
REFERENCES Applebaum, E., Bailey, T., Berg, P., & Kalleberg, A. (2000). Manufacturing advantage: Why high-performance work systems pay off. Washington, DC: Economic Policy Institute. Arthur, J. (1992). The link between business strategy and industrial relations systems in American steel minimills. Industrial and Labor Relations Review, 45(3), 488–505. Arthur, J. (1994). Effects of human resource systems on manufacturing performance and turnover. Academy of Management Journal, 37(3), 670–687. Arthur, M., & Hendry, C. (1992). HRM and the emergent strategy of small to medium sized business units. International Journal of Human Resource Management, 3(3), 233–250. Bacon, N., Ackers, P., Storey, J., & Coates, D. (1996). It’s a small world: Managing human resources in small business. International Journal of Human Resource Management, 7(1), 82–101. Baird, L., & Meshoulam, I. (1988). Managing two fits of strategic human resource management. Academy of Management Review, 13(1), 116–128. Baron, J., & Hannon, M. (2002). Organizational blueprints for success in high-technology startups: Lessons from the Stanford project on emerging companies. Californian Management Review, 44(3), 8–37. Baron, R. (2003). Human resource management and entrepreneurship: Some reciprocal benefits of closer links. Human Resource Management Review, 13(2), 253–256. Barrett, R., & Rainnie, A. (2002). What’s so special about small firms? Developing an integrated approach to analysing small firm industrial relations. Work, Employment and Society, 16(3), 415–431. Becker, B., & Gerhart, B. (1996). The impact of human resource management on organisational performance: Progress and prospects. Academy of Management Journal, 39(4), 779–801. Boxall, P., & Purcell, J. (2000). Strategic human resource management: Where have we come to and where should we be going? International Journal of Management Reviews, 2(2), 183–203. Boxall, P., & Purcell, J. (2003). Strategy and human resource management. London: Palgrave. Brand, M., & Bax, E. (2002). Strategic HRM for SMEs: Implications for firms and policy. Education and Training, 44(8/9), 451–463. Brewster, C. (1999). Strategic human resource management: The value of different paradigms. In: R. Schuler & S. Jackson (Eds), Strategic human resource management. Blackwell Business. Cardon, M. (2003). Contingent labor as an enabler of entrepreneurial growth. Human Resource Management Journal, 42(4), 357–373. Cardon, M., & Stevens, C. (2004). Managing human resources in small organizations: What do we know? Human Resource Management Review, 14(3), 295–323. Carroll, M., Marchington, M., Earnshaw, J., & Taylor, S. (1999). Recruitment in small firms: Process, methods and problems. Employee Relations, 23(9), 236–250. Cassell, C., Nadin, S., Gray, M., & Clegg, C. (2002). Exploring human resource management practices in small and medium sized enterprises. Personnel Review, 31(5/6), 671–693. Chamberlain, N. (1948). The union challenge to managerial control. New York: Harper. Child, J. (1972). Organisational structure, environment and performance: The role of strategic choice. Sociology, 6(1), 1–22. Churchill, N. C., & Lewis, V. L. (1983). The five stages of small business growth. Harvard Business Review, 61(3), 30–40.
148
BRIAN HARNEY AND TONY DUNDON
Cully, M., Woodland, S., O’Reilly, A., & Dix, G. (1999). Britain at work as depicted in the 1998 workplace employee relations survey. London: Routledge. Curran, J., & Stanworth, J. (1979). Worker involvement and social relations in the small firm. Sociological Review, 27(2), 317–342. Curran, J., & Stanworth, J. (1981). Size of workplace and attitudes to industrial relations in the printing and electronics industries. British Journal of Industrial Relations, 19(1), 14–25. D’Amboise, G., & Muldowney, M. (1988). Management theory for small business: Attempts and requirements. Academy of Management Review, 13(2), 226–240. Deakins, D., & Freel, M. (1998). Entrepreneurial learning and growth processes in SMEs. The Learning Organisation, 5(3), 144–155. De Kok, J., & Uhlaner, L. (2001). Organizational context and human resource management in the small firm. Small Business Economics, 17(4), 273–291. Delaney, J., & Godard, J. (2001). An industrial relations perspective on the high performance paradigm. Human Resource Management Review, 11, 395–429. Delbridge, R., & Whifield, K. (2001). Employee perceptions of job influence and organizational participation. Industrial Relations, 40, 472–489. Di Maggio, P., & Powell, W. (1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2), 147–160. Duberley, J., & Walley, P. (1995). Assessing the adoption of HRM by small and medium-sized manufacturing organizations. International Journal of Human Resource Management, 6(4), 891–909. Dundon, T., Grugulis, I., & Wilkinson, A. (1999). Looking out of the black hole non-union relations in an SME. Employee Relations, 21(3), 251–266. Dundon, T., Grugulis, I., & Wilkinson, A. (2001). New Management Techniques in small and medium enterprises. In: T. Redman, & A. Wilkinson (Eds), Contemporary human resource management: Text and cases. London: Prentice–Hall, Chapter 14. Dunlop, J. (1958). Industrial relations systems. London: Feffer and Simons Inc. DTI. (2001). Small to medium sized enterprise (SME) – Definitions, London: Department of Trade and Industry, Small Business Service www.sbs.gov.uk/statistics/smedefs.asp Edwards, P., Gilman, M., Ram, M., & Arrowsmith, J. (2003). Public policy, the performance of firms and the ‘missing middle’: The case of employment regulation and role for local business networks. Policy Studies, 23(1), 5–20. Entrialgo, M. (2002). Impact of the alignment of strategy and managerial characteristics on Spanish SMEs. Journal of Small Business Management, 40(3), 260–270. Ferris, G., Arthur, M., Berkson, H., Kaplan, D., Cook-Harrell, G., & Frink, D. (1998). Towards a social context theory of human resource management-organizational effectiveness relationship. Human Resource Management Review, 8(3), 235–264. Forfas. (1999). http://www.forfas.ie/ncc/reports/ncc99/smes.htm Fuller, T., & Moran, P. (2001). Small enterprises as complex adaptive systems: A methodological question? Entrepreneurship and Regional Development, 13(1), 47–63. Gall, G. (2004). British employer resistance to trade union recognition. Human Resource Management Journal, 14(2), 36–53. Godard, J. (2004). A critical assessment of the high-performance paradigm. British Journal of Industrial Relations, 42(2), 349–378. Gooderham, P. N., Nordhaug, O., & Ringdal, K. (1999). Institutional and rational determinants of organizational practices: HRM in European firms. Administrative Science Quarterly, 44, 507–531.
An Emergent Theory of HRM
149
Goss, D. (1991). Small business and society. London: Routledge. Greiner, L. E. (1998). Evolution and revolution as organisations grow. Harvard Business Review, HBR Classic, May/June, 76(3), 55–68. Guest, D. (1999). Human resource management: The workers verdict. Human Resource Management Journal, 9(3), 5–25. Guest, D. (2001). Human resource management: When research confronts theory. International Journal of Human Resource Management, 12(7), 1092–1106. Guest, D., & Hoque, K. (1994). The good, the bad and the ugly: Employment relations in new non-union workplaces. Human Resource Management Journal, 5(1), 1–14. Gunnigle, P., & Brady, T. (1984). Industrial relations in small firms. Employee Relations, 6(5), 21–25. Gunnigle, P., & Moore, S. (1992). Linking business strategy and human resource management: Issues and implications. Personnel Review, 23(1), 63–84. Hendry, C., Arthur, M., & Jones, A. (1995). Strategy through people: Adoption and learning in SMEs. London: Routledge. Heneman, R., Tansky, J., & Camp, M. (2000). Human resource management practices in small and medium sized enterprises: Unanswered questions and future research perspectives. Entrepreneurship: Theory and Practice, 25(1), 11–26. Hill, J., & Wright, L. (2001). A qualitative research agenda for small to medium-sized enterprises. Marketing Intelligence & Planning, 19(6), 433–443. Hornsby, J., & Kuratko, D. (1990). HRM in small business: Critical issues for the 1990s. Journal of Small Business Management, July, 28(3), 9–18. Hortsman, B. (1999). Decentralised and de-regulated Australian industrial relations. Employee Relations, 21(3), 325–341. Huselid, M. (1995). The impact of human resource management practices on turnover, productivity, and corporate financial performance. Academy of Management Journal, 38(3), 635–672. Huselid, M., & Rau, B. (1997). The determinants of high performance work systems: Crosssectional and longitudinal analysis. Academy of Management Annual Meetings, Human Resource Management Division. Ichniowski, C., Kochan, T. A., Levine, D., Olson, C., & Strauss, G. (1996). What works at work: Overview and assessment. Industrial Relations, 35(3), 299–333. Ichniowski, C., Shaw, K., & Prennushi, G. (1997). The effects of human resource management practices on productivity: A study of steel finishing lines. American Economic Review, June, 86(3), 291–313. Imrie, R. F. (1986). Work decentralisation from large to small firms: A preliminary analysis of subcontracting. Environment and Planning, 18, 949–965. International Labor Organization (ILO). (2005). Small scale activities and the productivity divide, In ILO. 2005. World Employment Report 2004–2005, ILO, Chapter 5. Jackson, S., & Schuler, R. (1995). Understanding human resource management in the context of organizations and their environment. Annual Review of Psychology, 46, 237–264. Jaffe, D. (2001). Organization theory, tension and change. Boston: McGraw Hill. Kaman, V., McCarthy, A., Gulbro, R., & Tucker, M. L. (2001). Bureaucratic and high commitment human resources practices in small service firms. Human Resource Planning, 24(1), 33–55. Katz, J., Aldrich, H., Welbourne, T., & Williams, P. (Eds.) (2000). Towards a new synthesis: Editors comments. Entrepreneurship: Theory and Practice, 25(1), 1–5
150
BRIAN HARNEY AND TONY DUNDON
Kerr, A., & McDougall, M. (1999). The small business of developing people. International Small Business Journal, 17(2), 25–37. Kinnie, N., Purcell, J., Hutchinson, S., Terry, M., Collinson, M., & Scarborough, H. (1999). Employment relations in SMEs: Market driven or customer shaped. Employee Relations, 21(3), 218–235. Klass, B., McClendon, J., & Gainey, T. (2002). Trust and the role of professional employer organizations: Managing HR in small and medium enterprises. Journal of Managerial Issues, 14(1), 31–49. Kochan, T., Katz, H. C., & McKersie, R. B. (1986). The transformation of American industrial relations. New York: Basic Books. Lawler, E. (1992). The ultimate advantage: Creating the high-involvement organization. San Francisco: Jossey-Bass. Lee, S. (1997). HRM and the legitimacy market. International Journal of Human Resource Management, 8(2), 226–243. Lee, M., & Johnson, N. (1998). Business environment, high-involvement management, and firm performance in Korea. In: L. David & K. Bruce (Eds), Advances in Industrial and Labor Relations, 8 (pp. 67–87). Greenwich, CT: JAI Press. Legge, K. (2001). HRM silver bullet or spent round. In: J. Storey (Ed.), HRM: A critical text. London: Thompson Learning. Lewin, D. (2001). IR and HR perspectives on workplace conflict: What can each learn from the other? Human Resource Management Review, 11, 453–485. Lewin, D., Feuille, P., Kochan, T. A., & Delaney, J. (1988). Public sector labor relations: Analysis and readings (3rd ed.). Lexington, MA: Lexington Books. Little, B. L. (1986). Performance of personnel duties in small Louisiana firms: A research note. Journal of Small Business Management, 22(4), 66–70. Lynch, J., & Roche, F. (1995). Business Management in Ireland. Dublin: Oak Tree Press. Mabey, C., & Salaman, G. (1995). Strategic human resource management. London: Blackwell Press. Mabey, C., Salaman, G., & Storey, J. (1998). HRM: A strategic introduction. Blackwell Publishing. MacDuffie, J. P. (1995). HR bundles and manufacturing performance: Organisational logic and flexible production systems in the world auto industry. Industrial and Labor Relations Review, 48(2), 197–221. Marchington, M., Carrol, M., & Boxall, P. (2003). Labour scarcity and the survival of small firms: a resource-based view of the road haulage industry. Human Resource Management Journal, 13(4), 5–22. Marchington, M., & Grugulis, I. (2000). Best Practice HRM: Perfect opportunity or dangerous illusion? International Journal of Human Resource Management, 11(4), 905–925. Marchington, M., & Parker, P. (1990). Changing patterns of employee relations. HarvesterWheatsheaf, Brighton. Matlay, H. (1999). Employee relations in the small firm. Employee Relations, 21(3), 285–295. Matlay, H. (2002). Industrial relations in the SME sector of the British economy: An empirical perspective. Journal of Small Business and Enterprise Development, 9(3), 307–318. McMahon, J. (1995). Employee relations in small manufacturing firms. In: T. Turner & M. Morley (Eds), Industrial relations and the new order: Case studies in conflict and co-operation. Dublin: Oak Tree Press, Chapter 5. McMahon, J. (1996). Employee relations in small firms in Ireland: An exploratory study of small manufacturing firms. Employee Relations, 18(5), 66–80.
An Emergent Theory of HRM
151
McMahon, J., & Murphy, E. (1999). Managerial effectiveness in the small enterprise: Implications for HRD. Journal of European Industrial Training, 23(1), 25–35. Mintzberg, H. (1994). The rise and fall of strategic planning. FT Prentice-Hall. Mintzberg, H., & Waters, J. (1985). Of strategies deliberate and emergent. Strategic Management Journal, 6(3), 257–272. Murray, A. I. (1988). A contingency view of Porter’s generic strategies. Academy of Management Review, 13(3), 390–401. Nadler, D. A., & Tushman, M. A. (1997). A congruence model for organization problem solving. In: M. Tushman & P. Anderson (Eds), Managing Strategic Innovation and Change. New York: Oxford University Press. Observatory of European SMEs (OES). (2002). SMEs in Europe, Report No. 2 (Available Online). Organisation for Economic Co-operation and Development (OECD). (2005). OECD SME and entrepreneurship outlook. Paris: OECD. Osterman, P. (1994). How common is workplace transformation and who adopts it? Industrial and Labor Relations Review, 47(2), 173–189. Paauwe, J. (2004). HRM and performance: Achieving long-term viability. Oxford: University Press. Paauwe, J., & Boselie, P. (2003). Challenging ‘strategic HRM’ and the relevance of the institutional setting. Human Resource Management Journal, 13(3), 56–70. Paauwe, J., & Boselie, P. (2005). HRM and performance: What next? Human Resource Management Journal, 15(4), 68–83. Pettigrew, A. M. (1985). Awakening giant. Oxford: Blackwell. Pfeffer, J. (1998). The human equation: Building profits by putting people first. Boston: Harvard Business School Press. Pil, F., & MacDuffie, J. P. (1996). The adoption of high-involvement work practices. Industrial Relations, 35(3), 423–455. Purcell, J. (1993). Developing research in comparative HRM. International Journal of Human Resource Management, 4(3), 507–510. Purcell, J. (1999). Best practice or best fit: Chimera or cul-de-sac? Human Resource Management Journal, 9(3), 26–41. Rainnie, A. (1989). Industrial relations in small firms. London: Routledge. Ram, M. (1991). Control and autonomy in small firms: The case of the West Midlands Clothing Industry. Work, Employment and Society, 5(4), 601–619. Ram, M., Edwards, P., & Gilman, M. (2001). The dynamics of informality: Employment relations in small firms and the effects of regulatory change. Work, Employment and Society, 15(4), 845–861. Ramsay, H., Scholarios, D., & Harley, B. (2000). Employees and high-performance work systems: Testing inside the black box. British Journal of Industrial Relations, 38(4), 501–531. Redman, T., & Wilkinson, A. (Eds) (2001). Contemporary Human Resource Management: Text and Cases. London: Prentice-Hall. Reid, R., & Adams, J. (2001). Human resource management – a survey of practices within family and non-family firms. Journal of European Industrial Training, 25(6), 310–321. Robinson, R. B., & Pearce, J. A. (1984). Research thrust in small firm strategic planning. Academy of Management Review, 9(1), 128–137. Roy, D. (1980). Repression and incorporation fear stuff, sweet stuff and evil stuff, management’s defensiveness against unionism in the south, In: T. Nicholas (Ed.), Capital and labor: A marxist primer fontana. Glasgow.
152
BRIAN HARNEY AND TONY DUNDON
Rubery, J., Earnshaw, J., Marchington, M. C. F. L., & Vincent, S. (2002). Changing organizational forms and the employment relationship. Journal of Management Studies, 39(5), 645–672. Rutherford, M. W., Buller, P. F., & McMullen, P. R. (2003). Human resource problems over the life cycle of small to medium-sized firms. Human Resource Management, 42(4), 321–336. Scase, R. (1996). Employment relations in small firms. In: P. Edwards (Ed.), Industrial relations theory and practice. Oxford: Blackwell. Schuler, R., & Jackson, S. (1987). Linking competitive strategies and human resource management practices. Academy of Management Executive, 1(3), 207–219. Scott, R. (2001). Organizations: Rational, natural, and open systems (4th ed.). Englewood Cliffs: Prentice-Hall. Sels, L., De Winne, S., Maes, J., Delmotte, J., Faems, D., & Forrier, A. (2006). Unravelling the HRM-performance link: Value-creating and cost-increasing effects of small business HRM. Journal of Management Studies, 43(2), 319–342. Sisson, K. (1993). In search of human resource management. British Journal of Industrial Relations, 31(2), 201–210. Small Firms Association (SFA). (2003). SFA employment law survey press release. Dublin: SFA. Sparrow, P. (1999). Is HRM in crisis, In: Schuler, R. & Jackson, S. (Eds), Strategic human resource management, Blackwell Business, Chapter 21. Stace, D., & Dunphy, D. (1991). Beyond traditional paternalistic and developmental approaches to organisational change and human resource strategies. The International Journal of Human Resource Management, 2(December), 263–284. Storey, D. (1994). Understanding the small business sector. London: Routledge. Tansky, J. W., & Heneman, R. (Eds) (2003). Guest editor’s note: Introduction to the special issue on human resource management in SMEs: A call for more research. Human Resource Management, 42(4), 299–303. Truss, C. (2001). Complexities and controversies in linking HRM with organisational outcomes. Journal of Management Studies, 38(8), 1121–1149. Truss, C. (2002). Shifting the paradigm in human resource management: From the resource based view to complex adaptive systems. Research Papers in Human Resource Management, Kingston Business School (http://business.kingston.ac.uk/research/hrm/hrmpaps.html). Tyson, S. (1997). Human resource strategy: A process for managing the contribution of HRM to organisational performance. International Journal of Human Resource Management, 8(3), 277–291. Wagar, T. (1998). Determinants of HRM in small firms: Some evidence from Atlantic Canada. Journal of Small Business Management, 36(2), 61–67. Wall, T., & Wood, S. (2005). The romance of human resource management and business performance, and the case for big science. Human Relations, 58(4), 429–462. Welsh, J., & White, J. (1981). A small business is not a little big business. Harvard Business Review, 59(4), 18–32. Whittington, R. (2001). What is Strategy – and does it matter? Thomson Learning. Wilkinson, A. (1999). Employment relations in SMES. Employee Relations, 21(3), 206–217. Williamson, I., & Cable, D. (2003). Organisation hiring patterns, inter-firm network ties, and interorganisational imitation. Academy of Management Journal, 46(3), 349–359. Wood, S. (1999). Human resource management and performance. International Journal of Management Reviews, 1(4), 367–414.
An Emergent Theory of HRM
153
Wright, P., & McMahon, G. (1992). Theoretical perspectives for strategic human resource management. Journal of Management, 18(2), 295–320. Wright, P. M., & Snell, S. A. (1991). Toward an integrative view of strategic human resource Management. Human Resource Management Review, 1(3), 203–225. Youndt, M., Dean, J., & Lepak, D. (1996). Human resource management, manufacturing strategy and firm performance. Academy of Management Journal, 39(4), 836–866. Zedeck, S., & Cascio, W. (1984). Psychological issues in personnel decisions. Annual Review of Psychology, 35, 461–518.
A NEW FRAMEWORK OF ENTERPRISE UNIONISM: A COMPARATIVE STUDY OF NINE ASIAN COUNTRIES Dae Yong Jeong and John Lawler ABSTRACT This paper proposes a new theoretical framework to explain enterprise unionism and conducts the first systematic comparative study of union structure in nine Asian countries. Our framework emphasizes political dynamics and the role of the state in labor relations and argues that the initial period of the collective bargaining era constituted a critical juncture (state labor policy) that occurred in distinctive ways in different countries and that these differences played a central role in shaping the different union structures in the following decades. The nine countries are mainly divided into three groups, depending on the type of state labor policy: enforcement of enterprise unionism; centralization/laissez-faire (non-enterprise unionism); and dual unionism/gradual transition (middle-ground). Governmental data were used for the study. A clear correspondence between state labor policy and union structure in each of these groups was found. We believe that our framework significantly enhances our understanding of the Asian cases. Future research should explore the validity of the proposed framework through comparative studies of Latin American cases where enterprise unions have also been observed. Advances in Industrial and Labor Relations, Volume 15, 155–211 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15004-0
155
156
DAE YONG JEONG AND JOHN LAWLER
Unlike forecasts of postwar industrial relations (IR) scholars, neither industrial (Kerr et al., 1960) nor general unionism (Galenson, 1959) has flourished in most Asian countries. What they largely missed was the development of enterprise unionism in this region. Why do workers in some countries organize themselves by enterprise contrary to their counterparts in other part of the world? This is the main research question of this paper. Great effort has indeed been devoted to explain enterprise unionism in Japan over a half century since the term enterprise union was dubbed (Social Science Research Institute at the University of Tokyo, 1950). Nevertheless, the leading frameworks (i.e., the internal labor market (ILM) hypothesis and the cultural hypothesis) rest on weak foundations and have been unsuccessful in providing a clear understanding of this phenomenon. The most crucial and common errors in these analyses have been the erroneous assumption that the established union structure is the manifestation of the nature of Japanese workers and the exclusion of the role of the state, which is a critical actor in labor relations (Dunlop, 1958), from these frameworks. More recent studies of Asian IR have largely centered on explanations of relative economic success and weak labor (and the possible relationship between them), leaving the structural characteristics of the unions in this region completely untouched. Frenkel (1993) has been the only crossnational study that pays significant attention to the structural characteristics of trade unions in this region. He found enterprise unions in nine of the 10 countries in the Asia-Pacific region and argued for the importance of the state in shaping patterns of unionism, among others. But, his study does not provide any theoretical mechanism for the emergence of enterprise unionism nor is it able to explain substantial national variations in the diffusion of enterprise unionism in Asia. In this paper, we propose a new theoretical framework to explain enterprise unionism and conduct the first systematic comparative study of union structure in Asian countries. The framework applies the concept of ‘‘critical junctures’’ – a period of significant change that typically occurs in distinct ways in different countries (or in other units of analysis) and which is hypothesized to produce distinct legacies (Lipset & Rokkan, 1967; Collier & Collier, 1991) – to the evolution of union systems in Asian countries. It maintains that the initial period of the collective bargaining era constituted a critical juncture (state labor policy) that occurred in distinctive ways in different countries and that these differences played a central role in shaping the different union structures in the following decades. This paper consists of two parts. In the first part, we first introduce the research methods we used for this study and then review existing
Enterprise Unionism in Asia
157
works on enterprise unionism and identify key weaknesses of these perspectives. We then propose a new framework, which emphasizes political dynamics and the role of the state in labor relations. In this section, we first describe strategic behavioral patterns of the tripartite actors (employers, unionists, and the state) under collective bargaining and establish the ‘‘model’’ in which the state seeks to decentralize the union movement, shifting the locus of power toward management in determining the union structure. We then suggest factors that significantly alter a state’s capacity and legitimacy for its labor policy and point out the unique conditions in Asia. In the second part, we test how well the proposed framework explains the various union systems across the nine Asian countries included in this study. The nine countries are mainly divided into three groups, depending on the type of state labor policy: (1) enforcement of enterprise unionism; (2) centralization/laissez-faire (non-enterprise unionism); and (3) dual unionism/ gradual transition (middle-ground). As such, this part is divided into three sections. In the first section, we investigate the countries in which the state has implemented an enterprise union system. Depending on the consistency of the state labor policy, the countries in this model are further divided into two subgroups. Thus, we first discuss one subgroup of these countries, where the state has consistently implemented an enterprise union system (Japan, Thailand, and Malaysia) and then discuss the other subgroup of countries, where the state has implemented an enterprise union system, but with some interruptions (Korea and the Philippines). In the second section, we delve into the countries in the non-enterprise unionism model. The countries in this model are Hong Kong, where the state labor policy has been laissez-faire; and Indonesia, where the state had implemented a centralization policy at least until 1994. In the third section, we examine the countries in the middle-ground of this spectrum. These are Taiwan, where the state has pursued a dual union system of craft and enterprise unions; and Singapore, where the state policy has gradually swung from laissez-faire toward an enterprise union system since 1984.
METHODOLOGY Methods The paper is largely based on comparative-historical methods mainly for the following reasons. First, the development of enterprise unionism is clearly a
158
DAE YONG JEONG AND JOHN LAWLER
historical phenomenon, which requires historical analysis. Second, statistical methods are unable to deal with a small number of relevant cases in this study. Through the comparative method, we distinguish between systemspecific regularities and universals, figure out categorical rules within those categories, and reexamine concepts and methodologies that are mandated by the discovery of exceptions. The major comparative method employed in this project is Mill’s (1843) method of difference. This inductive method works fine in verifying an argument by identifying a constant conjunction as evidence of causal relations, but it works poorly in falsifying an argument based on a failure in a constant conjunction. Our solution to this problem is to conduct a deductive project starting from a set of hardcore theoretical propositions (framework) and then apply the method of difference in a conservative way. We also compensate for certain drawbacks in the method of difference by adopting additional methods: Tilly’s (1984) variation-finding comparison and Przeworski and Teune’s (1970) most different systems design. The variation-finding comparison establishes a principle of variation in the character or intensity of a phenomenon having more than one form by examining systematic differences among cases. With this methodological design, this study identifies three broad types of state labor policy and seeks to discover whether corresponding contrasts emerge in the legacy of these policies. The logic of the most different systems design is that if the observed relationship holds in a range of contrasting settings, the argument of the research is better supported. This study thus rearranges the nine countries in the three different groups by the level of economic developments and investigates whether the corresponding smaller groups exhibit the same pattern of the observed relationship between state labor policy and union structure. These methods together enable us to discover whether corresponding contrasts emerge in the legacy of these policies, giving additional precision to a solution that the method of difference has found and enhancing the generalizability of the observed relationship across countries in different levels of economic developments. In this paper, a ‘‘horizontal’’ union includes all types of unions that organize workers across enterprises (i.e., industrial craft, general unions, etc.), as opposed to an enterprise union whose membership is restricted to employees in the particular enterprise, as commonly defined in the Japanese IR literature. However, unlike most researchers of Japanese industrial relations, we do not refer to an enterprise union as a ‘‘vertical’’ union, because an enterprise union does not necessarily organize all the eligible employees regardless of their status.1
Enterprise Unionism in Asia
159
Cases The user of the method of difference must be concerned about comparability of cases. Manipulations as precise as those called for by the method of difference are not possible in the social sciences or in field experiments generally. The only feasible strategy is, then, to design studies that approximate this condition as closely as possible. As such, some have suggested concentrating on some key variables to find ‘‘comparable cases’’ and have carried out their studies along this line (Liphart, 1971; DeFelice, 1980; Cole, 1979; Maurice, Sellier, & Silvestre, 1986; Kume & Thelen, 1999). The most fruitful approach, however, has been regional designs that include a range of countries, similar in as many ways as possible to control for ‘‘concomitant variation.’’ Numerous studies have been done of Anglo-American countries (Tolliday & Zeitlin, 1985; Card & Freeman, 1993), or of Latin American countries (Collier & Collier, 1991; Cohen, 1994), or of the ‘‘little dragons’’ in Asia (Deyo, 1987, 1989; Aoki et al., 1997). Following this tradition, we focus on the Asian region, and compare and contrast the development of the trade union system in nine Asian countries to test the validity of the framework proposed. We do not pretend to compare ‘‘ideal cases’’ that are different in only certain aspects, but we certainly intend to defy a claim that the cases are too different from each other to be compared for the purpose of this project. They all shared proximities in terms of democracy, geography, and economic conditions at the onset of the collective bargaining era. The Asian region is also an excellent candidate for this project in that great variations in union structure can be observed across countries whereas enterprise unions are concentrated in this region. In addition, the region includes distinct country categories: the core advanced late developer, Japan; the newly industrialized countries (Korea, Taiwan, Singapore, and Hong Kong); and the developing countries of Southeast Asia (Thailand, the Philippines, Malaysia, and Indonesia). Inclusion of these countries in different categories indicates how robust the framework is or if the ‘‘generalization holds’’ across countries in different stages and speeds of economic development. Data It required extra effort to collect reliable data on the proportions of different types of unions (i.e., enterprise, industrial, general, and craft) in each of the
160
DAE YONG JEONG AND JOHN LAWLER
countries, which is the dependent variable of the project. Fortunately, the Japanese, Korean, Taiwanese, and Thai governments annually publish such data. But, their counterparts in Singapore, Malaysia, Hong Kong, the Philippines, and Indonesia do not publish such data. But we were able to collect either the relevant data or at least the lists of all the unions from government agencies in these countries. Findings We found a highly dominant, stable enterprise union system in Japan (>90 percent), Thailand (90 percent), and Malaysia (80 percent), where the state has consistently implemented an enterprise union system; and a dominant, but unstable enterprise union system in Korea (>90 percent, but rapidly declining) and the Philippines (71 percent), where the state has implemented an enterprise union policy, with some interruptions. On the other hand, we found enterprise unions to be insignificant in Hong Kong (18 percent), where the state labor policy has been laissez-faire; and no enterprise union until 1994 in Indonesia, where the state had implemented a centralization policy. In the middle-ground of this spectrum are Singapore (43 percent), where the state policy has gradually swung from laissez-faire toward an enterprise union system since 1984; and Taiwan (28 percent), where the state has pursued a dual union system of craft/occupation unions (which are the majority) and enterprise unions. The reasons for the different state policies are explained in detail.
REVIEW OF EXTANT HYPOTHESES OF ENTERPRISE UNIONISM Research on enterprise unionism has relied on two main hypotheses: the cultural hypothesis and the ILM hypothesis. In the following pages, we review each of them in turn and identify key weaknesses of these perspectives. Cultural Hypothesis Culturalists argue that the Japanese industrial relations system (characterized by ‘‘lifetime employment,’’ the seniority wage system, and enterprise unionism) is a natural, even inevitable outcome of a peculiar Japanese mentality embedded in the culture. In this view, among the most important
Enterprise Unionism in Asia
161
assets which modern Japan owes to the Tokugawa period (1600–1867) are basic attitudes such as egalitarian traditions, dedication to hard work, respect for seniority and hierarchy, loyalty to communal discipline, submission to authority and adherence to tradition (e.g., Abegglen, 1958; Nakane, 1967; Iwada, 1977; Kahn, 1979). This cultural argument, however, rests on weak ground. First, it lacks a theoretically sound mechanism between the alleged group orientation and the development of enterprise unions in Japan. Even when we assume that Japanese workers were more group-oriented than their counterparts in Western countries, there still remains the great danger of falling into fallacies. In the post hoc fallacy, one assumes that what came first is necessarily the cause of a later event. In the fallacy of ‘‘obvious concretes,’’ one assumes that the most obvious and directly perceivable difference between two entities explains all other differences (Sakoh, 1990). What is then the causal mechanism between the alleged ‘‘group consciousness’’ and the development of enterprise unions? Why should we take for granted an enterprise as the unit of the ‘‘group?’’ What about other types of groups such as an occupational group, a business unit (and/or a plant) within an enterprise, or a trade union among workers across enterprises? The culturalist has not been successful in suggesting such a conceptual mechanism. We will discuss this issue further in the next section, which proposes behavior patterns of each of the tripartite actors on the topic. Would the culturalist also conceptualize the development of an enterprise union system in other Asian countries in the same way? In other words, would they argue that workers in other Asian countries share such group consciousness with Japanese workers? We are not aware of any writer who has attempted to link such ‘‘Asian values’’ with the structural characteristics of trade unions in these countries. In the first place, is the prevalence of group consciousness among prewar Japanese workers empirically true? Far from culturalists’ imagination of docile and diligent Japanese workers with group consciousness, the early industrial workers were assertive and possessed undisciplined characteristics (Gordon, 1988; Endo, 1985). They were traveling craftsmen seeking to polish their skills and to obtain a higher salary, as were their counterparts in Western countries (Okochi & Levine, 1973). Prewar Japanese workers never abandoned the traditional practices and values. Accordingly, almost all of the prewar Japanese unions were horizontal (i.e., industrial, craft, and general) unions (Taira, 1970; Komatsu, 1971). Furthermore, the cultural hypothesis is poorly equipped to explain variations in the organizing capacities of workers at different moments,
162
DAE YONG JEONG AND JOHN LAWLER
especially running into problems when confronted with historical and contemporary evidence of labor activism in the workplace and the national political arena (Hutchison & Brown, 2001). For instance, one should be reminded of the militant, class-based union movement in the postwar period. As Robinson (1996) points out, culturalists have no social theory to explain such individuals and developments other than in behavioral terms as cultural deviants or as mistaken options that contradict the social and cultural realities of Japan. Had the postwar horizontal union movement not been crushed by the state and management, would a type of horizontal unionism not dominate in Japan today? The state and management have, since the crush of industrial unionism, been forceful in implementing the ‘‘culture’’ while eradicating ‘‘heretics’’ from the workplace (Kumazawa, 1989, 1996; Gordon, 1998; Nomura, 1993; Tona, 1993). To the extent that the state and management have been successful in doing so, one observes little in the ‘‘culture’’ that suggests class-based forces that would sustain horizontal unionism. ‘‘Culture itself is a variable’’ (Dohse et al., 1985, p. 141). Internal Labor Market Hypothesis ILM theorists, in contrast, argue that Japan has, since the 1920s, developed highly internalized labor markets. Under this system, employees spend their entire working lives in an enterprise and are mainly paid for seniority within the enterprise rather than based on job type or occupational abilities. It is thus natural for them to develop enterprise-consciousness, without a sense of solidarity with workers in similar jobs and occupations across enterprises. In this view, the structure of (enterprise) unions is therefore a logical outcome of the labor markets and is what most Japanese workers desire (e.g., Koike, 1977, 1987, 1988; Shirai, 1983; Shimada & Shirai, 1978). This hypothesis, however, does not seem to meet one of the three criteria for causation – the cause precedes the effect. In order for this hypothesis to be valid, ILMs should have preceded enterprise unions, and the alleged enterprise-consciousness should have mediated the causal link between these two variables. Yet, there is little empirical evidence for the causal link between prewar paternalistic management practices (from the 1920s onward) and postwar enterprise unionism (Gordon, 1988). Behind the decline in labor mobility (which was still high) was the general stagnant state of employment rather than the asserted awakening of enterprise-consciousness (Ohkawa & Rosovsky, 1973; Taira, 1970). Rather, a reverse causal link (if any) between the two variables seems more plausible. It is natural for an
Enterprise Unionism in Asia
163
enterprise union to seek employment security within the enterprise. When this is matched with management strategies for labor-management stability, an equilibrium of adversarial force (ILM) can be established, whether or not employees hold enterprise-consciousness. Is this not what happened in postwar Japan? The formation of ILMs in fact came much later than the explosive growth of the plant unions in the early postwar years. The more straightforward example of the Korean experience confirms this reverse causal link, if any (see Lee, 1996, 1998a, 1998b; Jeong, 1995). The ground for this hypothesis is further weakened from a comparative perspective. It cannot explain why other Western countries did not develop enterprise union systems, despite having ILMs in large firms; and why other Asian countries with underdeveloped ILMs have enterprise union systems. Neither can it explain, by the same logic, why workers in medium-sized and small enterprises in Japan also tend to organize themselves primarily into enterprise unions.
A SOCIO-POLITICAL FRAMEWORK OF ENTERPRISE UNIONISM In the following pages, we propose a new framework of enterprise unionism, which emphasizes political dynamics and the role of the state in labor relations. We first describe strategic behavioral patterns of the tripartite actors under collective bargaining and establish the ‘‘model’’ in which the state seeks to decentralize the union movement, shifting the locus of power toward management in determining the union structure. We then suggest factors that significantly alter a state’s capacity and legitimacy for its labor policy and point out the unique conditions in Asia. Collective Bargaining and Union Structure Structural differences between earlier unions were largely due to their varying abilities to impose upon their members the rigorous discipline, without which unilateral regulation could not succeed (e.g., apprenticeship by craft unions). Collective bargaining, however, suits any union structure, and recognition by employers and the state for the purpose of bargaining stabilizes and legitimizes existing union structures (Clegg, 1976). Therefore, the claim of the union to represent its members no longer rests on its own strength alone under collective bargaining. Rather, the underlying position of the union becomes essentially defensive. Collective bargaining thus bring with it
164
DAE YONG JEONG AND JOHN LAWLER
intensified tension, conflicts, and struggles among the tripartite actors to institutionalize a union structure that will maximize its own benefits. This is why the locus of power – politics – among the actors at the onset of the collective bargaining era is critical in determining the dominance of a certain type of union structure in the following decades and why we propose a socio-political framework to explain it. The Unionist Under collective bargaining then, which type of union structure do unionists prefer and attempt to establish? Is there any universal tendency in this regard? In general, unionists aim for a centralized union structure against local exclusiveness (or monopoly). What about enterprise unionism in Japan, which is a type of local exclusive unionism? Does it mean that Japanese unionists are exceptional in this regard? To the contrary, the classic works of Webbs (1911) and Cole (1953) on Western unionism clearly reveal that even in Western countries horizontal unionism was not a built-in phenomenon, but was built out of unionists’ endeavors against the strong inherited tradition of local exclusiveness and natural selfishness of each branch in desiring to preserve its own local monopoly. In England, for example, ‘‘many of the earliest Trade Unions began in effect as works clubs or companionships, formal or informal associations of the workers employed by a particular establishment. Even when Trade Unionism became a general movement, many local trade societies and many branches of larger bodies continued to consist of workers drawn either from a single works or factory, or from a group of neighboring and closely related works’’ (Cole, 1953, p. 5). According to Webbs (1911), the irresistible tendency of expansion is in the union desire to secure uniform minimum conditions throughout each industry as a uniform policy can only be arrived at and maintained by a central body acting for the whole trade (i.e., to ‘‘take wages, health, safety, the environment, etc. out of competition’’ in contemporary terms). Therefore, the constant tendency for a centralized and bureaucratic administration is, in the union world, accepted and even welcomed by sturdy defenders of local autonomy. In addition, the union’s desire to increase bargaining power and political influence is also attributed to expansion. For unionists, success in institutionalizing a horizontal union system thus depends on whether they can overcome internal union politics, such as representational factors (Weber, 1967) and collective action problems (Olson, 1968), which make coercion and sanctions by the employer and the
Enterprise Unionism in Asia
165
state important elements in the creation and viability of labor unions. As Tilly (1978) points out, people act collectively to maximize their collective gains, but they do so within the constraints set by the costs of collective action. The Employer Employers desire to remain as non-union as possible or deal with the most moderate unions since unions necessarily interfere with management’s prerogatives to the extent that they function (Herman, Kuhn, & Seeber, 1987). But once collective bargaining becomes an issue, employers take upon themselves a share of the responsibility for enforcing the rules. Employers tend to disfavor centralization in union structure in principle. The rationale of employer opposition to centralization in union structure is twofold – economic and political. Decisions concerning the structure of bargaining are highly influenced by the structure of the union. As Kochan and Katz (1988) note, centralization in union structure tends to enable the union to achieve centralized bargaining, which in turn increases its ability to impose strike costs on the employers and strengthens its bargaining power particularly against large and financially strong employers.2 Besides this economic reason, centralization in union structure also represents political empowerment of unions. A centralized union can become a powerful interest group, for example, exerting pressure for favorable labor legislation. The cause is more political in nature, however, when union centralization is involved with ideology, as radical and ideology-oriented unions (which must go beyond the boundary of an enterprise to raise a common struggle on a class basis) could even threaten the foundation of management prerogatives. This sharp difference in the desirable union structure between the unionist and the employer makes the state labor policy a key determinant. The State The state is referred to as a set of continuous administrative, legal, bureaucratic, and coercive systems, headed, and more or less well coordinated, by an executive authority (Evans et al., 1985; Woo-Cummings, 1997). The overlap between the aims of state policy and the interests of any of the major economic groups is both partial and contingent, since states rarely pursue economic growth and/or welfare for their own sake, but rather as a means to their own political and military ends (Tilly, 1985; Zeitlin, 1985).
166
DAE YONG JEONG AND JOHN LAWLER
The political importance of the labor movement to the state is unions’ concrete collective power either to bestow political support or to mobilize opposition (Collier & Collier, 1991). Because of this double-edged capacity of a centralized labor movement, the state cannot be assumed to have an inherent preference for either centralization or decentralization in union structure. Rather, the state is supposed to make a strategic choice. If the state seeks centralization in union structure, the emergence of an enterprise union system is highly unlikely because no driving force toward an enterprise union system exists. In such a case, employers would also welcome centralization in union structure because if the centralized unions are tightly controlled by the state, they can have, free of charge, moderate unions to deal with. The Model The state might also decide to decentralize the labor movement. This will be certainly the case, in particular if the state perceives the labor movement to be too aggressive (strike-prone and/or ideology-oriented). Suppose that the state possesses enough resources and legitimacy to eliminate any strategic choice and actions. The state then would attempt to crush the aggressive centralized labor movement. The successful elimination of the aggressive centralized labor movement paves the way for an enterprise union system. To prevent a resurgence of an aggressive centralized labor movement, the state is likely to use overt and/or covert measures, such as legal enforcement of an enterprise union, prohibition of certain types of horizontal unions, or manipulation and surveillance of workers at the workplace. An enterprise union system then emerges and dominates the labor movement. We thus call these measures ‘‘mechanisms of production’’ of an enterprise union system. This legacy may then be reproduced through a continuous, consistent state policy (labor law and/or manipulation and surveillance), management endeavors (HRM, ILM, welfare programs, etc.), and ‘‘enterprise consciousness’’ among workers, which may develop along with the institutionalization of this system. These measures then can be thought of as ‘‘mechanisms of reproduction’’ of an enterprise union system. The completeness and stability of the system depends on the effectiveness and consistency of implementing these mechanisms. These mechanisms of producing and reproducing an enterprise system are illustrated in Fig. 1 below.3 When an aggressive centralized labor movement is absent, both centralization and decentralization policies are equally options for the state. As in
Enterprise Unionism in Asia
167
Legacy
Critical Juncture • Collective bargaining • State Labor Policy Autonomy? • Labor vs. Capital Resources? • Labor vs. State Legitimacy? Cleavages
Antecedent Conditions
Mechanisms of Production • Crush • Labor Law • Manipulation/ surveillance
Mechanisms of Reproduction • Labor Law • Manipulation/surveillance • Welfare Programs • HRM • ILM • Enterprise Consciousness
Stability of the Legacy
Mutual-aid Societies Socioeconomic Change • Commercial/Industrial Growth • New Production Method • Growth of Working Class • Left-wing Ideology
Fig. 1.
A General Model of Enterprise Unionism in a Critical Junctures Framework.
the former case, to the extent to which the state attempts to institutionalize an enterprise union system, and management is able to deal with shop-floor matters, the enterprise union system can dominate the IR system.4 Environment and Timing In reality, whether or not a state would attempt a decentralization policy depends on its capacity and legitimacy for the policy as it is directly against unionists’ will. We suggest two factors that significantly alter a state’s capacity and legitimacy, among others – the external environment in which the state is immersed and the timing of the policy. Geopolitics has a salience for internal politics of the state. As historians have observed, military and political competition between states is both one of the most conspicuous features of the modern world, and historically one of the most important means by which state managers have enhanced their independence from civil society (Tilly, 1985; Skocpol, 1985). Such an environment also enables the state to mobilize greater resources and to enhance legitimacy for its policies (if necessary) against the interests of any major economic groups. At the same time, the state can also gain support
168
DAE YONG JEONG AND JOHN LAWLER
and aids from allied powers, which may in turn help ease possible pressure from international (labor and human rights) organizations. The state could play a vital role in shaping its economic and industrial relations systems, but this is particularly so at the early stage of development (Clegg, 1976; Black, 2001). When the economy is underdeveloped, the availability of intermediaries is limited, the capabilities of firms are modest, and even the efficiency of markets is hampered by poor integration and underdevelopment of property-rights arrangements in the economy. Under these circumstances, the ability of the private sector to solve challenging coordination problems is suspect, and there may be a significant scope for government policy to facilitate development (Aoki, 1997). Union structure is not exceptional from this norm. There will be a better chance for the state to manipulate the union movement at the early stage of the collective bargaining era. As a certain type of union is institutionalized, it becomes more and more difficult for the state to turn back the tide. Why in Asia? In this regard, the states in Asia had unique conditions. First, the Western patterns of industrialization (e.g., England) in which a wealthy entrepreneurial class provided both the capital and the know-how was not the pattern found in these late developers (Gerschenkron, 1962). Rather, the state was the prime agent of industrialization. In the absence of strong regional or landed elites, consolidation of power by military groups or exclusionary political parties meant strong elite unity (Aoki, 1997; Koo, 1987; Deyo, 1987). Thus, with little incentive to distribute political rents in favor of any particular economic class, political leaders and bureaucrats in Asian countries were able to treat the government of their country as if it were the management of a household. More important is, however, the coincidence of the collective bargaining era with the Cold War. The countries in this region have historically been in different stages of industrialization. By the end of the second decade of the 20th century, the relatively earlier developers had already seen the emergence of substantial labor movements and dramatic episodes of worker protest against dismal working conditions under the demonstration effect of Western labor movements and the influence of the international Communist movement. Even in these countries, however, collective bargaining had to wait due to the military powers of imperialism until the end of World War II. In all these countries, therefore, collective bargaining (in substantial form) started under the Cold War.
Enterprise Unionism in Asia
169
The Cold War politics strengthened state autonomy, power, and legitimacy for repressive labor policies. As seen from the perspective of the American Cold-War strategy, there can be no doubt that the Pacific Rim region, extending from Japan to Indonesia, formed a continuous space for the establishment of hegemony in Asia (Yoshimi, 2003). American geopolitical concerns thus stimulated attacks on vested economic interests and supported economically interventionist states in this region (Evans, 1987). These unique conditions altogether created and maintained the powerful state and dictated the nature of industrial relations in Asia during the Cold War. Why Not in the West? Unlike in Asian countries, history of the collective bargaining era in the West goes back to the early 20th century, and industrial relations systems as we know them today had acquired their essential characteristics by the 1950s (Clarke, 1993). Thus, the socio-political environments in which the actors were immersed were clearly different from those in Asia. No Western state either could or did attempt to implement enterprise unionism at the onset of the collective bargaining, which in turn resulted in one or another type of horizontal unionism, as predicted in our framework. In Britain, collective bargaining had become a well-established practice by the end of the 19th century (Webbs, 1911). A major source of authority was ‘‘custom and practice’’ and the alternative of legal regulation was never a serious possibility. The state whose own power had been forged in opposition to royal absolutism rejected an authoritarian response to challenges from below, in fear of provoking even stronger challenges (Fox, 1985). A reluctance to use full weight of the state’s powers allowed space for workers’ organizations to build up a tenacious, if shadowy, existence (Edwards et al., 1998). As the role of the state became increasingly defined as that of maintaining the rule of law, the state was not available to employers to help them to resolve their problems, though it would certainly act to give any consistent support to a long-term policy of rationalizing control of the workplace. In this ‘‘early developer’’ (Gerschenkron, 1962), employers, unions, and the state were all strong enough to maintain defensive positions, while too weak to organize radical departures from the tradition of compromise and muddling through. The structure of contemporary British trade unions reflects their slow historical evolution and displays a complex pattern with no underlying organizational logic. Today, most significant unions are to some extent general unions: multi-occupational and often multi-industrial (Adams, 1995).
170
DAE YONG JEONG AND JOHN LAWLER
By 1914, Germany had the largest and best-organized working-class movement in Europe, which inspired fear among other classes, and growing social tensions and domestic problems. However, the government, which feared the worst situation of becoming involved both in a civil war and World War I, avoided an authoritarian response to labor challenges. During World War I, the working class was given a more important place in society and politics through its sacrifice, and trade unions and collective bargaining were legitimized (Berghahn & Karsten, 1988). When World War I was concluded the socialist government convened a meeting of employer and union representatives, which resulted in the Stinnes-Legien Agreement. The hope was to find, through consultations with them, a joint solution to the problems of demobilization and of restarting peacetime production in a defeated country. Through the Agreement, the employers recognized the right of the unions to represent the interests of workers and to negotiate collective agreements at multi-employer level, and works councils elected by all employees at the plant and enterprise level were established as a compromise (Jacobi et al., 1998). After 1945, Germany brought back the works councils and worker’s participation on boards of directors that it had initially introduced just after World War I. Union leaders established a single national federation composed of a small number of industrial unions under the influence of CIO representatives who had come with the American armed forces to Germany (Lo¨sche, 1973). Since then, the main German unions have faced little organizational competition (Jacobi et al., 1998). During and just after World War I, the U.S. government launched a ‘‘relentless campaign of suppression’’ against socialists, anarchists, and after the Russian revolution, communists (Taft, 1964). The major recession of the early 1920s and the conservative dominance of politics during that time further undermined the power of the socialist labor movement which declined propitiously (Laslett & Lipset, 1974). However, the AFL stood firm with business unionism and adroitly escaped from harsh repression by the state throughout this period of unrest. Even though union membership and militancy increased significantly in the U.S., just as it did in Europe, American workers did not pose a serious and credible threat to the continuation of the capitalist system (Dulles & Dubofsky, 1993). In 1932, F. Roosevelt supported both the unions and the practice of collective bargaining to revitalize the depressed economy. In the National Labor Relations Act, passed in 1935, many of the tactics still used by employers to thwart unionization and company unions were made illegal (Pencavel, 2003). In this new policy environment, there was a rapid
Enterprise Unionism in Asia
171
unionization of workers and a dramatic increase in the practice of collective bargaining. By certifying unions on a plant-by-plant basis, it also made any evolution toward multi-employer bargaining very difficult (Adams, 1995). In the postwar period, unionists emerged as heroes for their activities during the war, and labor movements reemerged strong and vigorous. The postwar militancy of labor then spread to industries and aroused the public. The state revised the basic labor law – Labor Management Relations Act of 1947 (the Taft-Hartley Act) – to restrict labor’s power, and the Act seriously harmed several strong unions (Burtt, 1979; LeRoy & Johnson, 2001). However, important here is that the state did never attempt to eliminate the horizontal labor unionism, nor did it attempt to implement an enterprise unionism. On the part of labor, ‘‘pure and simple’’ unionism continued to dominate the labor movement. In 1949–1950, the CIO expelled 11 national unions that were considered to be Communist-led, dropped its affiliation with the World Federation of Trade Unions (which the AFL had long opposed as an instrument of Soviet policy). In 1953, the AFL expelled the International Longshoremen’s Association on grounds of racketeering, following investigations by the New York State Crime Commission into criminal activities on the New York docks (Burtt, 1979). Since the merger of the two unions into the AFL-CIO in 1955, the ‘‘amorphous’’ industrial union structure, along with the decentralized bargaining structure, has largely remained intact.
A COMPARATIVE STUDY OF NINE ASIAN COUNTRIES In this part, we conduct a comparative study of nine countries in Asia to test the validity of the framework proposed. The nine countries are mainly divided into three groups, depending on the type of state labor policy: enforcement of enterprise unionism; centralization/laissez-faire (non-enterprise unionism); and dual unionism/gradual transition (middle-ground). In the following sections, we examine whether a clear correspondence between state labor policy and union structure has emerged in each of these groups. Model 1: Enforcement of Enterprise Unionism The first model can be dubbed Enforcement of Enterprise Unionism, in that having crushed radical and aggressive horizontal unions at the onset of the
172
DAE YONG JEONG AND JOHN LAWLER
collective bargaining era, the state (along with management) went further to enforce an enterprise union system. Japan, Korea, Thailand, Malaysia, and the Philippines belong to this model. Depending on the consistency of the state labor policy, the countries in this model are further divided into two subgroups: on the one hand, Japan, Thailand, and Malaysia, where the state has consistently implemented an enterprise union system; and on the other, Korea and the Philippines, where the state has implemented an enterprise union system, but with some interruptions. Group 1: Consistent Enforcement of Enterprise Unionism Japan. The Japanese labor movement emerged in the late 19th century, without any realistic possibility of establishing craft unionism, largely due to the unique but ruthless labor policy of the Japanese developmental state.5 Nonetheless, Japanese unionists were imbued with horizontal unionism, and almost all of their unions were horizontal (particularly, industrial) unions (Taira, 1970; Komatsu, 1971), despite the state suppression and management opposition, throughout prewar Japan. In spite of the substantial labor movement and dramatic episodes of worker protest in prewar Japan, collective bargaining, however, had to wait until the end of World War II due to the military powers of imperialism. In Japan, the so-called ‘‘three fundamental labor laws,’’ fostered by SCAP (Supreme Commander for the Allied Powers) in the 1945–1947 period, opened the collective bargaining era in real terms (Yanaga, 1949). Workers attempted to go beyond the plant, where they initially organized themselves to carry out ‘‘production control’’ for survival in the hellish economic crisis, to build powerful industrial unions and succeeded to a significant extent under the Communist leadership (Takahashi, 1965; Price, 1997; Yamamoto, 1977; Scalapino, 1965). The state, with SCAP, which was determined to make Japan into the ‘‘fortress against Communism in Asia,’’ denied trade union rights to Communist-backed organizations by revising the original laws, and with the breakout of the Korean War on June 25, 1950, it carried out the sweeping red purges, which dismissed 12,168 activists, many of whom filled key leadership posts in all the major industrial federations (Numata, 1972; Garon, 1987). The state, SCAP, and Nikkeiren went further to eradicate industrial unionism, which was perceived as the major source of the militant postwar labor movement. With the bitter defeats of the ‘‘prince of industrial unionism’’ (Densan) in the early 1950s and of labor through the Miike strike in 1960, actual functions and organizational units moved toward the enterprise
Enterprise Unionism in Asia
173
union under cooperative union leaders, who had initiated the ‘‘second’’ (or breakaway) unions (Kawanish, 1992; Fujita, 1955, 1959; Ota, 1971). Plant unions of a multi-plant enterprise later federated into an enterprise union to match the corporate structure of the firm (Nomura, 1994). Thus arose enterprise unionism, an equilibrium between job security conscious moderate enterprise unions, management in fear of a ‘‘ghost of labor militancy,’’ and the state with goals of political and industrial stability. This ‘‘micro corporatism’’6 has been reproduced through the gradual institutionalization of long-term employment for limited union members (with elastic buffers of auxiliary workers) from the late 1950s onward, ideological dissemination, surveillance and elimination of heretics from the workplace, and (when necessary) simple and brutal repression, among others (Kumazawa, 1989; Gordon, 1998; Nomura, 1993). Consequently, the enterprise union accounted for more than 90 percent of all the unions and organized workers in the late 1990s in Japan (Japanese Ministry of Labor, 1999). It is noteworthy that after the critical juncture (from the late 1940s through the 1950s), the state’s role in labor relations, including reproducing the enterprise union system, has been relatively indirect, though consistent, in Japan (Johnson, 1995), in contrast to the other ‘‘later-late developers’’ (Dore, 1979) included in this study. Japanese management, which successfully recovered its capability to manage shopfloor matters with the recovery of the Japanese economy after the Korean War, has been able to ‘‘fight with one hand behind its back and still achieve in most situations a verdict that it finds tolerable’’ (Fox, 1974, p. 279). Simply, it is not in the interests of the state to be directly or explicitly involved in labor relations when the direction of labor relations clearly points to the ends it desires. The continuous, consistent efforts of the state in implementing an enterprise union system, along with management’s serious endeavors to keep potential industrial conflicts within the boundary of the company, has left a strong legacy in the Japanese industrial relations system. Despite the erosion of Japanese-style management with a burst of non-regular, contingent workers under the recent Asian crisis and globalization, the enterprise union system has proven itself to be irreplaceable. In fact, several small community-based unions have recently been organized in the transformational stage of industrial order (Dean & Guevara, 2001; Benson & Debroux, 2000), but they have had little impact on the union system and remain left out of the main union system. The share of enterprise unions still remains over 90 percent today7 (Japan Institute for Labor Policy and Training, 2004). The enterprise union members and union officials, who are mostly ‘‘core’’ male
174
DAE YONG JEONG AND JOHN LAWLER
employees, are highly unlikely to give up their own privileges for the nonregular, contingent workers, the majority of whom are women (Broadbent, 2001; Takaki, 2002). Malaysia. In 1940, laws were introduced to regulate trade unions and trade disputes in Malaysia. The Trade Union Ordinance of 1940, however, could not be implemented because of the outbreak of World War II and the occupation of Malaysia by the Japanese military (Arudothy & Litter, 1993). The history of unionism and of collective bargaining in Malaysia thus begun after World War II with the state’s renewed enforcement of the Trade Union Ordinance in 1946. In the political liberalism of immediate postwar, workers in Malaysia rapidly formed unions under the direction of the Malayan Communist Party (MCP). The MCP, denied legal political status, sought to build up its support by forming unions in different industries and occupations under centrally organized general unions. The general unions were in turn centrally controlled under an umbrella organization, known as the Pan-Malayan Federation of Trade Unions (Ayadurai, 1993). By 1947, there were nearly 300 unions with a total membership of about 195,000 (Sharma, 1985, p. 50). However, a state of emergency was declared in 1948, and legislation was passed to outlaw omnibus unions (Henry, 1983). The state justified the declaration of a state of emergency throughout the country on the grounds of growing political insurgency and terrorism, and assumed extraordinary powers of arrest and detention without trial against anyone suspected of working for the Communist cause. Among those arrested were many unionists. During 1947 and 1948, union membership declined by 64 percent, while the number of registered unions fell by 42 percent. Within this context, unions were regulated as political rather than economic institutions (Arudothy & Litter, 1993). The state paved the way to an enterprise union system at this point. The state authorities have since then minimized union reach and size. Unions are not allowed to operate across trade, occupation or industry, each of which is narrowly defined (Galenson, 1992; Frenkel & Peetz, 1998). Under the Trade Union Ordinance of 1959, apex bodies are incorporated as societies, with no rights to represent member unions. In this way, general unions are effectively barred (Kuruvilla & Arudsothy, 1995). This indicates that the state is more resistant to a certain type of horizontal unionism that disturbed its stability at an earlier time. The state went further to balkanize the labor movement. Public sector unions have been prohibited from cooperating with private sector unions.
Enterprise Unionism in Asia
175
Within the public sector, workers in all statutory bodies have been restricted to forming enterprise unions.8 In addition to these direct methods, the state authorities also resorted to indirect methods to split the labor movement: ethnic antagonism, promoting moderate rival unions, military and police intimidation of activists, and restriction of multi-bargaining agreements, among others (Jomo & Todd, 1994; Ayadurai, 1993; Bhopal, 2001). The union movement suffered another major setback when the state enacted the 1967 Industrial Relations Act, barring collective bargaining or strikes over matters involving promotions, transfers, recruitment, retrenchment, dismissal, reinstatement, or the allocation of duties. Carefully cultivated relations between firms’ personnel officials and Ministry of Labor officials ensured quick interventions by the state in their favor. Such relationships helped to avert potentially disruptive labor activism in the 1970s and early 1980s when real manufacturing wages remained fairly stagnant (Rasiah, 2001). In the private sector, the registrar was the principal influence on the pattern and growth of unionization. The proliferation of small unions was due largely to the registrar’s application of the exclusionary policy. In fact, workers in the private sector attempted to establish industrial unions (e.g., the Electrical Industry Workers’ Union and the National Union of Transport Workers) in the early 1970s, but their efforts were thwarted by the registrar on the grounds that they were dissimilar industries. In response to a subsequent Malaysian Trade Union Congress (MTUC) announcement of its intention to ballot workers in support of a national union, the state announced that an industry union would not be allowed irrespective of the outcome; only enterprise unions would be permitted (Bhopal & Todd, 2000). Thus, enterprise unions were a distinct feature of Malaysian unions even before the arrival of the Look East Policy of the 1980s. Half of all unions were already organized on an enterprise basis in the early 1980s, though the private sector had a lower ratio of enterprise unions than the public sector (Wad, 1988, p. 126; Wad & Jomo, 1994, p. 218), reflecting the difference in the nuance of the state labor policies by that time between the two sectors. The ‘‘miracle’’ of Japanese economic success has provided Malaysia with the best developmental model to emulate, while the role of enterprise unions in Japan offered the Malaysian state a justification for its switch to a more explicit labor policy of establishing an enterprise union system in the private sector since 1983. In 1985–1986, out of 28 new unions registered, 27 were enterprise unions (Kuruvilla & Arudsothy, 1995, pp. 172–173). In early 1989, the Trade Union Act was amended specifically to accommodate
176
DAE YONG JEONG AND JOHN LAWLER
enterprise unions. Workers in any single enterprise can now form a union, regardless of the earlier occupation/trade/industry criteria, which had been used to emphasize the homogeneity of union membership. As a preliminary step, statistics were published on Malaysia’s enterprise unions. Until that time these statistics had remained unpublished because of the general criticism leveled against the Malaysian government that it fostered weak peanut unions (Arudothy & Litter, 1993). The state’s consistent labor policy has escalated the proportion of enterprise unions in the private sector. Today, enterprise unions account for 80 percent of all unions in Malaysia (Malaysian Ministry of Human Resources, 2005). Important to note is that the relative easiness with which enterprise unions can be established in Malaysia clearly derives from the early success of eliminating militant, horizontal unions, and following policies of limiting union reach and size. Thailand. Prior to the end of War World II, less than 2 percent of the workforce was engaged in industrial production. It is thus understandable that the first trade unions in Thailand emerged after World War II (Wehmhorner, 1983). The Thai experience shows a surprising similarity with the Malaysian case in the pattern of the state labor policy and its legacy throughout the collective bargaining era. Like the Malaysian case, the history of the Thai labor movement started with the state’s crackdown of pro-Communist unions under the Central Union of Thailand in the postwar political liberalism (Manusphaibool, 1993). Between 1946 and 1947, 168 strikes were reported, with 109 reported to have taken place in 1947 alone. This industrial activism was accompanied by the gradual building of labor organization at workplace, industry, provincial and national levels. Labor organizations were established in most industries under a centrally controlled national organization. Workers’ industrial militancy and the embracing of radical ideologies were perceived as a threat to the new military leaders and their political and developing business interests. The government stepped up its policy of repression. From mid-1948 through to late 1952, hundreds of union activists were banished, jailed, fined, deported, harassed, terrorized, and placed under police surveillance (Brown, 2004). The state was actively supported by the United States, which perceived Thailand as a front-line state in the global fight to curb the growing influence of communism (Fineman, 1997). Collective bargaining was formally recognized by the Labor Act of 1956. However, the state abolished this legislation in 1958 and inaugurated the socalled ‘‘dark ages of Thai labor’’ from 1958 until 1972 (Galenson, 1992). Strikes were made illegal and unions were banned under the claim that they
Enterprise Unionism in Asia
177
were ‘‘obstacles to economic development’’ and ‘‘gateways for communism to enter Thailand.’’ Many union leaders were arrested, interrogated, and imprisoned. Organized labor was effectively crushed and the remnants forced underground (Wehmhorner, 1983; Manusphaibool, 1993; Mabry & Srisermbhok, 1985). Importantly, these repressive controls, imposed under the authoritarian, military-led regimes at the onset of the collective bargaining era, established the foundation of an enterprise system in the following decades. The early to mid-1970s marked an important moment in the development of organized labor in Thailand. Particularly, during the short-lived democratic regime from 1973 to 1975, there was a burst of worker activism. For example, there were 500 strikes and 178,000 strikers in 1973 alone. Demonstrations were truly at the grassroots level, occurring in response to every conceivable complaint. Some 150 local unions were formed (Galenson, 1992). Through their activism, workers demonstrated an intention to build organizations at the workplace, industry and national levels, through which they could not only deal with employers, but also join with a coalition of social forces struggling for a range of broader social and political rights (Morell & Samudavanjja, 1981). However, the burst of worker activism in this period could not withstand the state balkanization policy through the Labor Protection and Labor Relations Ordinance of 1972, which limited the formation and legislation of trade unions to a single establishment or province. This policy was deliberately designed to encourage only the development of weak and small labor organizations (Sharma, 1985). On the employer’s side, from the late 1970s onward, some private-sector employers, aided by a ready supply of labor power, have systematically sought to combat and inhibit labor organizations. The strategies used to this end include the widespread use of short-term contracts; closing militant sections of factories; sacking the promoters of unions; drawing out bargaining negotiations and encouraging splits within union leadership; and closing factories to reopen them with newly recruited workforce. These and other measures have been reinforced with threats, beatings, and the strategic murder of a few union officials and activists (Siengthai, 1999; Brown, 2004). Not to mention, the employer’s anti-unionization measures would not have been effective without the state’s contribution; for example, the failure to close the many loopholes in the labor law, not enforcing employer compliance and a continuing refusal to ratify ILO conventions that cover workers’ basic rights. More particularly, however, specific elements within the state have sought to restrict worker autonomy. In the late 1970s, for example, the military (which has a long history of involvement in labor affairs) established its own organization – the National Free Labor
178
DAE YONG JEONG AND JOHN LAWLER
Congress – to compete with and promote divisions within organized labor at the peak council level. The military has also consistently fostered competition within labor councils (Brown & Frenkel, 1993; Brown, 2001). The combined effect of these changes was the dominance of enterprise unions, with surviving, but fragmented and weak horizontal unions. For example, of approximately 600 recognized unions that existed in 1989, about 400 were enterprise unions (Lawler & Suttawet, 2000, pp. 220–221). Similarly, Frenkel (1993, p. 311) reports that enterprise unions accounted for 65 percent of all unions in Thailand in the early 1990s. The Thai government went further to isolate state enterprise workers from the union movement in the early 1990s. Under the new legislation enacted in April 1991, unions in state enterprises, which had been the organizational heart of the union movement and included more than half of its unionized workers, had to be reorganized along strict enterprise lines. Furthermore, unions were banned, and more circumscribed employee ‘‘associations’’ were allowed instead. This three-year ban, spanning the 1991–1993 period, on state enterprise unions substantially weakened the labor movement (Levin, 1997). The new ‘‘associations’’ were stripped of the rights to strike and to bargain collectively. Furthermore, they were prevented from belonging to national labor federations (Manusphaibool, 1993). The new legal framework also placed further impediments in the path of union formation in the private sector. Before private sector workers engage in concerted action, 50 percent of the employees must vote to back a work stoppage. Action is also prohibited until the union negotiators have gone through the negotiation process with management, and unlawful strikes can result in the dismissal of the strikers. The state also facilitated competition and divisions among union leaders to fragment organized labor (Levin, 1997; Brown & Frenkel, 1993). Again, justifications for these changes were couched in terms of meeting the demands of the economy as well as maintaining national security (Brown, 2001). Consequently, the number of enterprise unions has increased from 43 in 1994 to 46 in 1998 in the public sector and from 888 in 1994 to 1,239 in 1998 in the private sector. Today, enterprise unions account for 90 percent of all labor unions in Thailand. There are only a few industrial unions in the metal and textile industries (Thai Department of Labor Protection and Social Welfare, 2005). Group 2: Inconsistent Enforcement of Enterprise Unionism Korea. During the Japanese colonial occupation (1910–1945), Korea had already seen the emergence of substantial labor movements and dramatic
Enterprise Unionism in Asia
179
episodes of worker protest against dismal conditions of work and Japanese colonization under the influence of the international Communist movement. Labor unions were established by industrial line under a centrally controlled umbrella organization. Extreme repression by the colonial government radicalized the labor movement, forcing it to go underground with closer ties to the Communist movement (Kim & Kim, 1970; Kim, 1987; Yun, 1991). Collective bargaining was thus unthinkable until the end of World War II. After liberation from Japanese colonization in 1945, the radicalized workers resurged as a political force and sought to take over factories abandoned by Japanese owners. Labor unions were rapidly established by industrial line under a communist-led National Council of Korean Trade Unions (NCKTA), which grew to include more than 500,000 members. The violent struggle between NCKTA and the rightist forces backed by the U.S. military government was brought to an end when the Korean Communist Party was outlawed by the U.S. military government in South Korea in March 1947. By this time, hundreds of leftist labor leaders had been killed or executed, and thousands were imprisoned (Kim, 1999a; Park & Lee, 1995; Haggard & Cheng, 1987; Lee, 1984). The U.S. military government went further to constrain collective bargaining agreements within ‘‘plant, establishments, or other business units’’ by the Trade Union Act of 1953 (Kim, 2005, pp. 26–27). These events paved the way to the development of enterprise (plant) unions and of enterprise level collective bargaining under the Rhee regime (1953–1961) (Gong, 2000). Under the rhetoric of national security against Communist North Korea, the Park regime destroyed the vertical linkages of unions in 1971–1972, invalidated previous labor legislation by freezing workers’ rights of collective agreement and collective action (Article 9, the Emergency Decree on National Security of 1971) during the Yusin period (1973–1978), and eradicated all forms of political opposition by the Presidential Emergency Measure of 1974 (Hong, 1994). Furthermore, the following Chun regime sent labor leaders to the notorious ‘‘purification’’ camps, crushed independent unions, and divested national industrial unions of any meaningful role. The involvement of outside groups and activists was outlawed in 1980 (You, 1995). The Chun regime finally mandated the formation of enterprise unions by law in early 1981, along with the enforcement of the establishment of labor-management councils within the workplace through the Labor-Management Council Act in 1980 (Kim, 1993). By 1987, all collective activities of workers had been forcefully sanctioned with the justification of national economic success and the threat of Communist North Korea. Accordingly, it was not the Ministry
180
DAE YONG JEONG AND JOHN LAWLER
of Labor but the Korean Central Intelligence Agency (KCIA) or the Blue House (the counterpart of the White House in the U.S.) that held actual control over labor policies. In this sense, Kim (1995, p. 174) argues that labor relations in Korea is ‘‘state-labor relations’’ rather than ‘‘labormanagement relations.’’ The enterprise union structure was seriously contested by labor for the first time after the reinstallation of rights to bargain collectively and to select union structures in 1987. Labor’s challenge was highlighted by its establishment of the ‘‘National Headquarters Struggle for Revising the Labor Law’’ from mid-1988 to late 1989 when the state’s repressive policy returned against the autonomous labor movement, resulting in a dramatic decrease in the number of these unions and their membership (Jeong, 1995). Important to note is that the enterprise union system remained intact against this background. The survival of the enterprise union system has proven that an enterprise union system, even mandated and maintained by force, can generate to a certain extent ‘‘enterprise-consciousness’’ among its members of the privileged sector. The legacy has been reinforced through ‘‘microcorporatism’’ between privileged enterprise unions (core workers), management (in fear of labor militancy), and the state (with the desire for political and industrial stability) at the expense of the mass of peripheral workers since then,9 as was the case in Japan. Having said that, we should point out two major factors which made the Korean experience to deviate from the Japanese case. One factor is the unsuccessful attempt of centralization and manipulation labor policy by the Park regime in the 1961–1971 period. The Park regime after the military coup of 1961 reorganized unions along industrial lines under a new labor federation (FKTU),10 allowing relative freedom of collective bargaining and collective actions in this period (Lee, 1993; Park & Park, 1989). In fact, industrial unions extended their organizational bases through penetration into industrial complexes. However, upon intensification of political conflicts between the state and the FKTU in 1969–1972, the state severely repressed labor by purging labor leaders and destroyed the vertical linkages of the FKTU and the industrial unions with local unions (Song, 1989; Koo, 2001). Importantly, this relatively liberal, industrial union period complicated labor relations in years to come, as it provided workers with chances to learn the importance of centralized labor movements and to develop close ties with outside activists, such as church groups and student activists, which played significant roles in supporting grassroots labor struggles (Lee et al., 2001). These were crucial resources for the reemergence of militant labor organizations and labor actions beyond the enterprise boundary.
Enterprise Unionism in Asia
181
On the other hand, Korean enterprises – ‘‘dependent bourgeoisie’’ (Koo, 2000; Amsden, 1997) – lacked effective mechanisms to internally reproduce the legacy at least until the late 1980s. When management was unable to handle labor struggles, state agents had to be brought in. The harsh but intermittent suppression, instead of silencing labor dissent, had only laid bare the role of the government in labor discipline, and consequently, served to convince the workers of the need for political struggles. The state labor policy, therefore, created the preconditions for the cyclical surge of workingclass struggles. In short, these two major factors provided the grounds for the unique features of Korean enterprise unionism today – relatively adversarial labor relations, the persistence of political unionism, and the esprit de corps – despite the dominance of enterprise unions.11 These unique features of the Korean union movement were well manifested in the swift reemergence of industrial unions in some industries after massive layoffs following the 1997–1998 financial crisis and the 1997 revision of labor laws, which made it easier to establish industrial unions. In the two-year period of 1998–2000, almost 20 industrial unions formed and their combined membership reached 32.8 percent of the total union membership in Korea as of July 2001 (Kim & Kim, 2003, pp. 363–364), though the number of enterprise unions still accounted for over 90 percent in 2003 (Korea Labor Institute, 2004). In this context, Kim and Kim (2003) went further to predict a rapid, continuous increase in the membership of industrial unions up to 40–50 percent of the total union membership within a couple of years, and for future labor unionism to be a mixture of enterprise and industrial unions in Korea. This observation seems to be reasonable enough when we consider the historically confined tension within the current labor movement. Whereas industrial unionism is in real force, the current enterprise unions would not easily give up their vested interests in the system. The Philippines. Owing to Spanish and American colonialism that severely suppressed union activities, collective bargaining was infeasible in the prewar Philippines (Amante, 2004; Jimenz, 1993). The serious balance of payments crisis after World War II coincided with serious labor and peasant unrest, and this became part of a larger politico-economic crisis highlighted by the threat of a communist-led armed rebellion in the Philippines. The government response to the labor unrest initially took the form of military harassment. Then, it adopted an iron-fist policy, driving a number of the communist-led Congress of Labor Organizations (CLO) leaders to go underground and join the communist guerrillas. Eventually, the government
182
DAE YONG JEONG AND JOHN LAWLER
outlawed the CLO in early 1951 (Ofreneo, 1995). The Industrial Peace Act of 1953, which opened the collective bargaining era in the Philippines (Wurfel, 1959), officially encouraged enterprise unions to remove communist influences from the labor movement. The state broke the union movement’s links with militant peasant organizations in Central Luzon and furthered ‘‘economic unionism’’ through collective bargaining (Villegas, 1988; Hutchison, 2001). The foundation of an enterprise union system was thus established at the onset of the collective bargaining era. By the late 1960s and early 1970s, Filipino workers found collective bargaining to be inadequate and increasingly resorted to unauthorized collective action to improve their economic position. In 1971, the number of workers’ strikes rose by 50 percent to 157. Increased government repression and the failure of the moderates to gain concessions played a major role in the formation of the radical worker faction (Parsa, 2000), though communists did not exist in the early 1970s in the Philippines (Kang, 2002). However, the Martial Law of 1972 radically demobilized organized labor and its allies, leftist students, and intellectuals. The leaders of militant unions were harassed, imprisoned, and even killed (Nemenzo, 1988). With the help of the Unites States, Marcos restructured the military and the police to centralize control under his own command (Kang, 2002) and drastically turned its labor policy to centralization and manipulation in this period. Under Martial Law, the highly decentralized unions were restructured along industrial lines, and all trade unions were required to be affiliated with the government-controlled Trade Union Congress of the Philippines (Kuruvilla, 1996; Villegas, 1988). In fact, this centralization and manipulation strategy turned out to be successful in retaining the relative calm on the labor front (Parsa, 2000). Ironically, however, this period of centralization in the 1970s left an ambiguous legacy on the union movement in the years to come. Despite its rhetoric, the registration requirement was not strictly enforced. The government allowed the unions to initiate and carry out restructuring among themselves. This caused serious dislocation within the union movement, as various groups clashed over a unified and rational structure. Ramos (1990, p. 111) suggests that the Marcos government may have had a political interest in creating this situation. The national economic collapse in the early 1980s enabled the Muslim and communist insurgencies to revive again and the radical and militant KMU (May First Movement) to undertake ‘‘social movement unionism’’12 against the authoritarian Marcos regime (West, 1997). James Nach of the U.S. Embassy estimated in 1986 that the communists had between 8,000 and 10,000 regular troops and perhaps a half-million supporters in the
Enterprise Unionism in Asia
183
Philippines (Kang, 2002). However, the strikes lost and labor militancy declined in the late 1980s when the Aquino regime enforced strict rules on strikes and management refused to bow to the demands of striking workers, accusing the strike-prone militant federations of being communistdominated (Ofreneo, 1995). With the National Labor Relations Law enacted in 1989, the Philippine government returned to its earlier policy of decentralization. Since then, the government has made various legal and administrative efforts to depoliticize industrial bargaining through the enforcement of enterprise bargaining and proactive bipartite programs such as labor-management councils at the plant level (Hutchison, 2001). This change in the state labor policy certainly furthered the fragmentation of the union movement, and there has been no major change in the state labor policy throughout the Aquino, Ramos, and Estrada governments (Frenkel & Kuruvilla, 2002; Bello, 2000). On the part of unions, almost all the labor federations have engaged in general unionism, attempting to organize workers across industries (Ofreneo, 1995). But they have not been able to develop a unified union movement in repeated national economic crises with manufacturing declines and chronically high levels of unemployment and underemployment (Deyo, 1997). The dispersion of industries across a country with over 7,100 islands has also been attributed to the frustration of efforts toward unity (Jimenz, 1993). The 1997–1998 financial crisis accelerated growing labor causalization, subcontracting and labor-only contracting in a context of growing unemployment and sharpened divisions within the working class. Union weakness has led to little organized opposition to employer strategies based on the logic of competition (Hutchison, 2001; Frenkel & Kuruvilla, 2002). Therefore, it is not surprising to observe the highly fragmented union movement today. The movement is weak with only 9 percent of the workforce (3.6 million) organized and highly fragmented with 8,822 unions affiliated with more than 174 federations (Frenkel & Kuruvilla, 2002). Furthermore, only 3.7 percent of workers are covered by collective agreements (ILO, 1997, p. 248). Reflecting the state labor policies throughout the collective bargaining era, enterprise unions account for 71 percent of all the unions, whereas industrial, regional, and occupational unions comprise the rest as of July 2004 (Philippine Bureau of Labor Relations, 2004). Model 2: Non-Enterprise Unionism (Centralization/Laissez-Faire) Indonesia and Hong Kong significantly deviate from the other cases in this study, in that neither state has ever seriously attempted to balkanize the
184
DAE YONG JEONG AND JOHN LAWLER
labor movement, which is hypothesized to result in an enterprise union system. The state labor policies of these cases have been clearly different from each other, however. In contrast to the labor policy of centralization in Indonesia, the Hong Kong state has adopted a laissez-faire policy, mainly due to its political relationship with China. Indonesia. Under Dutch occupation, no mass-based union movements could develop in Indonesia until 1945 (Gall, 1998). In 1946, the All-Indonesia Central Workers’ Organization (SOBSI) was organized along industrial lines (but with craft unions having full equality) under the direction of the Indonesian Communist Party (PKI) (Hawkins, 1962). By the mid-1950s, union membership reached to about 2 million, representing 20 percent of workers, with half of those in the SOBSI (Manning, 1993). Law No. 18 of 1956 and Basic Law (No. 14) of 1969 opened the collective bargaining era in Indonesia (Nayar, 1996; Edwards, 1996). Having realized the inevitability of dealing with the unions under the labor laws, some employers tried to foster unions within their own plants, and thus a number of this type of unions existed. Yet, they were far from the common pattern of union structure. The SOBSI organized almost all unions either by industrial line or as (geographically confined) general unions like the AFL-CIO in the U.S. (Hawkins, 1962). However, an ‘‘attempted coup’’ in 1965 by young army officers (Vasil, 1997; Kingsbury, 1998) was used as the pretext for the armed force’s taking power and destroying the PKI and the SOBSI. The military killed about half a million Indonesians, and imprisoned 200,000 to 300,000 individuals, many of whom died in prison or on the infamous Buru Island (Botz, 2001, p. 114; Hadiz, 2000, p. 246). At the height of the Cold War, the military, which executed the slaughter, was portrayed as the ‘‘savior’’ against the communist threat (Neher, 2002). The Suharto government then instituted a bureaucratic authoritarian state where power was limited to the state itself, and the state power was further strengthened when the United States and international agencies poured aid and grants into the country during the era of the New Order (1965–1999) to ensure that Indonesia would not fall to the Communists (Robinson, 1997). Under the New Order, the military was cast again as the ‘‘protector’’ of national stability and industrial peace. The armed forces’ involvement in its military capacity in industrial relations was widely known as the ‘‘security approach’’ (Ford, 1999). In terms of union structure, the state through the bloodbath accomplished the necessary dislocations for the development of an enterprise union system at the onset of the collective bargaining era. As seen in other countries, the
Enterprise Unionism in Asia
185
Indonesian state could have utilized an enterprise union system for ‘‘the perceived need to pre-empt the re-emergence of independent, militant tendencies within organized labor’’ (Hadiz, 2002, p. 131). Yet the state preferred another option, centralization in union structure. Perhaps, this choice was taken to overcome its almost insuperable obstacles in achieving nationhood – geographic, linguistic, ethnic, and social heterogeneity. Almost half of Southeast Asians live in Indonesia, which is composed of thousands of islands and is populated by 13 major and hundreds of minor ethnic groups (Neher, 2002). In 1973, the state reorganized the remaining labor organizations into 21 industrial unions under one easily controllable state-sanctioned labor federation, the All Indonesia Labor Federation (FBSI), whose role was strictly confined to the economic realm (Bourchier, 1994; Hadiz, 2001). The state-controlled unions functioned much like Nazi or fascist fronts to control and discipline workers for capital. Workers had no right to organize an independent labor union, could not engage in meaningful collective bargaining, and were denied the right to strike. Those who attempted to organize strikes or independent labor unions were fired and blacklisted, were beaten and tortured, and were sometimes murdered (Botz, 2001). State labor control was further strengthened by eliminating the less directly controllable industrial unions and transforming the FBSI into the even more centralized and hierarchical All-Indonesia Workers’ Union (SPSI) with nine industrial departments in 1985, following the labor unrest in the late 1970s and early 1980s (Rinakit, 1999; Hadiz, 2000). In 1993, the SPSI was divided again into 13 unions for different industries as an attempt to head off the growth of the independent unions and make the SPSI more relevant to workers. Anti-class rhetoric and references to the ‘‘latent threat’’ of communism were actively used in attempts to curtail protests and strike activity, even in the last decade before Suharto was forced from power. Government officials repeatedly claimed that strikes were coordinated and manipulated by outsiders, third parties, and communists (Ford, 1999; Hadiz, 2002), despite the absence of radicalism in the movement (Gall, 1998). Despite a considerable amount of repression, the 1990s saw a resurgence of labor activism driven by local community-based organizational vehicles, often in association with non-governmental organizations (NGOs) (Hadiz, 2001). This was a direct result of brutal state repression of the labor movement and severe restrictions on organizing in the workplace during past decades. It was necessary for labor and student activists to disperse to maintain the seeds of future resistance (Katoppo, 2000).
186
DAE YONG JEONG AND JOHN LAWLER
Meanwhile, the government made political gestures of accommodating to the aspirations of workers as well as to international criticism (Gall, 1998). While transforming the SPSI into the Federation of All Indonesia Workers’ Unions (FSPSI) again in 1995, the government gave workers the nominal right to register industrial unions autonomous from its central body (Hadiz, 2000). Yet, it must be added that the government decrees effectively prevented the formation of independent industrial unions, as special requirements had to be met for recognition13 (Nayar, 1996). In short, the state labor policy of centralization in union structure did not undergo any significant changes for the past four decades at least until 1994. Consequently, there existed only 13 industrial unions under the sole legal trade union center, the SPSI, in 1994. No enterprise unions had existed by 1994 in Indonesia14 (Indonesian Ministry of Manpower and Transmigrations, 2005). The Indonesian case indicates a clear correspondence between the state labor policy of centralization and the legacy (union structure) against the other examples of an enterprise union system. We conclude our analysis of the Indonesian case here for the purpose of cross-national comparisons. In the following paragraphs, we touch upon the rapid change in the state’s labor policy since 1994, which resulted in a dramatic change in the union structure in Indonesia. Faced by a resurgence of labor activism and increasing international criticism of the authoritarian labor policy in the early 1990s, some state officials had begun to recognize the limits of a strategy of control premised on repression (Hadiz, 2002), and the Suharto government enacted PER-OI/ NTN in 1994, which was the first but critical step in developing an enterprise union system by the Indonesian government. This law established employee organizations (SPTP), restricting the activities of these organizations to representing employees of the company in collective bargaining and required the potential leaders with no legal protection to gain the ‘‘confidence’’ of employers (Sulistyaningsih, 2001). About 60 SPTPs were established at the plant level in 1994 alone (Nayar, 1996, p. 6), and around 1,000 companies already operated with SPTPs in the late 1990s (Sulistyaningsih, 2001, p. 7). As noted earlier, this law accompanied the state’s prevention of independent industrial unions in effect. Furthermore, the state, with the 1997 Manpower Law, limited union interests to the company level (Ford, 2000). These altogether clearly indicate the intention of the Indonesian government in this period. The government, which could no longer sustain its tight control over labor by centralization in the changing political context, turned its labor policy around to prevent the union movement from gaining political presentation at the national level, and established the foundation of an enterprise union system for the near future.
Enterprise Unionism in Asia
187
After the fall of Suharto in May 1998, the Habibie government signed ILO Convention 87, recognizing the right of workers to form independent unions (Ford, 2000). In 2000, the Abdurrahman government went further to encourage the establishment of enterprise unions, by the enactment of Law No. 21, which significantly reduced the minimum membership requirement of a union to 10 employees at the plant level (Hanartani, 2005). Most of the existing SPTPs have since turned into enterprise unions.15 The inertia of SPSI-led unions to adequately and effectively respond to workers’ demands at shopfloors has also triggered the formation of unions at the company level. By January 2001, 2,175 enterprise unions had been registered. By the end of 2001, more than 144 national labor unions and approximately 11,000 enterprise unions were registered (Rahayu & Sumarto, 2003, pp. 7–13). The demise of the New Order presented a small window of opportunity for organized labor in Indonesia. However, workers were ill-equipped and unprepared to capitalize a unified labor movement since the fall of Suharto (Hadiz, 2002). The fragmentation of the union movement seems to be also attributed to the geographic character of the archipelago country with more than 20,000 islands (Hanartani, 2005). As of December 6, 2004, 86 labor federations and 145 national unions are registered, and 18,532 enterprise unions (over 90 percent of all unions) dominate workplaces across Indonesia16 (Indonesian Ministry of Manpower and Transmigration, 2005). Weakening of the authoritarian rule and recognizing the right of workers to form independent unions may be a prerequisite for the growth of independent, horizontal unions, but it is not sufficient. Hong Kong. When Hong Kong was still primarily an entrepot economy, a relatively militant and strong union movement emerged, peaking in the 1920s. Craft unions burgeoned in the traditional trades, and the major industries that served mainly the local community and the trading sector proved to be congenial to the growth of industrial unions: large firm size, a stable employment relationship, and a skilled and predominantly male labor force (Ng, Stewart, & Chan, 1997). Labor unions reemerged immediately after World War II. With the enactment of the Trade Unions and Trade Disputes Ordinance of 1948, major features of the postwar industrial relations system were taking shape by the early 1950s. By 1951, there were 197 labor unions (mostly in the private sector) under either the pro-Taiwan Trade Union Council (TUC) or the proBeijing Federation of Trade Unions (FTU) (Chan, 1988, p. 36). Interestingly enough, the state in Hong Kong did not attempt to break down the prevailing Communist-dominated unions in this period, however. Neither
188
DAE YONG JEONG AND JOHN LAWLER
has there been any systematic attempt to incorporate the labor movement or to suppress its more militant elements since then. Rather, the state has deliberately avoided involvement in industrial disputes (Wilkinson, 1994). The Trade Unions Ordinance and Part IVA of the Employment Ordinance have been the two principal statutory instruments that confer and regulate the rights and freedoms of trade unions in Hong Kong, but there are no serious legal impediments upon combinations of trade unions, provided that the relevant union has registered with the Registrar of Trade Unions as required by law (Ng et al., 1997). This unique labor policy of the state is mainly attributed to a combination of political factors: British colonization over the territory until 1997 and its relationship with Communist China (England, 1980). An overt attempt to break down the left-wing unions (mostly industrial unions) would have been foolish of the colonial government (which obviously was militarily indefensible), as it could have provided China with justification for an invasion into the territory (Fosh et al., 1999). On China’s part, the stability of the territory was crucial, for the territory was the gateway for a large proportion of its foreign exchange and foreign trade that provided essential raw materials and the equipment necessary for modernization. Accordingly, the FTU was under instruction to do nothing that might interrupt the flow of China’s earnings (Turner, 1980). The FTU unions thus avoided confrontation with the authorities or employers from the early 1950s onward (with few exceptions – principally, the events of 1967)17 and even became advocators of sound labor management relations for Hong Kong’s stability and prosperity after the 1984 Sino-British agreement on the future of Hong Kong and local political reforms (England & Rear, 1981; Miners, 1988). Non-Communist unions also did not seek an alteration in the political status quo, since British rule was the only realistic alternative to absorption into China (England, 1989). The ‘‘success’’ of voluntarism was indeed built on the equilibrium of the parties involved in the political economy. The state’s ‘‘voluntarism’’ meant that neither recognition of a union nor collective bargaining is mandatory for employers. Given the power imbalance in the workplace, this approach served to buttress managerial prerogatives rather than facilitate labor organization (Levin & Chiu, 1993; Cheetham, 1988). Union organizers and sympathizers are likely to be faced with dismissal (Wilkinson, 1994). With employer preference, the principle of collective bargaining at the plant level was accepted (Levin & Ng, 1995). In short, the colonial government in its wisdom sat back and had employers play the game against weak and poorly organized workers at the workplace level.
Enterprise Unionism in Asia
189
It is also true that the colonial government did favor a policy of ‘‘divide and conquer’’ and attempted to balkanize the union structure as much as it could. The Trade Unions Ordinance placed restrictions upon amalgamation and federation to mitigate against the creation of large power blocs. The law confined amalgamation of unions within the same trade or industry, and required a 50 percent turnout of voting members as well as a 60 percent of the votes cast in a secret ballot. Federation was also restricted to the same trade, industry, or occupation (Ng & Sit, 1989; Turner, 1980). The net result of the law has been a proliferation of small unions. There had been only a few instances of amalgamation (despite the relaxation of the requirement for amalgamation in 1961) by 1979 and one federation by 1981 (England & Rear, 1981, pp. 157–159). These restrictions were also a clear obstacle to the development of general unions. There was only one general union, which was obviously not competitive with unions of other types, in 1991 (Turner, Fosh, & Ng, 1991, p. 49). Consequently, the ‘‘voluntarism’’ has resulted in diverse types of unions, excluding general unions. In 1951, the majority of the 197 labor unions were industrial unions while there were some craft unions and a few enterprise unions (Chan, 1988, p. 36). In 1967, among the 41 National Trade Unions Congress (NTUC) affiliates, 11 were industrial unions and 12 were craft unions, 3 were general unions, and 15 were enterprise unions. Industrial and craft unions shared the majority of the membership (Leggett, 1993a, pp. 225–227). As Hong Kong was being transformed to an export-oriented economy in the 1950s and 1960s, the structural features of the union movement hardly changed. There were 27 enterprise unions in 1978, excluding those organized within specific government departments. Interestingly, in most cases they were based upon Western companies (England & Rear, 1981, p. 156). In the early 1990s, the number of enterprise unions accounted for less than 20 percent among the 450 Hong Kong unions (Frenkel, 1993, p. 311). The legislation governing trade unions changed little during the postwar years (Nish et al., 1996). Neither the FTU nor the TUC did manage to make substantial inroads among the growing industrial working class. Collective bargaining remained voluntary and minimal in scope, there was hardly any industry-wide negotiation machinery, and many workers lacked access to effective channels for seeking redress of grievances and claims arising under existing contracts of employment. Collective bargaining resulting in written agreements at the enterprise or industry level was rare. Wage levels were normally fixed by individual agreements between employers and workers. Therefore, the commonly accepted weakness of Hong Kong labor unions at
190
DAE YONG JEONG AND JOHN LAWLER
the workplace level still applied on the whole to the private sector in the 1990s (Levin & Ng, 1995; Levin & Chiu, 1993). Upon the handover of Hong Kong to Mainland China in 1997, the Hong Kong labor movement substantially restructured its program, being incorporated into the power structure as a new electoral force. As a sequel to the government-sponsored political reforms which democratized the legislative constituency sitting on the Legislative Council, Hong Kong labor unions and their centers began to participate in mainstream politics as quasipolitical parties from the late 1980s onward (Ng & Rowley, 2000; Ng & Poon, 2004). Nonetheless, without any direct government interference in union activities, the major union structure has remained intact. Enterprise unions account for only 18 percent of all unions in Hong Kong today (Hong Kong Labor Department, 2004). The labor movement is still fragmented though the principal groups are now the FTU and the independent HKCTU (mostly in the public sector), with the waning of the influence of the TUC. The workplace culture of labor union recognition has also remained as feeble as it was in the 1970s and 1980s. Enterprises have been lukewarm and equivocal in institutionalizing any practice of ‘‘bilateral collectivism’’ on a formal basis. The practice of workplace labor relations has remained as pluralistic, diversified, and even loose as it was three or four decades ago, except where the enlarged scale of business and the growth of human resource management practices have led to better communication (Ng & Poon, 2004). Model 3: Middle-Ground (Dual Unionism/Gradual Transition) As discussed earlier, a state would not necessarily exercise a repressive labor policy even when it has enough resources, since it must be concerned about potential negative impacts of its repressive actions and its reputation. Instead, the state would choose relatively benign strategies to deal with ‘‘labor problems’’ when such benign policies are perceived to be ‘‘satisfying.’’ Taiwan and Singapore belong to this model. These states attempted to subordinate and splinter relatively docile labor to eliminate the possibility of centralized opposition. Taiwan. In the years after the establishment of Republic of China in 1912, trade unions were formed to protest low wages and worsening working conditions in foreign companies. The early labor movement was transformed by the political struggle between the Kuomintang (KMT) and the
Enterprise Unionism in Asia
191
Chinese Communist Party. Trade unions at that time were pro-Communist. The KMT tried to attract the leadership of the trade unions. However, this effort was unsuccessful and affected the attitudes of the KMT toward workers and the labor movement after 1949 (Chen, Ko, & Lawler, 2003). Immediately following World War II, Japan, which had colonized Taiwan for 50 years and had prohibited labor unionism, was forced to hand Taiwan back to China. Unionization was encouraged, resulting in the growth of more than 100 mainly craft unions. However, after several strikes by the miners’ unions in 1947, labor organizations were banned by the KMT (Frenkel, Hong, & Lee, 1993). After the defeat of the KMT by the Communists on the mainland in 1949, more than one and half million mainlanders, most of whom were KMT officials, soldiers and their families, rushed to the island. They identified mainland China as their homeland and established the island as a base from which to reclaim the mainland while forming the political and military leadership in Taiwan. Political diversification and tension were visible between the rulers (the alien KMT party and capitalists that came from mainland China) and the ruled (the majority of whom were ‘‘local’’ working-class people having come to the island before 1949) in the faltering economic and social conditions (Chen & Wong, 2002). Having learned from its defeat on the mainland, the KMT established a land reform, reforms of the government (and military), and major economic reforms to stay in power. The strong KMT state was the guarantor for social and political stability against the threat of invasion from the Communists, which indeed tried to invade in 1950, 1954, and 1958 (Aspalter, 2001). The martial law of 1947 banned any political opposition and parties. It is for these reasons that the Taiwanese population (i.e., mainly Fujian Chinese and Hakka) focused on economic development and accepted the rule of the mainland Chinese in a very pragmatic way. Economic success was fostered by the institutional setting of the Lenninist structure of the KMT state with enormous economic aid and military assistance from the United States (Frenkel et al., 1993; Cheng, 2001). The KMT government, having learned from bitter experiences with organized labor on the mainland, also prohibited any type of trade union organizations (Lee, 1995). The military security apparatus had extensive manpower and resources to harass and intimidate labor organizers. They threatened to charge labor organizers as communist agitators or other for anti-government conspiracies, and indeed formally charged and convicted them for such accusations in many cases. National security provided the state with a justification for the tight security measures against unions and
192
DAE YONG JEONG AND JOHN LAWLER
potential labor organizers (Cheng, 2001). Under these circumstances it was almost impossible for the workers to develop autonomous unions. In the absence of a mobilized mass, the KMT quickly learned that it could use labor unions not only in administering recently instituted welfare programs, but also in mobilizing workers for defense in case of an attack by the Communists on the mainland. In 1951, the KMT established a working group, which was composed of representatives of various state agents and some pro-government union leaders, to direct the Chinese Federation of Labor (CFL) (Sharma & Sephton, 1991). The KMT then actively organized workers of public enterprises and large private enterprises, nominating or appointing ruling party members as union leaders. The explicit purposes were to consolidate the KMT’s power and to prevent communist infiltration into the workplace (Lee, 1995). Since then, labor controls have remained central in the enterprise. The Labor Union Law (Article 6) of 1975 permitted only enterprise and craft unions,18 excluding industrial and general unions, which could possibly threaten the security of the national system. Craft unionism was permitted, perhaps because it is (or was perceived as) a form of horizontal unionism, which was unlikely (or least likely) to threaten political and industrial stability at the national level; in fact, the amended law of 1975 further restricted craft unions to the district level (Galenson, 1992). The control function of such unions was tightened by many devices; for instance, the prohibition of strikes before the lifting of martial law in 1987 (since then many road-blocks in the path to legal strikes); involvement in union elections, nominating candidates and mobilizing supporters; pressure on employers to encourage the formation of enterprise unions; mandated welfare programs at the enterprise level; and requirement of factory councils in large organizations (Wilkinson, 1994; Kleingartner & Peng, 1991; Sharma & Sephton, 1991; Frenkel et al., 1993; Chen et al., 2003). The success of these repressive labor practices was ensured by local KMT cadres and the prominence and leadership of enterprise management in union affairs (Deyo, 1987). The law complemented this mechanism by making the statesponsored CFL a singular labor confederation enjoying the exclusive representation of local unions and workers. Therefore, trade unions had long been merely auxiliary institutions and administrative arms of the government at least until 1987 (Chen et al., 2003; Lee, 1995), and craft and enterprise unions together have dominated the Taiwanese union movement. Craft unions are much larger than enterprise unions, both in terms of the number of local unions and members for the following reasons (San, 1993). First, craft union membership is very
Enterprise Unionism in Asia
193
inclusive; it can be joined by workers excluded from full union representation rights under the Labor Union Law, by the self-employed, and by those from the non-excluded categories who work for small companies. Second, the huge workforce in small companies tends to join craft unions, as it is quite difficult for workers in small companies to establish enterprise unions.19 Lastly, craft unions are in a monopoly position vis-a-vis labor insurance coverage on the amended Labor Insurance Law in 1979, and the state even subsidizes the union dues of workers joining craft unions to encourage labor insurance coverage (Frenkel et al., 1993; Kuruvilla & Erickson, 2002). However, it must be added that the fundamental reason for the craft union dominance is the state’s positive attitude toward these unions; after all, the ‘‘public health insurance agents’’ have been little threat to the state. After the lifting of martial law in 1987, the independent union movement launched with two major strategies: reforming the ‘‘yellow unions’’ and forming new unions. Importantly, most of these unions were enterprisecentered organizations that sprung up in response to a specific incident or set of grievances, and many of them dissolved after the specific event or grievance had been resolved (Kleingartner & Peng, 1991). This indicates that the consistent labor controls through continuous martial law and close manipulation have successfully balkanized the union movement in Taiwan. The Democratic Progressive Party (DPP), the opposition party, took over the Presidency of Taiwan in 2000, ending more than 50 years rule of the KMT government. Ironically, the situation is not any different now with the change of government, however. The new DPP government has a new union ally, the Taiwanese Confederation of Trade Union (TCTU), which was also formed in 2000. The silent and evasive position of the TCTU on the passing of a number of regressive labor policies and law amendments by the new government aroused skepticism about whether the TCTU would challenge the new government on labor issues. Ironically again, the CFL, which in the past was criticized as KMT controlled ‘‘yellow unions,’’ is now reversing itself to attack the DPP government and the TCTU for betraying the rights of workers. Long-term division over unification and independence has not only dominated Taiwan politics, but it is also politicizing and dividing the labor movement. While the geographical federations of unions have divergent sectoral interests around which to unite and bargain, unions and workers within the same sector have difficulty in uniting into industrial unions and bargaining genuine sectoral agreement (Chen & Wong, 2002). Therefore, the union structure has not experienced any fundamental or dramatic change since the 1980s. Craft and enterprise unions still dominate
194
DAE YONG JEONG AND JOHN LAWLER
the union movement, though the share of craft unions has jumped both in number and membership, reflecting the advantages granted to craft unions in organizing. The share of enterprise unions has decreased from 47 percent in number (34 percent in membership) in 1987 to 29 percent in number (20 percent in membership) in 2002 (Taiwanese Council of Labor Affairs, 2001). Singapore. In Singapore, there was no direct collision between the state and the horizontal labor movement at the onset of the collective bargaining era. Rather, the actual conflict and struggle was between the left-wing faction and the Lee Kuan Yew faction (English-educated intellectuals) within the People’s Action Party (PAP), which had been established to achieve independence from the British colonial rule in 1954. Differences in political ideology led to a split within the party, resulting in the left-wing faction breaking away to form the Socialist Party and the Singapore Association of Trade Unions (SATU) in 1961 (Yuen & Lim, 2000). The struggle continued until 1963, when the PAP defeated the Socialist Party in the general elections. The PAP government dissolved the leftist grassroots organizations, accusing them of being communist groups (Tremewan, 1994). The leftists thus had been rendered politically impotent before Singapore left Malaysia in 1965 to become an independent sovereign state (Leggett, 1993a). In the relative vacuum on the left, the PAP government reorganized unions under the NTUC (National Trade Unions Congress), the exact counterpart of the CFL in Taiwan, and financially sponsored it to consolidate its power and mobilize workers against external threats – Indonesia and Malaysia (1960s), the Soviet Union (1960s–1970s), and China (Wilkinson, 1994; Neher, 2002; Tamney, 1996). The NTUC affiliated various types of unions; that is, there were 11 industrial, 12 craft/occupational, 3 general, and 15 enterprise unions in 1967 (Leggett, 1993a, p. 225). The PAP government tightly controlled local unions by nominating or appointing ruling party members as union leaders and by its guidance of affiliate union activities through the NTUC Research Unit, which played an advisory role in the formation of union bargaining claims and strategies. The PAP-NTUC symbiosis became so complete during the 1970s that it became increasingly difficult to distinguish between the political and labor elite. The PAP government, through the special PAP-NTUC symbiotic relationship and labor legislation, determined not just the regulation of unions but even the purpose of trade unionism itself (Leggett, 1993a; Wilkinson, 1994; Fong, 1981). The role of the PAP government was strengthened with Western aid to prevent communism and overseas investment during the Vietnam War and the 1980s (Neher, 2002).
Enterprise Unionism in Asia
195
Interestingly, however, Singaporean labor legislation imposed little structural restrictions on labor unions, unlike its Taiwanese counterpart. The major reason for this appears to be the PAP government’s stronger confidence in handling ‘‘labor problems.’’ On the one hand, the city-state’s small size is a great advantage in labor control. Living in fewer than 225 square miles (smaller than the Jakarta metropolitan area), Singaporeans have little room for non-conformity (Neher, 2002). On the other, the successful elimination of left-wing elements from the island before its independence in 1965 might have provided the PAP government with great confidence, in contrast to the KMT government that had a painful experience in losing the mainland in which the organized labor played a crucial part. Consequently, union structure did not change much in the 1970s. In 1979, for example, the 59 registered labor unions were organized by various lines; that is, there were 12 industrial, 22 craft/occupational, 4 general, and 21 enterprise unions, with membership concentrated in industrial and general unions (Leggett, 1993a, p. 227). In the early 1980s, however, the PAP government reorganized the union structure into what would better serve its current political and economic ends. The first restructuring was toward industrial unions, inspired by the West German model (Blum & Pataranapicht, 1987). By 1982, the NTUC had created nine industrial unions by breaking up the general unions, the Singapore Industrial Labor Organization (SILO) and the Pioneer Industries Employees’ Union (PIEU), which were the largest unions in the private sector (Kuruvilla & Erickson, 2002). Economically, this was to align workers more closely with industries for higher productivity. Politically, however, this was to prevent a recurrence of the concentration of individual power within the NTUC that Phey Yew Kok had wielded. Accordingly, the NTUC unions consisted of 18 industrial, 20 craft/occupational, 4 general, and 28 enterprise unions in 1984; composing 42, 10, 18, and 20 percent of NTUC membership, respectively (Leggett, 1993a, p. 230). The state labor policy swung again toward the establishment of enterprise unions, when it became clear that national- and industry-level bargaining posed a threat to firms whose different economic circumstances mandated different wage levels and increases. New unions were mandated to organize by enterprise line in 1984 (Kuruvilla & Erickson, 2002). Some enterprise unions were also created out of existing industrial unions, despite resistance of some union officials and members (Blum & Pataranapicht, 1987; Leggett, 1993b). The proportion of enterprise unions slightly increased from 40 percent in 1984 to 45 percent in the early 1990s (Frenkel, 1993, p. 311), reflecting the mild change in state labor policy.
196
DAE YONG JEONG AND JOHN LAWLER
With the increasingly important role trade unions play in society, there has recently been a shift in the relationship between the PAP and the labor movement. In the past, the labor movement was apparently the subservient party in the symbiotic relationship. With globalization and the regional economic crisis, trade unions have reached out to the public and grown closer to the population. Recognizing the unions’ increasing influence, the government has worked closely with the labor movement in communicating and marshaling support for tough economic policies. In the process, the NTUC has become a partner (Yuen & Lim, 2000). However, Singapore’s industrial relations system has not experienced rapid or fundamental change (Kuruvilla & Erickson, 2002), neither has Singapore’s union structure since the early 1990s. As of December 2005, the NTUC unions consisted of 22 industrial, 14 craft/occupational, 3 general, and 29 enterprise unions (Singaporean Ministry of Manpower, 2005). It is unlikely that the enterprise union would dominate in Singapore in the near future, as the state would not adopt overly extreme measures to that end in order not to jeopardize the special symbiotic relationship with the existing unions. In this sense, the Singaporean experience indicates a sharp contrast to the Indonesian experience in which the state dramatically turned its labor policy from authoritarian centralization to decentralization when it could not any longer sustain centralized control under both internal and external pressures.
SUMMARY AND CONCLUDING REMARKS This study proposed a new theoretical framework to explain enterprise unionism and conducted the first systematic comparative study of union structure in Asian countries. The proposed framework applies the idea of ‘‘critical junctures’’ and their legacies to the evolution of union systems in Asian countries, focusing on the initial period of the collective bargaining era. It maintains that the initial period of the collective bargaining era constitutes a critical juncture (state labor policy) that occurs in distinctive ways in different countries, and that these differences played a central role in shaping the different union structures in the following decades, amid the sharp difference in the desirable union structure between unionist and employer. In general, Asian countries had late developer advantages for state autonomy and power; and the Cold War further strengthened the autonomy, power, and legitimacy of the states for their repressive labor policies. However, the conditions that allowed state autonomy, power, and legitimacy for
Enterprise Unionism in Asia
197
their policies have been uneven across countries and over time even within one country. Some of the states have been successful in implementing their labor policies, whereas others have either miscalculated the situations handed to them, or have been unable to push their policies to their end at some points, resulting in union systems that are not necessarily in line with their intentions. The internal and external conditions that had major influences on the different state policies were analyzed. This study found a concrete enterprise union system in Japan (>90 percent), Thailand (90 percent), and Malaysia (80 percent), where the state has consistently implemented an enterprise union system; and an unstable enterprise union system in Korea (>90 percent, but rapidly declining) and the Philippines (71 percent), where the state has implemented an enterprise union policy, with some interruptions. On the other hand, this study found enterprise unions to be insignificant in Hong Kong (18 percent), where the state labor policy has been laissez-faire; and no enterprise unions until 1994 in Indonesia, where the state had implemented a centralization policy. In the middle-ground of this spectrum are Taiwan (29 percent), where the state has pursued a dual union system of craft unions (which are the majority) and enterprise unions; and Singapore (43 percent), where the state policy has gradually swung from laissez-faire toward an enterprise union system since 1984. The state labor policies and the share of enterprise unions are summarized in Table 1 below. On the other hand, empirical evidence found in this study is clearly contrary to both the cultural hypothesis and the ILM hypothesis. First, in none of the countries had ILMs existed prior to the collective bargaining era and the development of an enterprise union system. The ILM hypothesis is thus seriously disconfirmed. Second, in all these countries, most unions were organized by craft, industry line, or geographically across companies, despite state suppression and management opposition, throughout the precollective bargaining era. This fact does not support the cultural hypothesis, as we are forced to assume that earlier industrial workers held stronger Table 1. State Policy
A Typology of State Labor Policies and the Proportion of Enterprise Unions. Enterprise Unionism Consistent
More developed Japan (>90%) Less developed Malaysia (80%) Thailand (90%)
Inconsistent
Middle-Ground
Non-Enterprise Unionism
Dual unionism Gradual transition Centralization
Korea (>90%k) Taiwan (29%) Philippines (71%)
Singapore (43%)
Laissez-faire HK (18%)
Indonesia(0%)
198
DAE YONG JEONG AND JOHN LAWLER
‘‘Confucian ideas’’ than their counterparts in the collective bargaining era in Asia (if they indeed do). Third, if the Asian workers embraced the unique nature of ‘‘collectivism,’’ which supposedly led them to organize themselves by enterprise in the collective bargaining era, what necessity would be there for their states to forcefully enforce them to do so even by law? Fourth and finally, neither the cultural hypothesis nor the ILM hypothesis is able to explain the variations in the union system across the Asian countries (or the changing patterns of union structure over time even within a single country). For example, do Thailand and Malaysia have more sophisticated ILMs than Hong Kong and/or Indonesia? Do Thai and Malaysian workers hold stronger ‘‘collectivism’’ (which stemmed from Confucian ideas) than their counterparts in Hong Kong and Indonesia? Do workers in Hong Kong and Indonesia not share the ‘‘Asian culture’’ with their counterparts in the other Asian countries? This study investigated the evolution of unionism in Asian countries based on a new theoretical framework of enterprise unionism. We believe that the proposed framework significantly enhances our understanding of the Asian experiences and hope that it also aids in understanding similar phenomena in other countries. In fact, enterprise unionism is not a unique feature of Asian countries. Enterprise unions are also the dominant union structure in some Latin American countries, such as Chile and Columbia (Cook, 1998, pp. 318–319). Future research should further test the validity of the proposed framework with Latin American countries. Comparative studies of Asian and Latin American cases will further our confidence in the rejection of the alternative hypotheses.
NOTES 1. A mixed (or vertical) union of white-collar and blue-collar employees is empirically exceptional from a comparative perspective. Even in Japan, only about 40 percent of enterprise unions fall into this category (Japanese Ministry of Labor, 1979; Sakurabayashi, 1986). 2. This may explain the more severe opposition of larger employers to centralization in union structure than small and highly competitive employers in localized product and labor markets that may consent to centralized bargaining for fear of being ‘‘whipsawed’’ (see Hendricks & Kahn, 1982; Beaumount, Thomson, & Gregory, 1980). 3. For this study, we modified Fig. 1.1 in Collier and Collier (1991, p. 30). 4. Dunlop’s systems model emphasizes the importance of interaction between the actors whose behaviors are tightly constrained by the three major contexts (technology of the workplace, market or budgetary constraints on firm or industry, and
Enterprise Unionism in Asia
199
the locus and distribution of power in the larger society) and the resulting ‘‘web of rules.’’ However, it just stops there without providing any explanations of how the ‘‘locus and distribution of power’’ shapes the system in realty. We try to overcome this weakness and provide explanations of interactions among the actors through the actual ‘‘locus and distribution of power,’’ which shape union structure. We do not accept the ‘‘functionalist’’ assumption that a common ideology exists among actors as a necessary condition for creation of a stable system, either. Rather, our model sees the direction of causation (if any) the other way around. That is, a common ideology may be developed in the process of ‘‘reproduction’’ of the system. Neither do we accept the deterministic view of the ‘‘technological convergence hypothesis’’ that predicts the development of common forms of work organization and rules across countries. Instead, we take the strategic choice theory suggested by contemporary scholars (e.g., Kochan, McKersie, & Cappelli, 1984; Kochan, Katz, & McKersie, 1986 Kochan, et al. 1984, 1986) more seriously. In particular, our model considers the state’s strategic choices beyond the constraints given to them to be most significant in shaping union structure. 5. The Meiji government, which needed a mass of laborers for manufacturing factories for rapid industrialization, completely abolished the prestigious Tokugawaera guilds of artisans and granted freedom of occupation by the early 1870s as part of the Meiji Restoration (Yokoyama, 1899, p. 314). The early industrial workers were frustrated in the sudden changes brought about by the developmental state, with no prepared means of protecting their own status in society. They became too helpless to regulate apprenticeship and the overall labor market in contrast to their Western counterparts. As Max Weber (1927) observed, none of the Western states attempted to destroy guilds, though some of them endeavored to attract factories into their countries by granting privileges, which would be a protection against the guilds in the process of industrialization. 6. The concept of micro-corporatism refers to a situation where consent forged at the point of production plays a determinant role in maintaining the accord between labor and capital even without the global compromise in the state (Lee, 1996, 1998a). 7. Japanese Ministry of Labor stopped to report the proportions of different types of unions in Rodo Kumiai Kiso Chosa (Basic Survey on Labor Unions) from 2000. 8. The only exceptions were the teachers’ unions, which had members in both government and non-government schools, and the Amalgamated National Union of Local Authorities whose members could not form enterprise unions because of the very small size of their establishment (Arudsothy & Litter, 1993). 9. As of the end of 2001, there were 417 relatively large unions with 500 or more members. These unions accounted for only 6.8 percent of the total number of unions but 73.5 percent of the total membership. This centralization of membership in these large enterprise unions tends to encourage them to use their organizational power and financial capability to seek to create an enterprise-based empire and to focus on ego-centrism rather than the common good (Bae & Cho, 2004, pp. 149–151). See Park (1990) for the gooey relationship between enterprises and their enterprise unions. Today, the number of workers in non-standard employment exceeds that of regular workers. The proportion of temporary and daily workers in the total employed increased from 45.9 percent in 1997 to 52.9 percent in 2000 (Lee, 2002; Kim & Kim, 2003; Kim, 2005).
200
DAE YONG JEONG AND JOHN LAWLER
10. The Labor Law of 1963 indeed did not mandate unions to form by industrial lines. The reorganization of unions had been first forced by the state power and the structure was then maintained by the Labor Law of 1963 that prohibited multiple unions. Importantly, most of the unions were ‘‘amorphous industrial unions,’’ based on the old system of enterprise unions, which maintained their functions but simply changed names (Kim, 1999b) and not individual workers, but enterprise unions were the members of industrial unions (Song, 1989). Thus, though it lasted almost a decade, the legacy of ‘‘industrial unionism’’ on the later union movement was limited. 11. See Kim (1995) and Shin (1994) for Korean workers’ consciousness. 12. Social movement unions are defined by their commitment to an agenda of social change beyond the workplace and by their extensive links and alliances with other social and political movements (Hutchison, 2001, p. 75). 13. In order to gain union recognition, according to the Manpower Ministerial Relation No. Per-03/MEN/1993, a labor union must be able to meet the following conditions: (1) union officers in at least five of the country’s 27 provinces; (2) branch offices in 25 districts; (3) 100 plant-level units; and (4) 10,000 members if it is so geographically confined that it would be unable to meet standards (1)–(3) because of the nature of the industry, for example, in mining. Thus, only the SPSI was able to meet the strict requirements for official recognition (Sulistyaningsih, 2001, pp. 1–5). 14. This information was obtained from Endang Sulistyaningsih, Head of Research and Development Bureau at Indonesian Ministry of Manpower and Transmigration in January 2005. 15. This information was obtained from Muzni Tambusai, Director General of Industrial Relations at the Ministry of Manpower and Transmigration in January 2005. 16. This information was obtained from Muzni Tambusai in January 2005. 17. See England and Rear (1975, pp. 5–9) for the disturbances of 1966 and 1967. 18. There are two types of unions defined under Taiwanese law: industrial and craft (Taiwanese Council of Labor Affairs of the Executive Yuan, 1999, p. 7). According to Chen et al. (2003), ‘‘industrial union,’’ which is a literal translation from Chinese, is more accurately termed ‘‘enterprise union’’ in English in most circumstances (p. 320). Thus, the ‘‘industrial unions’’ in Taiwan should be considered as enterprise unions for a cross-national analysis. As such, the term, enterprise union, is used instead of industrial union in this paper. 19. In 1995, companies that had less than 30 employees employed some 65 percent of workforce in the private sector and companies that had more than 500 employees employed only five and half percent of the workforce (Chen et al., 2003).
ACKNOWLEDGMENTS We would like to thank the scholars and government officials in various countries, who directly or indirectly helped us locate the relevant data: Shyh-jer Chen, Sununta Siengthai, Rajah Rasiah, Vedi Hadiz, Ending Sulistyaningsih, Wee Kwek Ng, Zuraini Bt Abu Kassim, Sarosh Kuruvilla,
Enterprise Unionism in Asia
201
Anil Verma, Stephen Chiu, Michele Ford, and David Wan Tai Wai. We would also like to thank John Price, Ruth Aguilera, Michael LeRoy, Edward Hertenstein, and Helena Worthen for their valuable feedback on the earlier drafts of this manuscript. We would also like to thank the editors, David Lewin and Bruce E. Kaufman, and an anonymous referee for putting us through a stimulating reviewing process, which has dramatically improved the quality of this article.
REFERENCES In English Abegglen, J. (1958). The Japanese factory: Aspects of its social organization. Glencoe, IL: The Free Press. Adams, R. (1995). Industrial relations under liberal democracy: North America in comparative perspective. Columbia, SC: University of South Carolina. Amante, M. S. V. (2004). Industrial democracy in the rough seas: The case of Philippine seafarers. IRRA 56th Proceedings, pp. 81–89. Amsden, A. H. (1997). South Korea: Enterprising groups and entrepreneurial government. In: A. Chandler, F. Amatori & T. Hikino (Eds), Big business and the wealth of nations. Cambridge: Cambridge University Press. Aoki, M. (1997). Unintended fit: Organizational evolution and government design of institutions in Japan. In: M. Aoki, K. Kim & M. Okuno-Fujiwara (Eds), The role of government in East Asian economic development: Comparative institutional analysis. Oxford: Clarendon Press. Aoki, M., Murdock, K., & Djuno-Fujwara, M. (1997). Beyond the East Asian miracle: Introducing the market-enhancing view. In: M. Aoki & K. Kim (Eds), The role of government in East Asian economic development: Comparative institutional analysis. Oxford: Clarendon Press. Arudothy, P., & Litter, C. (1993). State regulation and union fragmentation in Malaysia. In: S. Frenkel (Ed.), Organized labor in the Asia-Pacific region: A comparative study of trade unionism in the countries. New York: ILR Press. Aspalter, C. (Ed.). (2001). The Taiwanese economic miracle: From sugarcane to hightechnology. In: Understanding modern Taiwan: Essays in economics, politics and social policy. Hampshire, UK: Ashgate Publishing Limited. Ayadurai, D. (1993). Malaysia. In: S. Deery & R. Mitchell (Eds), Labor law and industrial relations in Asia. Melbourne, Australia: Longman Cheshire. Bae, K., & Cho, S. (2004). Labor movement. In: Korea Labor Institute (Ed.), Labor in Korea. Seoul: Korea Labor Institute. Beaumount, P., Thomson, A. W. J., & Gregory, M. B. (1980). Bargaining structure. Management Decision, 18(3), 103–170. Bello, W. (2000). The Philippines: The making of a neo-classical tragedy. In: R. Robinson, M. Beeson, K. Jayasuriya & H. Kim (Eds), Politics and markets in the wake of the Asian crisis. London: Routledge.
202
DAE YONG JEONG AND JOHN LAWLER
Benson, J., & Debroux, P. (2000). Japanese trade unions at the crossroads: Dilemmas and opportunities created by globalization. In: C. Rowley & J. Benson (Eds), Globalization and labor in the Asia Pacific region. London: Frank Cass. Berghahn, V., & Karsten, D. (1988). Industrial relations in West Germany. New York: St. Martin’s Press. Bhopal, M. (2001). Malaysian unions in political crisis: Assessing the impact of the Asian contagion. Asian Pacific Business Review, 13(45), 73–100. Bhopal, M., & Todd, P. (2000). Multinational corporations and trade union development in Malaysia. In: C. Rowley & J. Benson (Eds), Globalization and labor in the Asian Pacific region. London: Frank Cass. Black, B. (2001). National culture and industrial relations and pay structures. Labor, 15(2), 257–277. Blum, A., & Pataranapicht, S. (1987). Productivity and the path to house unionism: Structural change in the Singapore labor movement. British Journal of Industrial Relations, 25(3), 389–400. Botz, D. L. (2001). Made in Indonesia: Indonesian workers since Suharto. Cambridge, MA: South End Press. Bourchier, D. (Ed.). (1994). Solidarity: The new order’s first free trade union. In: Indonesia’s emerging proletariat: Workers and their struggles. Clayton, Australia: The Center of Southeast Asian Studies Monash University. Broadbent, K. (2001). Power in the union?: Part-time workers and enterprise unionism in Japan. International Journal of Manpower, 22(4), 318–332. Brown, A. (2001). After the Kader fire: Labor organizing for health and safety standards in Thailand. In: J. Hutchison & A. Brown (Eds), Organizing labor in globalizing Asia. New York: Routledge. Brown, A. (2004). Labor, politics and the state in industrializing Thailand. London: Routledge Curzon. Brown, A., & Frenkel, S. (1993). Union unevenness and insecurity in Thailand. In: S. Frenkel (Ed.), Organized labor in the Asia-Pacific region: A comparative study of trade unionism in nine countries. Ithaca, NY: ILR Press. Burtt, E. J. (1979). Labor in the American economy. New York: St. Martins’ Press. Card, D., & Freeman, R. (1993). Small differences that matter: Labor markets and income maintenance in Canada and the United States. Chicago: University of Chicago. Chan, A. (1988). The trade unions ordinance: Its enactment and subsequent development. In: Y. C. Jao, D. A. Levin, S. H. Ng & E. Sinn (Eds), Labor movement in a changing societies: The experience of Hong Kong. Hong Kong: University of Hong Kong. Cheetham, J. (1988). The perspective of employers on trade unions. In: Y. C. Jao, D. A. Levin, S. H. Ng & E. Sinn (Eds), Labor movement in a changing societies: The experience of Hong Kong. Hong Kong: University of Hong Kong. Chen, S., Ko, J., & Lawler, J. (2003). Changing patterns of industrial relations in Taiwan. Industrial Relations, 42(3), 315–340. Chen, Y., & Wong, M. (2002). New bondage and old resistance: Realities and challenges of the labor movement in Taiwan. Hong Kong: D C Output Production Center. Cheng, W. (2001). Labor politics in transition in democratizing Taiwan. In: C. Aspalter (Ed.), Understanding modern Taiwan: Essays in economics, politics and social policy. Hampshire, UK: Ashgate Publishing Limited.
Enterprise Unionism in Asia
203
Clarke, R. (1993). Conclusions: Towards a synthesis of international and comparative experience of nine countries. In: G. Bamber & R. Lansbury (Eds), International and comparative industrial relations. St. Leonards, Australia: Allen & Unwin. Clegg, H. (1976). Trade unionism under collective bargaining: A theory based on comparisons of six countries. Oxford: Basil Blackwell. Cohen, Y. (1994). Radicals, reformers, and reactionaries: The prisoner’s dilemma and the collapse of democracy in Latin America. Chicago: University of Chicago Press. Cole, G. D. H. (1953). An introduction to trade unionism. London: George Allen & Unwin Ltd. Cole, R. (1979). Work, mobility, and participation: A comparative study of American and Japanese industry. Berkeley: University of California Press. Collier, R., & Collier, D. (1991). Shaping the political arena: Critical junctures, the labor movement, and regime dynamics in Latin America. Princeton, NJ: Princeton University Press. Cook, L. (1998). Toward flexible industrial relations? Neo-liberalism, democracy, and labor reform in Latin America. Industrial Relations, 37(3), 311–336. Dean, A. B., & Guevara, P. J. (2001). New economic challenge to Japan’s trade union movement. WorkingUSA, 4(4), 59–83. Defelice, G. (1980). Comparison misconceived. Comparative Politics, 13, 119–124. Deyo, F. (Ed.). (1987). State and labor: Modes of political exclusion in East Asian development. In: The political economy of the new Asian industrialism. Ithaca, NY: Cornell University Press. Deyo, F. (1989). Beneath the miracle: Labor subordination in the new Asian industrialization. Berkeley, CA: University of California Press. Deyo, F. (1997). Labor and industrial restructuring in South-East Asia. In: G. Rodon, K. Hewison & R. Robison (Eds), The political economy of South-East Asia: An introduction. Melbourne: Oxford University Press. Dohse, K., Jiirgens, U., & Malsch, T. (1985). From ‘Fordism’ to ‘Toyotism’? The social organization of the labor process in the Japanese automobile industry. Politics and Society, 14(2), 115–146. Dore, R. (1979). Industrial relations in Japan and elsewhere. In: A. M. Craig (Ed.), Japan: A comparative view. Princeton: Princeton University Press. Dulles, F. R., & Dubofsky, M. (1993). Labor in America. Wheeling, IL: Harlan Davidson, INC. Dunlop, J. (1958). Industrial relations system. New York: Henry Holt. Edwards, C. (1996). Labor regulations and industrial relations in Indonesia. Policy Research Working Paper 1640. The World Bank. Edwards, P., et al. (1998). Great Britain: From partial collectivism to neo-liberalism to where? In: A. Ferner & R. Hyman (Eds), Changing industrial relations in Europe. Oxford, UK: Blackwell. England, J. (1980). Strategic factors in trade unionism and industrial disputes in Hong Kong. Labor and Society, 5(3), 267–290. England, J. (1989). Industrial relations and law in Hong Kong. New York: Oxford University Press. England, J., & Rear, J. (1975). Chinese labour under British rule. Hong Kong: Oxford University Press. England, J., & Rear, J. (1981). Industrial relations and law in Hong Kong. New York: Oxford University Press. Evans, P. (1987). Class, state, and dependence in East Asia. In: F. Deyo (Ed.), The political economy of the new Asian industrialism. Ithaca, NY: Cornell University Press.
204
DAE YONG JEONG AND JOHN LAWLER
Evans, P., Reuschemeyer, D., & Skocpol, T. (Eds). (1985). On the road toward a more adequate understanding of the state. In: Bringing the state back in. New York: Cambridge University Press. Fineman, D. M. (1997). A special relationship: The United States and military government in Thailand, 1947–1958. Honolulu: University of Hawaii Press. Fong, P. E. (1981). Singapore. In: A. Blum (Ed.), International handbook of industrial relations. Westport: Greenwood Press. Ford, M. (1999). Testing the limits of corporatism: Industrial relations in Suharto’s Indonesia. The Journal of Industrial Relations, 41(3), 372–392. Ford, M. (2000). Research note: Indonesian trade union developments since the fall of Suharto. Labor and Management in Development, 1(3), 2–10. Fosh, P., Chan, A., Chow, W., Snape, Ed., & Westwood, R. (Eds). (1999). Government supervision of trade unions in Hong Kong: Colonial powers, patterns of enforcement and prospects for change. In: Hong Kong management and labor: Change and continuity. New York: Routledge. Fox, A. (1974). Beyond contract: Work, power, and trust relations. London: Faber & Faber Limited. Fox, A. (1985). History and heritage: The social origins of the British industrial relations system. London: Allen & Unwin. Frenkel, S. (Ed.). (1993). Variations in patterns of trade unionism: A synthesis. In: Organized labor in the Asia-Pacific region: A comparative study of trade unionism in nine countries. Ithaca, NY: ILR Press. Frenkel, S., Hong, J., & Lee, B. (1993). The resurgence and fragility of trade unions in Taiwan. In: S. Frenkel (Ed.), Organized labor in the Asia-Pacific region: A comparative study of trade unionism in nine countries. Ithaca, NY: ILR Press. Frenkel, S., & Kuruvilla, S. (2002). Logics of action, globalization, and changing employment relations in China, India, Malaysia and the Philippines. Industrial & Labor Relations Review, 55(3), 387–413. Frenkel, S., & Peetz, D. (1998). Globalization and industrial relations in East Asia: A threecountry comparison. Industrial Relations, 37(3), 282–310. Galenson, W. (1959). Labor and economic development. New York: Wiley. Galenson, W. (1992). Labor and economic growth in five Asian countries: South Korea, Malaysia, Taiwan, Thailand, and the Philippines. New York: Praeger. Gall, G. (1998). The development of the Indonesian labor movement. The International Journal of Human Resource Management, 9(2), 359–376. Garon, S. (1987). The state and labor in modern Japan. Berkeley, CA: University of California Press. Gerschenkron, A. (1962). Economic backwardness in historical perspective. Cambridge, MA: Harvard University Press. Gordon, A. (1988). The evolution of labor relations in Japan: Heavy industry, 1853–1955. Cambridge, MA: Harvard University Press. Gordon, A. (1998). The wages of affluence: Labor and management in postwar Japan. Cambridge, MA: Harvard University Press. Hadiz, V. (2000). Globalization, labor and the state: The case of Indonesia. In: C. Rowley & J. Benson (Eds), Globalization and labor in the Asian Pacific region. London: Frank Cass. Hadiz, V. (2002). The Indonesian labor movement: Resurgent or constrained? Southeast Asian Affairs 2002, 130–142.
Enterprise Unionism in Asia
205
Hadiz, V. R. (2001). New organizing vehicles in Indonesia. In: J. Hutchison & A. Brown (Eds), Organizing labor in globalizing Asia. New York: Routledge. Haggard, S., & Cheng, T. (1987). State and capital in the East Asian NICs. In: F. Deyo (Ed.), The political economy of the new Asian industrialism. Ithaca, NY: Cornell University Press. Hanartani, M. (2005). Decent work program in Indonesia. Report submitted to 2nd Sub Regional Tripartite Forum on Decent Work in South East Asia and the Pacific. Melbourne, Australia, 5–8 April 2005. Hawkins, E. (1962). Indonesia. In: W. Galenson (Ed.), Labor in developing economies. Westport, CT: Greenwood Press. Hendricks, W., & Kahn, L. (1982). Determinants of bargaining structure in U.S. manufacturing industries. Industrial and Labor Relations Review, 35, 181–195. Henry, J. (1983). Corporate strategy and employment relations in multinational corporations: Some evidence from Kenya and Malaysia. In: K. Thurley & S. Wood (Eds), Industrial relations and management strategy. Cambridge: Cambridge University Press. Herman, E., Kuhn, A., & Seeber, R. (1987). Collective bargaining and labor relations. Englewood, NJ: Prentice-Hall. Hong Kong Department of Labor. (2004). www.labor.gov.hk/eng/public/rtu/report2003 Hutchison, J. (2001). Export opportunities: Unions in the Philippine garments industry. In: J. Hutchison & A. Brown (Eds), Organizing labor in globalizing Asia. New York: Routledge. Hutchison, J., & Brown, A. (Eds). (2001). Organizing labor in globalizing Asia: An introduction. In: Organizing labor in globalizing Asia. New York: Routledge. ILO. (1997). World labor report 1997–98: Industrial relations, democracy and social stability. Geneva: International Labor Organization. Indonesian Ministry of Manpower and Transmigrations (2005). Information from Muzni Tambusai, Director General of Industrial Relations at the Ministry of Manpower and Transmigration. Jacobi, O., Keller, B., & Muller-Jentsch, W. (1998). Germany: Facing new challenges. In: A. Ferner & R. Hyman (Eds), Changing industrial relations in Europe. Oxford, UK: Blackwell. Japan Institute for Labor Policy and Training. (2004). Labor situation in Japan and analysis 2004/2005. Jeong, J. (1995). Enterprise unionism from a Korean perspective. Economic and Industrial Democracy, 16, 253–273. Jimenz, R. (1993). The Philippines. In: S. Deery & R. Mitchell (Eds), Labor law and industrial relations in Asia. Melbourne, Australia: Longman Cheshire. Johnson, C. (1995). Japan: Who governs: The rise of the development state. W.W. Norton & Company: New York. Jomo, K. S., & Todd, P. (1994). Trade unions and the state in peninsular Malaysia. Kuala Lumpur: Oxford University Press. Kahn, H. (1979). World economic development: 1979 and Beyond. Boulder, Colo.: Westview Press. Kang, D. (2002). Crony capitalism: Corruption and development in South Korea and the Philippines. Cambridge: Cambridge University. Katoppo, A. (2000). The role of community groups in the environment movement. In: C. Manning & P. V. Diermen (Eds), Indonesia in transition. London: ZED Books. Kawanish, H. (1992). Enterprise unionism in Japan. New York: Chapman and Hall. Kerr, C., Dunlop, J., Harbison, F., & Myers, C. (1960). Industrialism and industrial man. Cambridge, MA: Harvard University Press.
206
DAE YONG JEONG AND JOHN LAWLER
Kim, D., & Kim, S. (2003). Globalization, financial crisis, and industrial relations. Industrial Relations, 42(3), 341–367. Kim, H. (1993). The Korean union movement in transition. In: S. Frenkel (Ed.), Organized labor in the Asia-Pacific region: A comparative study of trade unionism in nine countries. Ithaca, NY: ILR Press. Kim, Y. H., & Kim, N. J. (1970). Hangook nodong Undongshi. Seoul: Iljogak. Kingsbury, D. (1998). The politics of Indonesia. New York: Oxford University Press. Kleingartner, A., & Peng, H. (1991). Taiwan: An exploration of labor relations in transition. British Journal of Industrial Relations, 29(3), 427–445. Kochan, T., & Katz, H. (1988). Collective bargaining and industrial relations. Homewood, IL: Irwin. Kochan, T., Katz, H., & McKersie, R. (1986). The transformation of American industrial relations. New York: Basic Books. Kochan, T., McKersie, R., & Cappelli, P. (1984). Strategic choice and industrial relations theory. Industrial Relations, 23(1), 16–39. Koike, K. (1987). Human resource development and labor-management relations. In: K. Yamamura & Y. Yasubara (Eds), The political economy of Japan, volume 1: The domestic transformation. Stanford: Stanford University Press. Koike, K. (1988). Understanding industrial relations in modern Japan. New York: St. Martin’s Press. Koo, H. (1987). East Asian development. In: F. Deyo (Ed.), The political economy of the new Asian industrialism. Ithaca, NY: Cornell University Press. Koo, H. (2000). The dilemmas of empowered labor in Korea. Asian Survey, 11(2), 697–715. Koo, H. (2001). Korean workers: The culture and politics of class formation. Ithaca, NY: Cornell University Press. Kumazawa, M. (1996). Portraits of the Japanese workplace. Boulder, CO: Westview Press. Kume, I., & Thelen, K. (1999). The rise of non-market training regimes: Germany and Japan compared. Journal of Japanese Studies, 25(1), 33–64. Kuruvilla, S. (1996). The relationship between economic development strategies and industrial relations: India, Malaysia, Singapore and the Philippines. Industrial and Labor Relations Review, 49(4), 635–757. Kuruvilla, S., & Arudsothy, P. (1995). Economic development strategy. In: A. Verma, T. Kochan, & R. Lansbury (Eds), Government labor policy and firm-level industrial relations practices in Malaysia. Employment relations in growing Asian economies. New York: Routledge. Kuruvilla, S., & Erickson, C. (2002). Change and transformation in Asian industrial relations. Industrial Relations, 41(2), 171–227. Laslett, J. H. M., & Lipset, S. M. (1974). Failure of a dream? Essays in the history of American socialism. Garden City, NY: Anchor. Lawler, J., & Suttawet, C. (2000). Labor unions, globalization and deregulation in Thailand. In: C. Rowley & J. Benson (Eds), Globalization and labor in the Asia Pacific region. London: Frank Cass. Lee, C. (1998b). New unionism and the transformation of the Korean industrial relations system. Economic and Industrial Democracy, 19, 347–373. Lee, J. (1995). Economic development and the evolution of industrial relations in Taiwan, 1950–1993. In: A. Verma, T. Kochan & R. Lansbury (Eds), Employment relations in the growing Asian economics. New York: Routledge. Lee, J. (1996). Micro-corporatist class compromise and democratic stability: The South Korean case. Ph.D. dissertation, University of Wisconsin at Madison.
Enterprise Unionism in Asia
207
Lee, J. (1998a). Micro-corporatism in South Korea: A comparative analysis of enterprise-level industrial relations. Economic and Industrial Democracy, 19, 443–474. Lee, J. (2002). Labor key to success amid intensified global competition. The Korea Herald, June 5. Lee, M. B. (1993). Korea. In: M. Rothman, D. Briscoe & R. Nacamulli (Eds), Industrial relations around the world: Labor relations for multinational companies. New York: Walter de Gruyter. Leggett, C. (1993a). Corporatist trade unionism in Singapore. In: S. Frenkel (Ed.), Organized labor in the Asian-Pacific region. Ithaca, New York: ILR Press. Leggett, C. (1993b). Labor law and industrial relations in Asia: Eight country studies. Melbourne, Australia: Longman Cheshire. LeRoy, M., & Johnson, J. (2001). Death by lethal injunction: National emergency strikes under the Taft-Hartley Act and the moribund right to strike. Arizona Law Review, 63(1), 63–134. Levin, D., & Chiu, C. (1993). Dependent capitalism, a colonial state, and marginal unions: The case of Hong Kong. In: S. Frenkel (Ed.), Organized labor in the Asian-Pacific region. Ithaca, NY: ILR Press. Levin, D., & Ng, S. (1995). From an industrial to a post-industrial economy: Challenges for human resource management in Hong Kong. In: A. Verma, T. Kochan & R. Lansbury (Eds), Employment relations in the growing Asian economics. New York: Routledge. Levin, M. (1997). Worker rights and labor standards in Asia’s four new tigers: A comparative perspective. New York: Plenum Press. Liphart, A. (1971). Comparative politics and the comparative method. American Political Science Review, 65(3), 682–695. Lipset, M., & Rokkan, S. (Eds). (1967). Cleavage structures, party systems, and voter alignments: An introduction. In: Party systems and voter alignments: Cross-national perspectives. New York: The Free Press. Lo¨sche, P. (1973). Stages in the evolution of the German labor movement. In: A. Sturmthal & J. G. Scoville (Eds), The international labor movement in transition. Urbana, IL: University of Illinois Press. Mabry, B. D., & Srisermbhok, K. (1985). Labor relations under Martial Law: The Thailand experience. Asian Survey, 25(6), 613–638. Malaysian Ministry of Human Resources. (2005). Information from Zuraini bt Abu Kassim, Assistant Director, Trade Union Affairs Department. Manning, C. (1993). Structural change and industrial relations during the Soeharto period: An approaching crisis? Bulletin of Indonesian Economic Studies, 29(2), 59–95. Manusphaibool, S. (1993). Thailand. In: S. Deery & R. Mitchell (Eds), Labor law & industrial relations in Asia. Melbourne, Australia: Frank Cass. Maurice, M., Sellier, F., & Silvestre, J. (1986). The social foundations of industrial power: A comparison of France and Germany. Cambridge, MA: MIT Press. Mill, J. S. (1843). A system of logic. New York: Harper & Brothers Publishers. Miners, N. (1988). The representation and participation of trade unions. In: Y. C. Jao, D. A. Levin, S. H. Ng & E. Sinn (Eds), Labor movement in a changing societies: The experience of Hong Kong. Hong Kong: University of Hong Kong. Morell, D., & Samudavanjja, C. (1981). Political conflict in Thailand: Reform, reaction, revolution. Cambridge, MA: Gunn & Hain. Nayar, R. (1996). Indonesian labor legislation in a comparative perspective: A study of six APEC countries. Policy Research Working Paper 1673. The World Bank. Neher, C. (2002). Southeast Asia in the new international era. Boulder, CO: Westview Press.
208
DAE YONG JEONG AND JOHN LAWLER
Nemenzo, F. (1988). From autocracy to elite democracy. In: A. J. Dios, P. Daroy & L. Kalaw Tirol (Eds), Dictatorship and revolution: Roots of people’s power. Manila: Conspectus. Ng, S., & Poon, C. Y. W. (2004). Business restructuring in Hong Kong: Strengths and limits of post-industrial capitalism. Oxford: Oxford University Press. Ng, S., & Rowley, C. (2000). Globalization and Hong Kong’s labor market: The deregulation paradox. In: C. Rowley & J. Benson (Eds), Globalization and labor in the Asia Pacific region. London: Frank Cass. Ng, S., & Sit, V. (1989). Labor relations and labor conditions in Hong Kong. London: MacMillan. Ng, S., Stewart, S., & Chan, F. (1997). Current issues of workplace relations & management in Hong Kong. Hong Kong: The University of Hong Kong. Nish, I., et al. (1996). Industrial harmony, the trade union movement and labor administration in Hong Kong. In: I. Nish, G. Redding & S. Ng (Eds), Work and society: Labor and human resources in East Asia. Hong Kong: Hong Kong University Press. Ofreneo, R. (1995). Philippine industrialization and industrial relations. In: A. Verma, T. Kochan & R. Lansbury (Eds), Employment relations in the growing Asian economics. New York: Routledge. Ohkawa, K., & Rosovsky, H. (1973). Japanese economic growth. Stanford, CA: Stanford University Press. Okochi, K., & Levine, S. (1973). Japanese industrial relations system. In: K. Okochi, B. Karsh & S. Levine (Eds), Workers and employment in Japan. Tokyo: Tokyo University Press. Olson, M. (1968). The logic of collective action: Public goods and the theory of groups. Cambridge, MA: Harvard University Press. Park, Y., & Lee, M. B. (1995). Economic development, globalization, and practices in industrial relations and human resource management in Korea. In: A. Verma, T. Kochan & R. Lansbury (Eds), Employment relations in the growing Asian economies. London: Routledge. Parsa, M. (2000). States, ideologies, & social revolutions: A comparative analysis of Iran, Nicaragua and the Philippines. Cambridge: Cambridge University. Pencavel, J. (2003). Company unions, wages, and hours. Advances in Industrial and Labor Relations, 12, 7–38. Philippine Bureau of Labor Relations. (2004). http://www.blr.dole.gov, October 4. Price, J. (1997). Japan works: Power and paradox in postwar industrial relations. Ithaca, NY: ILR Press. Przeworski, A., & Teune, H. (1970). The logic of comparative social inquiry. Malabar, FL: R.E. Krieger Pub. Co. Rahayu, S. K., & Sumarto, S. (2003). The practice of industrial relations in Indonesia. SMERU Working Paper. Jakarta, Indonesia. Ramos, E. T. (1990). Dualistic unionism and industrial relations. Quezon: New Day Publishers. Rasiah, R. (2001). Labor and work organization in Malaysia’s Proton. In: J. Hutchison & A. Brown (Eds), Organizing labor in globalizing Asia. New York: Routledge. Rinakit, S. (1999). Trade unions and labor unrest. In: R. Baker, J. Kristiadi, D. Ramage & M. Soesastro (Eds), Indonesia: The challenge of change. Singapore: Pasir Panjang. Robinson, R. (1996). The politics of ‘Asian values’. The Pacific Review, 9(3), 309–327. Robinson, R. (1997). Indonesia: Crisis, oligarchy, and reform. In: G. Rodan, K. Hewison & R. Robison (Eds), The political economy of South-East Asia: Conflicts, crises, and change. New York: Oxford University Press.
Enterprise Unionism in Asia
209
Sakoh, K. (1990). Economic implications of enterprise unionism. Journal of Labor Research, 11(3), 257–267. San, G. (1993). Taiwan. In: M. Rothman, et al. (Eds), Industrial relations around the world. New York: Walter de Gruyter. Scalapino, R. (1965). Labor and politics. In: L. William (Ed.), The state and economic enterprise in Japan. Princeton, NJ: Princeton University Press. Sharma, B. (1985). Aspects of industrial relations in ASEAN. Singapore: Institute of South East Asian Studies. Sharma, B., & Sephton, P. (1991). The determinants of union membership growth in Taiwan. Journal of Labor Research, 12(4), 429–437. Shimada, H., & Shirai, T. (1978). Japan. In: J. Dunlop & W. Galenson (Eds), Labor in the twentieth century. New York: Academic Press. Shirai, T. (Ed.). (1983). A theory of enterprise unionism. In: Contemporary industrial relations in Japan. Madison: University of Wisconsin Press. Siengthai, S. (1999). Industrial relations and the recession in Thailand. Report to the ILO/ EASTMAT. Asian Institute of Technology, Bangkok. Singaporean Ministry of Manpower. (2005). Information from Kwek Wee Ng, the Registrar of Trade Unions. Skocpol, T. (1985). Bringing the state back in: Strategies of analysis in current research. In: P. Evans, D. Reuschemeyer & T. Skocpol (Eds), Bringing the state back in. New York: Cambridge University Press. Song, H. K. (1989). State and the working class-labor market in South Korea: 1961–1987. Ph.D. dissertation, Harvard University. Sulistyaningsih, E. (2001). New Indonesia industrial relations system: Promoting freedom of association and the right to collective bargaining. Symposium. APEC HRD WG, Mexico City, June 25–26. Taft, P. (1964). Organized labor in American history. New York: Russel and Russel. Taira, K. (1970). Economic development and the labor market in Japan. New York: Columbia University Press. Taiwanese Council of Labor Affairs. (2001). Monthly bulletin of labor statistics. Taiwan Area, Republic of China. Taiwanese Council of Labor Affairs of the Executive Yuan. (1999). Labor laws and regulations of the Republic of China. Council of Labor Affairs of the Executive Yuan, R.O.C. Takaki, A. (2002). Re-examining the role of labor unions in the era of the diversified workforce. Japan Labor Bulletin, 41(5), 6–14. Tamney, J. (1996). The struggle over Singapore’s soul. New York: Walter de Gruyter & Co. Thai Department of Labor Protection and Social Welfare. (2005). Information from Sununta Siengthai. Tilly, C. (1978). From mobilization to revolution. Reading, MA: Addison-Wesley Publisher Co. Tilly, C. (1984). Big structures, large processes, huge comparisons. New York: Russell Sage Foundation. Tilly, C. (1985). War making and state making as organized crime. In: P. Evans, D. Reuschemeyer & T. Skocpol (Eds), Bringing the state back in. New York: Cambridge University Press. Tolliday, S., & Zietlin, J. (1985). Shopfloor bargaining and the state: Historical and comparative perspectives. Cambridge: Cambridge University Press. Tremewan, C. (1994). The political economy of social control in Singapore. New York: St. Martin’s Press, Inc.
210
DAE YONG JEONG AND JOHN LAWLER
Turner, H. A. (1980). The last colony: But whose? New York: Cambridge University Press. Turner, H. A., Fosh, P., & Ng, S. (1991). Between two societies: Hong Kong labor in transition. Hong Kong: University of Hong Kong. Vasil, R. (1997). Governing Indonesia: National development and democracy. Singapore: Kin Keong Printing Co. Villegas, E. (1988). The political economy of Philippine labor laws. Quezon City: Foundation for Nationalist Studies. Wad, P. (1988). The Japanization of the Malaysian trade union movement. In: R. Southall (Ed.), Trade unions and the new industrialization of the Third World. London: Zed Books. Wad, P., & Jomo, K. S. (1994). In-house unions: ‘Looking East’ for industrial relations. In: K. S. Jomo (Ed.), Japan and Malaysian development. London: Routledge. Webb, S., & Webb, B. (1911). Industrial democracy. New York: Longmans, Green and Co. Weber, A. (1967). Stability and change in the structure of collective bargaining. In: L. Ulman (Ed.), Changes to collective bargaining. Englewood Cliffs, NJ: Prientice-Hall. Weber, M. (1927). General economic history. Glencoe, IL: The Free Press. Wehmhorner, A. (1983). Trade unionism in Thailand: A new dimension in a modernized society. Journal of Contemporary Asia, 13(4). West, L. A. (1997). Militant labor in the Philippines. Philadelphia: Temple University Press. Wilkinson, B. (1994). Labor and industry in the Asia-Pacific: Lessons from the newly industrialized countries. New York: Walter de Gruyter. Woo-Cummings, M. (1997). The political economy of growth in East Asia: A perspective on the state, market, and ideology. In: M. Aoki, H. Kinn & M. Okuno-Fujiwara (Eds), The role of government in East Asian economic development: Comparative institutional analysis. Oxford: Clarendon Press. Wurfel, D. (1959). Trade union development and labor relations policy in the Philippines. Industrial and Labor Relations Review, 12, 562–608. Yanaga, C. (1949). Japan since Perry. New York: McGraw-Hill. Yoshimi, S. (2003). ‘America’ as desire and violence: Americanization in postwar Japan and Asia during the Cold War. Inter-Asia Cultural Studies, 4(3), 433–450. You, J. (1995). Changing capital-labor relations in South Korea. In: J. Achor & J. You (Eds), Capital, the state, and labor: A global perspective. Brookfield, VT: Edward Elgar. Yuen, C. C., & Lim, G. S. (2000). Globalization, labor market deregulation and trade unions in Singapore. In: C. Rowley & J. Benson (Eds), Globalization and labor in the Asia Pacific region. London: Frank Cass. Zeitlin, J. (1985). Shop floor bargaining and the state: A contradictory relationship. In: S. Tolliday & J. Zeitlin (Eds), Shop floor bargaining and the state: Historical and comparative perspectives. Cambridge: Cambridge University Press.
In Japanese Endo, M. (1985). Shokunin to seikatsu bunka: Nihon shokuninshi no kenkyu iv. Tokyo: Yuzangaku. Fujita, W. (1955). Dai ni kumiai. Tokyo: Nihon Hyoronsha. Fujita, W. (1959). Kumiai to sutoraiki: Tenkanki ni tatsu rodo kumiai. Tokyo: University of Tokyo Press. Iwada, Y. (1977). Nihonteki keiei no hensei genri. Tokyo: Bunsinto. Japanese Ministry of Labor. (1979). Rodokumiai kisochosa hokoku. Tokyo.
Enterprise Unionism in Asia
211
Japanese Ministry of Labor. (1999). Rodokumiai kisochosa. Tokyo. Koike, K. (1977). Shokuba no Rodokumiai to sanka. Tokyo: Toyo Keizai Shinposha. Komatsu, R. (1971). Kigyobetsu kumiai no seise. Tokyo: Ochanomizu Shobo. Kumazawa, M. (1989). Nihonteki keiei no meian. Tokyo: Chikumashobo. Nakane, C. (1967). Tateshakai no ningenkankei. Tokyo: Kodansha Gendai Shinsho. Nomura, M. (1993). Toyotism: Nihongata seisan shisutemu no seizuku to kanyo. Tokyo: Mineruwa Shobo. Nomura, M. (1994). Shushinkoyo. Tokyo: Iwanami Shoten. Numata, I. (1972). Rodoundo no kenri. Tokyo: Horitsu Bunkasha. Ota, K. (1971). Tatakai no nakade. Tokyo: Aoki Shoten. Sakurabayashi, M. (1986). SekiyushokkuIko no kigyobetsu kumiai to tansan. Tokyo: Kosaisangyo & Rodo Kenkyu Senta. Social Science Research Institute at the University of Tokyo. (1950). Sengo rodokumiai no jittai. Tokyo: University of Tokyo Press. Takahashi, K. (1965). Nihonteki roshikankei no kenkyu. Tokyo: Miraisha. Tona, N. (1993). Nihongata flexibility no kozo. Tokyo: Horitsu Bunkasha. Yamamoto, K. (1977). Sengokiki ni okeru rodokumiai. Tokyo: Ochanomizu Shobo. Yokoyama, G. (1899). Nihon no kasoshakai. Tokyo: Iwanami Shoten.
In Korean Gong, D. S. (2000). Hangookeui nodongjohapkwa nodong chyungchi. Seoul: Kyungjinsa. Hong, S. W. (1994). Nodong. In: J. H. Yun (Ed.), Hangook kyungjaeron kangeui. Seoul: Hanul. Kim, D. C. (1995). Hangooksahoe nodongja yongu: 1987-yon ihu rul chungsim uro. Seoul: Yoksa Pipyongsa. Kim, J. (1987). Iljehaui nodongundong eui banghyang junwhan he kanwhan yungu. In: H. S. Yunguhoi (Ed.), Iljeha ui Sahoiundong. Seoul, Korea: Moonhakkwa Jisungsa. Kim, J. (1999a). 5.16 Ihoo nodongjohap eui jepyungkwa ‘Hankook Nochyong Chejae’ eui sungrip. Sahoiwa Yoksa, 55, 103–144. Kim, J. (1999b). 20 segi hankook eui nodong: Yuksajuk kyunghum eui banchu. Kyungjaewa Sahoi, 44, 59–90. Kim, J. H. (2005). Giupbul nojo eui shildewa gujojuk hangae. Nodong Rybyu, 3, 24–37. Korea Labor Institute. (2004). 2003 Nodongjohap jojik hyunwhang bunsuk. Seoul: Korea Labor Institute. Lee, B. H., No, K. P., Oh, K. H. & In, S. B. (2001). Nodong johap sanggun ganbu yungu. Hangook Nodongsahoi Yunguso. Lee, H. C. (1984). 8–15 ui sahoikungjesajuk inshik. In: D. G. Lee & J. W. Young (Eds), Hankook zabonjuuiron. Seoul, Korea: Kachi. Park, D. J. (1990). Kooknae sanupbyul nojo nosagwangae heckshimjuck yukhal motago innyunga? Hangook Sabop Haengjung Hakhoi Pusol Nosa Munjae Yunguso, 27, 86–97. Park, D. J., & Park, G. S. (1989). Hangook eui nodongjohap (I). Seoul: Hangook Nodong Yeonguwon. Shin, K. Y. (1994). Gegupgwa nodongundong eui sahoihak. Seoul, Korea: Nanam. Yun, Y. D. (1991). Hankook chogi nodong undong uungu. Seoul: Iljogak.
CHINESE UNIONS: AN ALICE IN WONDERLAND DREAM WORLD$ David Metcalf and Jianwei Li ABSTRACT China has, apparently, more trade union members than the rest of the world put together, but the unions are subservient to the Party-state. The theme of the paper is the gap between rhetoric and reality. Issues analysed include union structure, membership, representation, and the interaction between unions and the Party-state. We suggest that Chinese unions inhabit an Alice in Wonderland dream world and that they are virtually impotent when it comes to representing workers. Because the Party-state recognises that such frailty may lead to instability it has passed new laws promoting collective contracts and established new tripartite institutions to mediate and arbitrate disputes. While such laws are welcome they are largely hollow: collective contracts are very different from collective bargaining and the incidence of cases dealt with by the tripartite institutions is tiny. Much supporting evidence is presented drawing on detailed case studies undertaken in Hainan Province (the largest and one of the oldest special economic zones) in 2004 and 2005. The need for more effective representation is appreciated by some All China Federation of Trade Unions (ACFTU) officials, but it seems a long way off, so unions in China $
Quotes are from Lewis Carroll (1865) Alice’s Adventures in Wonderland, (AAW) and (Carroll, 1872) Through the Looking Glass and What Alice Found There, (TLG). All page references are to the centenary edition Carroll (1865) and (1872), edited by Hugh Haughton (1998).
Advances in Industrial and Labor Relations, Volume 15, 213–268 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15005-2
213
214
DAVID METCALF AND JIANWEI LI
will continue to echo the White Queen: ‘‘The rule is, jam tomorrow and jam yesterday – but never jam today’’ and, alas, tomorrow never comes.
1. INTRODUCTION ‘‘Curiouser and curiouser’’ said Alice (TLG, 1998, p. 16)
China has, apparently, more trade union members than the rest of the world put together (Visser, 2003). But the unions do not function in the same way as western trade unions. In particular, Chinese unions are subservient to, and a component of, the Party-state. The theme of this paper is the gap between rhetoric and reality. Issues analysed include union structure, membership, representation and new laws (e.g. promoting collective contracts), new tripartite institutions and the interaction between unions and the Partystate. Regarding each of these issues, and others, we suggest that Chinese unions inhabit an Alice in Wonderland dream world. Alternatively, we try to tell the true story: Chinese unions have many members (though probably not as many as official documents suggest), but are virtually impotent when it comes to representing workers. Because, the Party-state recognises that such frailty may lead to instability (for example strikes were called riots [saoluan] by all the officials we spoke to) it has passed new labour laws promoting collective contracts and established new tripartite institutions to mediate and arbitrate in individual (but seldom collective) disputes. While these new laws and institutions are welcome they are largely hollow. Collective contracts are very different from collective bargaining and the incidence of cases dealt with by the tripartite institutions is tiny. Although this paper argues that China does not have properly functioning unions – rather the reverse – two caveats are necessary. First, we are seized by the fact that the whole notion of a ‘‘labour market’’ [laodongli shichang], the market in which unions must function, is only a decade or so old. The first time the phrase ‘‘labour market’’ was used in official documents was 1993 (Qiao, Lin, & Jiang, 2004), reflecting the previous Marxist aversion to exchanging labour for money. This was fundamentally altered in the 1994 Labour Law, which introduced labour contracts. Nevertheless our argument is that even though the labour market is now firmly established – 97% of workers are on the labour contract system – there is no evidence whatsoever of parallel development in functioning trade unions. The inability of workers to develop proper representation for their common
Chinese Unions: An Alice in Wonderland Dream World
215
interests, coupled with the rapid spread and deepening of the market mechanisms, implies that unions – despite their huge membership – are likely to remain largely nugatory (ineffective) in Chinese labour relations. But, second, neither China, nor the All China Federation of Trade Unions (ACFTU), are monolithic so there are (normally short-lived) exceptions to the rule that unions are nugatory. Some such exceptional examples will be noted as the paper proceeds. Section 2 describes unions’ structure, membership, voice and finances. In Section 3 we consider what unions do – to both industrial relations and to workplace efficiency and equity. Detailed case study evidence from Hainan Province, the largest and one of the oldest special economic zones, is set out in Section 4. The future prospects of Chinese unions is the subject of Section 5 that examines the interaction between unions and other parties and various challenges including declining legitimacy. Conclusions are presented in Section 6. In addition to documentary sources, the evidence here is from seven sets of interviews. These covered: three workplaces (a food conglomerate, tinplate producer, coconut juice producer); the Labour, Personnel and Social Security Bureau (Haikou City office); the Haikou City office of the All China Federation of Trade Unions; the Enterprise Confederation and Industrial and Business Confederation in Haikou City; and the ACFTU Headquarters in Beijing, China Institute of Industrial Relations. Throughout the text these are referred to, respectively, as Case 1, Case 2 etc., further details are given in Appendix B. Section 4 focuses entirely on the case study evidence from the workplaces of the three producers (Cases 1–3) in order to present a detailed description and analysis of what unions do and do not do in the workplace.
2. UNION ORGANISATION MEMBERSHIP, VOICE AND FINANCE 2.1. Organisational Structure ‘‘When I use a word’’, Humpty Dumpty said in a rather scornful tone, ‘‘it means just what I choose it to mean’’ (TLG, 1998, p. 186).
The formal structure of national Chinese unions was established in the early 1920s with the ACFTU setting up in Canton (now Guangzhou) in 1925. But
216
DAVID METCALF AND JIANWEI LI
it was after the liberation in 1949 that unions’ functions were consolidated and the structure formalised. Though unions have been through ups and downs during the past 50 years, their original organisational structure remains largely intact (Ng & Warner, 1998; Warner & Zhu, 2000). Unions’ organisational structure has three interrelated elements: democratic centralism, top down control and a dual local and industrial structure. The constitution (revised in 2003) of the ACFTU and its constituent unions stipulates unions’ organisational principle as ‘‘democratic centralism’’. And at ACFTU headquarters we were told that the relationship between the Party-state, unions and workers is one of ‘‘representative democracy’’ or ‘‘participative democracy’’. ‘‘Democratic’’ reflects the ostensible democracy enjoyed by the mass of workers. ‘‘Centralism’’ implies an authoritarian style of administration. This principle requires ‘‘individuals being obliged to the organization, the minority obliged to the majority, and subordinates obliged to superiors’’ (Ng & Warner, 1998). Therefore the logic emphasises a top-down control while allowing some ‘‘freedom’’ of opinions and actions at lower levels. Unions’ organisational structure follows this logic. According to the Trade Union Law 2001 and the Trade Union Constitution 2003, all workers enjoy the freedom to join a union, but this union must be approved by and be under the leadership of ACFTU, the only permitted official union organisation. In recent years, a few experiments have occurred in grass root unions to elect a union chairperson directly by the members of the workplace (Taylor, Chang, & Li, 2003). However, these experiments have not been widely replicated and such appointments must be ‘‘approved by local Party organs’’ in order to weed out ‘‘troublemakers’’ (Taylor et al., 2003). Union leaders of higher local (county, city and province) levels belong to the government administration and are appointed by the Chinese Communist Party (CCP). It is readily apparent that the ‘‘democratic’’ element of democratic centralism is in the Humpty Dumpty tradition. This will only be remedied when workers are allowed to choose and vote for their own officials and leaders. Chinese unions form a hierarchy with the ACFTU at the top, and with two strands representing industrial and geographic boundaries (see Appendix A). The basic organisational unit is the workplace-based grass root union in each administrative unit (enterprise, undertaking or state organ). There can be only one grass root union in each workplace. Above the grass root unions are two local levels – county/city and province. Grass root unions come under the leadership of both the city-level federation of unions and the appropriate industrial union. The city federation is under the direct leadership of the higher-level provincial federation. And the industrial union at
Chinese Unions: An Alice in Wonderland Dream World
217
this city level has two bosses – the higher national industrial union and the provincial federation. At the top of the union structure is ACFTU. There are now 31 federations of trade unions based on provinces, autonomous regions and metropolitan areas and ten national industrial unions. All lower-level unions are responsible to their immediate upper-level unions and report their work to them. (‘‘About ACFTU’’ from http://www.acftu.org.cn/ about.htm). Within the unions, interaction between union officials and members is close to non-existent. In almost all cases, union officials are not elected by members but chosen by appropriate Party organ. Union officials are responsible to the Party and government administration and empowered by the state. Their career and promotion does not intersect with union members whom, according to trade union law, they should represent and protect (Taylor et al., 2003). This dichotomy between union officials and members weakens any power base, which might be derived from the members and simultaneously demotivates union officials from working for their constituency. Union members – who have almost no rights in choosing and changing union officials – naturally have little trust in union officials. Chinese unions essentially operate on behalf of the state and management rather than workers. Both the organisational principle and the structure of unions enhance unity in thought and action within unions, all under the leadership of ACFTU whose officials are chosen by and are responsible for the Party-state. The Trade Union Law 1992 and its 2001 revision assigned the ACFTU the role of a two-way transmission belt between the mass of workers and the Party: ‘‘by top-down transmission, mobilization of workers for labour production on behalf of the state; and by bottom–up transmission protection of workers’ rights and interests’’ (Chan, 2000). In reality, Chan concludes: ‘‘the Chinese state was so powerful that the top-down transmission of Party directives regularly suppressed any bottom–up transmission relating to workers’ interests. The union merely functioned as an arm of the Party-state’’. Workers are left with little possibility of channelling their grievances upward: such a lack of upward voice suggests, as we shall see, future potential social unrest. Chinese unions have close relations with the Party-state. This closeness dates back to the revolutionary history in the early 1900s, when unions were setup to form an alliance with the CCP to fight against military warlords, foreign invaders and the National Party. Chinese workers’ class interests were considered a source of power for the CCP, but have always been subordinate to the Party’s needs. When the power base of the CCP moved to the countryside, unions became a sideshow. It was the promotion of
218
DAVID METCALF AND JIANWEI LI
industrialisation soon after liberation in 1949 that restored and reinforced the importance of the ACFTU and its affiliated unions. Then, during the Cultural Revolution of 1966–1976 unions were dissolved: it was held that the Party-state represented the interests of all, including workers, so there was no need for the existence of an organisation emphasising separate interests. Once the new round of economic reforms started in 1978, unions regained their legitimacy, ‘‘sponsored’’ by the Party-state, as before. Unions’ evolution demonstrates that they have never been a separate institution just for their members, but the junior partner of the CCP. Though the ACFTU’s ostensible constitutional obligation, reinforced by Trade Union Law 2001, is to represent workers’ interests, it also imposes a strong political role of safeguarding the interests of the country. One of the four cardinal principles that ACFTU must abide by is to uphold the CCP’s leadership (Trade Union Law 2001). This requires ACFTU to have its own policies and activities congruent with the Central Committee of the CCP. Unions’ political subordination to CCP permeates all their actions and priorities so some authorities describe Chinese unions as a ‘‘party organ’’ (Taylor et al., 2003). The banning of the right of strike in the 1982 Chinese Constitution eliminates a key source of union power. Indeed, unions now depend on the powerful state for legislative empowerment in order to survive. Hence, Child (1994) points out that Chinese unions ‘‘continue to be subordinated to the authority of the Party’’ and function as top-down oneway ‘‘transmission belt’’ between workers and the Party-state. However, the move towards a market economy has encouraged the Partystate not to ignore unions’ other organisational responsibilities – representing their constituency of workers. Any social unrest from laid-off workers may destabilise the Party’s control, so the Party-state needs a channel for workers to effectively voice their worries and grievances. And with the Party withdrawing from the direct control of public-owned enterprises and the fast growth of the non-public sector, the Party needs unions if there should be a desire to counter the power of management in China’s monopsonic labour market. Under the rhetoric of the 2001 Trade Union Law, unions are now charged with protecting workers from exploitation by management, unfortunately, the reality is that such protection is seldom present. At national and provincial level the ACFTU have been working hard to get legislative changes on behalf of workers, with some success. Recent favourable clauses in the Labour Law 1994 and the Trade Union Law 2001 include shorter maximum working hours and overtime hours per week, greater supervision of working conditions and the signing of collective contracts. But at micro-enterprise level when union officials challenge
Chinese Unions: An Alice in Wonderland Dream World
219
management to try to protect workers’ rights, they are often dismissed or have their pay cut (e.g. Workers’ Daily, 6 February 2004; http:// www.sina.com.cn, 18 April 2004). Several factors constrain unions’ representative role. First, the Party worries that too strong a union could become a power adversary. Second, to promote economic performance and attract investors, local government urges unions and workers to cooperate with, rather than confront, management. Third, employers are prepared to violate the law in order to remain non-union. Fourth, some local union officials, who are linked more to the Party, are reluctant to act on workers’ behalf. Fifth, enterprise level research shows (Zhu & Warner, 2003; Cooke, 2002; Ng & Warner, 1998; Leung, 1988) that unions are aligned with management rather than workers. Indeed, their positions are often interchangeable with the union chairperson being part of management. Such constraints on unions’ representative role sometimes lead workers to risk breaking the law to fight for their rights by, for example, organizing wildcat strikes or even setting up independent unions. The Party-state has a dilemma concerning its control and empowerment of unions: without empowerment such illegal action is likely to become more frequent; but empowerment also risks some union members and officials challenging party orthodoxy – for example, on retrenchment of employment in the State Owned Enterprises (SOEs) and the harmony of interests between capital and labour. 2.2. Membership and Density Virtually in every country the world union membership echoes the Red Queen: It takes all the running you can do to keep in the same place (TLG, 1998, p. 143).
but in China in the early part of the new millennium there was a 50 million boost in the number of members in just three years. Such a hike suggests that the ACFTU takes a similar approach to the Queen of Hearts: Sentence first – verdict afterwards, (AAW, 1998, p. 107).
Employment and unionisation in China is set out in Table 1. Total employment was 752 million in 2004 split roughly one-third urban (265 million) and two thirds rural (487 million). Employment has risen monotonically since 1970 – apparently little influenced by cyclical variation. Staff and workers – the core of state employment – are mostly urban based and are
220
Table 1. Year
1
Employment (Millions)
Unions
Density (%)
Total
Urban
Rural
Staff & workers
Number of union members (millions)
Number of grass root unions
6/2
6/3
6/5
2 ¼ 3+4
3
4
5
6
7
8
9
10
207.29 258.10 345.32 423.61 437.25 452.95 464.36 481.97 498.73 512.82 527.83 543.34 553.29 647.49 654.91 661.52 668.08
24.86 45.37 64.12 105.25 110.53 114.28 117.46 122.29 128.08 132.92 137.83 142.67 143.90 170.41 174.65 178.61 182.62
182.43 212.73 281.20 318.36 326.72 338.67 346.90 359.68 370.65 379.90 390.00 400.67 409.39 477.08 480.26 482.91 485.46
16.03 43.21 62.16 104.44 109.40 112.81 115.15 118.90 123.58 128.09 132.14 136.08 137.42 140.59 145.08 147.92 148.49
10.03 17.67
4.8 6.8
40.3 38.9
62.6 40.9
61.17 68.44 73.32 76.93 80.29 85.26 89.09 93.37 96.29 99.09 101.36 103.89 103.23 101.76
14.4 15.7 16.2 16.6 16.7 17.1 17.4 17.7 17.7 17.9 15.7 15.9 15.6 15.2
58.1 61.9 64.2 65.5 65.7 66.6 67.0 67.7 67.5 68.9 59.5 59.5 57.8 55.7
58.6 62.6 65.0 66.8 67.5 69.0 69.6 70.7 70.8 72.1 72.1 71.6 69.8 68.5
589,212 606,045 616,852 626,666
DAVID METCALF AND JIANWEI LI
1952 1962 1970 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
Employment and Union Membership and Density 1952–2004.
674.55 680.65 689.50 698.20 706.37 713.94 720.85 730.25 737.40 744.32 752.00
186.53 190.40 199.22 207.81 216.16 224.12 231.51 239.40 247.80 256.39 264.76
488.02 490.25 490.28 490.39 490.21 489.82 489.34 490.85 489.60 487.93 487.24
148.49 149.08 148.45 146.68 123.37 117.73 112.59 107.92 105.58 104.92 105.76
102.03 104.00 102.19 91.31 89.13 86.90 103.13
582,780 593,113 586,672 510,276 503,532 508,592 858,592
15.1 15.3 14.8 13.1 12.6 12.2 14.3
54.7 54.6 51.3 43.9 41.2 38.8 44.5
133.98 123.40 136.94
1,712,528 905,516 1,020,045
18.2 16.6 18.2
54.1 48.1 51.7
68.7 69.8 68.8 62.4 72.2 73.8 91.6 127 118 129
Note: 1. Staff and workers refers to those who work and get paid at state-owned, collective-owned, joint-stock, foreign and Hong Kong (HK), Macao and Taiwan owned firms, other economic units and their attached institutions. It excludes re-employed retired and resigned people, locally hired teachers, investors of private enterprises, foreigners and people from HK, Macao and Taiwan who are working at the state-owned units. It includes those who have left their jobs and are not engaged in work, but who still keep work relationship with the unit and have not found relatively stable work (p. 60 and 119 of Chinese Trade Unions Statistics Yearbook 2001). Most of staff and workers are in urban employment. 2. The most reasonable density measure is Column 9 (see text). But it overstates true density among urban workers because a small fraction of rural employees – those in township and village enterprises for example – are unionised and are included in the membership Column 6. Source: 1. Columns 2, 3 and 4 from Table 1 to 5 of China Labour Statistical Yearbook 2003. 2. Column 5 from Table 1 to 13 of China Labour Statistical Yearbook 2003 and 2 to 6 of Chinese Trade Unions Statistics Yearbook 2005. 3. Column 6, data of union membership from 1980 to 2000 is adapted from the website of the Industrial Relations Department of Cornell University and Table 3–12 of Chinese Trade Unions Statistics Yearbook 2005. The 1952 and 1962 union membership data are from Table 22 to 3 of China Labour Statistical Yearbook 2003. 4. Column 7 is from Table 3–9 of Chinese Trade Unions Statistics Yearbook 2005.
Chinese Unions: An Alice in Wonderland Dream World
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
221
222
DAVID METCALF AND JIANWEI LI
the most likely to belong to a union; the number of staff and workers has fallen rapidly in the last decade, reflecting the changes in the patterns of ownership of Chinese industry. Union membership and the number of grass roots unions are detailed in Columns 6 and 7. In China, ‘‘all manual and brain workers in enterprises, institutions and government departments within the territory of China who rely on wages or salaries as their main source of income, irrespective of their nationality, race, sex, occupation, religious belief or educational background, have the right to organize or join trade unions according to law’’ (Article 3, Chapter I, Trade Union Law, 2001 Revision). The definition of union members is very broad or ‘‘generous’’ (Ng & Warner, 2000) in order for the ACFTU to claim substantial membership. The grass root union is workplace based. It refers to a union in a workplace, an enterprise, an institution or a government department with a membership of 25 or more, or a union of joint workplaces with each having fewer than 25 members (Article 10 of Chapter II, Trade Union Law, 2001 Revision). The People’s Republic of China (PRC) was founded in 1949, and for 10 years membership remained below 20 million. During the Cultural Revolution 1966–1976 unions were dissolved, but once revived membership grew rapidly and hit 100 m by 1990. From 1995 to 1999, the number of staff and workers and union members haemorrhaged with the shedding of workers in state-owned and collective-owned units, which comprise the bedrock of union membership. But, according to official statistics, in the new millennium unions are, apparently, thriving. The number of grass roots unions more than trebled between 1999 and 2002, but has since fallen back to around 1 million. Union membership rose by over half between 1999 and 2002 and now stands at 137 million. Presently, there are 470,000 full-time union cadres – in local federations and enterprises – and other part-time cadres and activists. Thus, there is one full-time official for every 280 union members, a ratio well below that laid down in Chinese labour law (Case 7). There are a number of reasons why unions appear to suddenly be flourishing despite the continuing crumbling away of jobs in state-owned and collectively owned units. First, a law introduced in 1998 and revised in 2001 requires new workplaces to establish a union branch. In 1998, ACFTU circulated Opinions on Strengthening Unions’ Work in Restructured Small and Medium Sized State Owned Enterprises clearly demanding them to set up and perfect union organisations. This was to offset the erosion of union members, which resulted from downsizing and restructuring among the larger SOEs. In 2001, the ACFTU circulated a further official document Opinions on Strengthening Organising Work in Newly Set-up Enterprises to
Chinese Unions: An Alice in Wonderland Dream World
223
encourage organisation of workers in new workplaces. Subsequently, these documents were put in the 2001 revision of trade union laws. Second, the ACFTU has begun to admit migrant workers into membership. Third, recently, unions put much greater effort into organising private and foreignfunded workplaces. The Thirteenth Congress of national trade unions in 1998 stressed that ‘‘where there are staff and workers trade unions must be set up’’ (ACFTU, 2003) and agreed for the first time that ‘‘organising staff and workers into a union’’ is central to their work. Then the 2000 ACFTU presidium held a symposium on organising newly established enterprises, and agreed that this was the most urgent task for the ACFTU. It set a target of establishing 1 m new branches and an extra 36 m members by the end of 2002. These ambitious targets have, ostensibly, been met. But there is grave doubt about the veracity of these figures: ‘‘the consensus is that these numbers are grossly inflated’’ (American Center for International Labor Solidarity (ACILS), 2004; Clarke, 2005). Leung (2002) suggested, for example, that from 2000 to 2002 grass roots branches were counted by workplace instead of enterprise, such that enterprises with a number of plants suddenly multiply the number of union branches. However, this practice stopped in 2003 such that the number of grass roots unions fell back. Further, the number of branches and members were simply overstated by some local officials in order to hit their targets. For example, in some places a local federation of unions was set up to include all local enterprises and although all the employees were then counted as members their enrolment is on paper rather than in fact. We were told by an ACFTU official (Case 7) that ‘‘although I believe the (membership) figures because they are official, there is a lot of water in the figures’’ [jinguan wo xiangxin zhexie guanfang shuzi, dan qizhong you xuduo shuifen]. Union density depends on which measure of employment is used as the denominator. There are three possibilities from the available statistical series: total employment, urban employment or staff and workers. As the majority of rural workers are in farming they will not be in a union so total employment is not a sensible denominator. Similarly, membership spreads beyond staff and workers to other urban employees like short-term contract workers or part-time staff. So total urban employment seems the most reasonable series. The evidence in Table 1 suggests that around half such employees are presently unionised. Density rose steadily in the 1980s and peaked at 70% in 1989. In the next decade it fell back to 39% by 1999. This was because the huge growth in urban jobs outside state-owned and collective-owned enterprises was not matched by a corresponding boost in membership. Most new urban employees were non-union: many did not
224
DAVID METCALF AND JIANWEI LI
have urban residential status and/or were not permanent employees but were on temporary contracts, e.g., migrant peasant workers. But, if the figures are to be believed, density rose rapidly in the last few years to stand at 52% in 2004. This reflects unions’ organising efforts in the previously non-union private enterprises and foreign-funded firms and among migrant peasant workers. In 2004, 48% of urban employment was not unionised. Such non-union groups include: the self employed; private business owners; a significant fraction of the peasants who have migrated to urban areas and who are often temporary or part-time workers or sub-contractors; non-union employees in unionised workplaces including new and/or young workers and contract workers; military personnel; and religious professionals. Female employees account for 38% of employment in urban units (this is a narrower definition than total urban employment: it includes only workers with contracts and excludes, e.g., temporary workers in urban workplaces; the measure is very similar to the total number of ‘‘staff and workers’’) and a virtually identical proportion (37.8%) of union membership (see Table 2). Thus, union density rates for men and women are the same. It is worth remarking that the downsizing of employment in state-owned and collective-owned units hit female jobs relatively more than male ones, and this led through to a lower density figure for females than previously. Membership and density information by industrial sector is given in Table 3, using three alternative measures of employment. The evidence on Table 2.
Union Membership by Gender, 2000.
Employment in Urban Units (Em, Ef, Et)
Male Female Total
Union Members (Um, Uf, Ut)
Density (%)
Numbers (million)
Shares (%)
Number (million)
Shares (%)
Um/Em or Uf/Ef
1
2
3
4
5
72.012 44.113 116.125
62.0 38.0 100
63.997 39.137 103.134
62.2 37.8 100
88.9 88.7 88.8
Source: 1. Column 1, male, female and total employment in urban units (Em, Ef, Et), from Tables 1 to 9 and 1 to 10 of China Labour Statistical Yearbook 2003. 2. Column 3, male, female and total union members (Um, Uf, Ut), from Table 3.2 of Chinese Trade Unions Statistics Yearbook 2001.
Union Membership and Density by Sector, 1999.
Sector
1
Total Employment (Millions)
Employment in Urban Units (Millions)
Staff and Workers (Millions)
Union Members (Millions)
Density (%)
5/2
5/3
5/4
2
3
4
5
6
7
8
Farming, forestry animal husbandry and fishery Mining and quarrying Manufacturing Production and supply of electricity, gas and water Construction Geological prospecting and water conservancy Transport, storage, post and telecommunications Wholesale and retail trade & catering services Finance and insurance Real estate trade Social services Health care, sports and social welfare Education, culture and arts, radio, film and television Scientific research and Polytechnic services Government agencies, party agencies and social organisations Others
334.93 6.67 81.09 2.85 34.12 1.11 20.22 47.51 3.28 0.96 9.23 4.82 15.68
5.37 6.55 35.54 2.85 8.15 1.11 7.04 11.42 3.28 0.97 4.76 4.82 15.68
5.19 6.50 34.96 2.83 7.78 1.10 6.82 11.10 3.00 0.90 4.53 4.73 14.80
3.97 5.14 30.16 2.53 4.13 0.89 6.73 7.87 1.30 0.50 2.37 2.96 10.34
1.18 77.1 37.2 88.8 12.1 80.2 33.3 16.6 39.6 52.1 25.7 61.4 65.9
73.9 78.5 84.9 88.8 57.7 80.2 95.6 68.9 39.6 51.5 49.8 61.4 65.9
76.5 79.1 86.3 89.4 53.1 80.9 98.7 70.9 43.3 55.6 52.3 62.6 69.9
1.74 11.02
1.74 11.02
1.65 10.88
0.83 5.90
47.7 53.5
47.7 53.5
50.3 54.2
49.69
1.01
0.96
1.29
2.6
Total
713.94
121.30
117.73
86.90
12.2
71.6
134 73.8
225
Source: 1. Columns 2, 3 and 4 from Table 1–6, 1–9 and 1–14 of China Labour Statistical Yearbook 2003, respectively. 2. Column 5 from unpublished ACFTU document of Union Membership by Sectors, 1995–1999. Note: Data are as in the source. Total row is not exactly the sum of the sectors for columns 2, 3, 4.
127
Chinese Unions: An Alice in Wonderland Dream World
Table 3.
226
DAVID METCALF AND JIANWEI LI
union membership is from an internal ACFTU document and is only available for 1999. Around one-third of union members work in manufacturing and a further one-third in transport and communications, wholesale and retail trades, education and government agencies. Density rates (using employment in urban units as the denominator) are highest in manufacturing, the gas, water and electrical utilities and transport and communication. The very high density in these sectors is found in other countries (e.g. UK), and reflects common features such as the ease of organisation in large workplaces. Union membership and density vary sharply according to the form of ownership (see Table 4). State owned enterprises account for two fifths of members, and state institutions and government agencies add a further quarter of all members. The bulk of the remainder are employed by collectively owned enterprises, limited liability corporations and township enterprises. It is difficult to calculate accurate density figures by ownership because we only have total employment or employment of staff and workers. But the evidence in Table 4 suggests density is highest where the state is involved, namely state-owned enterprises and state institutions and government agencies. Township enterprises in rural areas employ 128 million people, but have just 5 m union members so their density rate is low. Union membership is concentrated in the East and Central regions of China (Table 5) reflecting their total employment share, larger workplaces and the importance of state owned enterprises in these regions. Each region has a range of density by province reflecting the aim of achieving an even spread of industries across the various regions. The three coastal provinces of Jiangsu, Jiangxi and Guandong have relatively low-union density because they have many small, private businesses and many relatives of overseas Chinese who sponsor such businesses. These private businesses, and foreign owned firms which locate here, also absorb many migrating peasant workers (from inside or outside the province) who are normally not union members. 2.3. Representative and Direct Voice In principle there are a variety of mechanisms providing both representative and direct voice to workers. Representative voice occurs via an institutional intermediary such as a trade union, whereas direct voice permits the employee to influence management without any mediated institution. Representative voice through a trade union includes collective consultation, collective contracts (i.e. collective agreements), Labour Dispute Mediation/ Arbitration Committees, and the tripartite system of relations embracing the
Table 4. Union Membership and Density by Workplace Ownership, 2000. Total Employment (Millions)
1 Domestic funded enterprises
Hongkong, Macao and Taiwan funded enterprises Foreign funded enterprises (FFEs) Town & township enterprises (TTEs) Institutions Government agencies Others Total
State-owned enterprises (SOEs) Collective-owned enterprises (COEs) Share holding enterprises (SHEs) Joint-owned enterprises Limited liability corporations Share holding corporations (SHCs) Private enterprises Other units
Staff and Workers (Millions)
Number of Union Grass Root Members Unions (Millions)
Density (%)
5/2
5/3
2
3
4
5
6
7
81.02 (SOUs) 14.99 (COUs) 1.55 0.42 6.87 4.57 24.07 50.70
44.35 8.25 3.10 0.31 6.36 5.55 5.77 0.43
139,804 70,941 22,034 1,791 20,659 13,630 152,435 6,989
41.37 7.33 2.71 0.27 5.86 5.07 4.37 0.37
51.1 48.9 174.8 64.3 85.3 110.9 18.1 0.07
93.3 88.8 87.4 85.9 92.1 91.4 75.7 86.0
3.10
2.94
19,759
2.03
65.5
69.0
3.32 128.20
20,335 92,967 196,056 118,623 13,284
2.41 5.23 18.19 6.60 0.96
72.6 4.1
392.69
3.06 6.31 19.27 6.97 1.04
0.2
78.8 82.9 94.4 94.7 92.5
711.50
113.71
889,307
103.13
14.4
90.7
227
Note: Since State-owned Units (SOUs) include State-owned Enterprises (SOEs), Government Agencies and part of Institutions, the number of employees in SOUs is larger than that in SOEs. The same situation is to Collective-owned Units (SOUs) and Collective-owned Enterprises (COEs). Source: 1. Column 2 are adapted from Table 2.3 of Chinese Trade Unions Statistics Yearbook 2001. (1) 81.02 m in column 2 include employees in SOEs, government and party agencies and part of institutions. Likewise 14.99 m include employees in SOEs, collective institutions in urban area. (2) Employees of Private Enterprise are calculated by the number of private enterprises in urban areas (12.68 m) plus that in rural areas (11.39 m). (3) The categories of self-employed individuals in both urban and rural areas are put under Other units in Domestic Funded Enterprises; therefore 50.70 m is got by adding the two together: 21.36 m for urban and 29.34 m for rural. (4) The last category of ‘‘Others’’ (392.69 m) includes the residues of urban employment (62.86 m) and peasants (329.83 m). 2. Columns 3, 4 and 5 are from Table 3.5, 3.8 and 3.11 of Chinese Trade Unions Statistics Yearbook 2001, respectively.
Chinese Unions: An Alice in Wonderland Dream World
Types of Ownership
228
DAVID METCALF AND JIANWEI LI
Table 5. Region
Trade Unions: Density and Distribution by Region, 2000. Province/Metropolitan/ Autonomous Region
Union Numbers (millions)
Share of all Members(%)
Union Density (%)
2
3
4
5
1 North China
Beijing Tianjin Hebei Shanxi Inner Mongolia Sub-total
2.281 2.327 5.524 3.564 2.680 16.375
2.2 2.3 5.3 3.4 2.6 15.8
47.1 64.3 61.4 63.1 65.8 60.25
Northeast China
Liaoning Jilin Heilongjiang Sub-total
7.516 2.719 4.556 14.791
7.3 2.6 4.4 14.3
69.1 51.5 55.4 60.67
East China
Shanghai Jiangsu Zhejiang Anhui Fujian Jiangxi Shandong Sub-total
3.739 6.267 4.494 3.193 2.666 2.373 7.665 30.396
3.6 6.0 4.3 3.1 2.6 2.3 7.4 29.3
60.1 41.3 45.8 46.7 47.8 44.6 52.2 47.77
Central China
Henan Hubei Hunan Guangdong Guangxi Hainan Sub-total
6.077 4.756 4.585 6.159 2.192 0.674 24.443
5.9 4.6 4.4 6.0 2.1 0.7 23.6
50.6 49.5 50.3 37.7 52.3 52.9 46.53
Southeast China
Chongqing Sichuan Guizhou Yunnan Tibet Sub-total
1.753 4.455 1.580 2.155 0.136 10.078
1.7 4.3 1.5 2.1 0.1 9.7
44.9 49.9 55.2 53.9 79.5 50.74
West China
Shaanxi Gansu Qinghai Ningxia Xinjiang Sub-total
2.789 1.547 0.464 0.524 2.008 7.332
2.7 1.5 0.4 0.5 1.9 7.1
52.7 54.1 57.4 54.4 61.4 55.57
Chinese Unions: An Alice in Wonderland Dream World
229
Table 5. (Continued ) Region
1
Province/Metropolitan/ Autonomous Region
Union Numbers (millions)
Share of all Members(%)
Union Density (%)
2
3
4
5
0.159 0.041 103.615
0.2 0.0 100
– – 51.5
Government Agencies Agencies Directly Under CCCPC Total
Source: 1. Column 3 from Table 2.5 of Chinese Trade Unions Statistics Yearbook 2001. 2. Columns 4 and 5 from Table 3.10 of Chinese Trade Unions Statistics Yearbook 2001.
Ministers of Labour, ACFTU and the National Enterprises Association. Representative voice also happens via a Workers Congress – analogous to European Union (EU) works councils – in some enterprises. Direct voice includes worker representation on the Board of Directors or Supervision Committees and, more recently, through various components of human resource management (HRM) including teams, performance appraisals and profit sharing. Here we discuss the coverage of such arrangements. Their effectiveness is altogether another matter – discussed here, in our case study and in later sections. Evidence on workers’ voice – both representative and direct – in Chinese workplaces is contained in Table 6. Unions represent their members when collective consultations are held with management, for example about workplace safety or welfare arrangements, or when collective contracts are signed. In year 2004, there were some 1.020 m grass root trade unions but only 0.327 m collective consultation systems and only 0.331 m enterprises signed collective contracts. In most workplaces where there is a collective contract there is an arrangement for collective consultation. Thus, the mere existence of a grass roots union is insufficient to provide representative voice because some three fifths of enterprises with such unions have no collective voice arrangements. This can also be seen from another angle: there are just 60 m workers covered by collective agreements, only around half the total of union members and a quarter of urban employees. Further, many commentators (e.g. Qiao et al., 2004) question the content of such collective agreements. Around a fifth of units with grass roots trade unions (1.02 m) have an internal workplace-based Labour Dispute Mediation Committee (LDMC) (0.20 m) and there are 0.43 m worker representatives on such LDMCs. These committees mediated in 0.192 m disputes of that only a tiny number
230
Table 6. Voice Mechanism Representative
Unions
Representative and/or Direct Voice, 2004. Numbers
Grass root trade unions Enterprises with equal consultation system Enterprises signing collective Enterprises signing contracts collective contracts independently Enterprises not signing collective contracts independently but covered by regional industrial collective contracts Staff and workers in enterprises covered by collective contracts Grass root units with labour dispute mediation committee (LDMC) Trade union and staff and workers representatives in LDMC Collective labour dispute Cases accepted by LDMC Successfully mediated labour dispute
1,020,045 327,763 235,306
96,927
59,933,097 195,403 26,460 6,752 54,537
DAVID METCALF AND JIANWEI LI
Indicators
Direct
One scheme of direct voice
Source: 1. Data on unions are from Tables 3.13, 3.15 and 3.31 of Chinese Trade Unions Statistics Yearbook 2005. 2. Data on Workers’ Congress are from Table 3.14 of Chinese Trade Unions Statistics Yearbook 2005. 3. Data on One Scheme of Direct Voice are from Table 3.16 of Chinese Trade Unions Statistics Yearbook 2005.
15,585 14,447 14,731 368,731 214,865 253,130 316,258 41,430 95,730 40,105 79,517 6,610,729
Chinese Unions: An Alice in Wonderland Dream World
Workers’ congress
Trade union representatives in labour dispute arbitration committee (LDAC) Trade union cadres with labour dispute arbitrator qualification Labour dispute cases handled by trade union labour dispute arbitrators Enterprises with staff and workers’ congress system Enterprises reporting business entertainment expenses to staff and workers’ congress Enterprises conducting democratic assessment on leaders and cadres Enterprises with transparent business affairs system Enterprises with staff and workers representatives in board of directors Staff and workers representatives in board of directors Enterprises with staff and workers representatives in supervision committees Staff and workers representatives in supervision committees Number of rational suggestions made by staff
231
232
DAVID METCALF AND JIANWEI LI
(0.068 m) were about collective issues. The vast majority concerned individual issues such as dismissals and promotions. Mediation was successful in one quarter of the cases. Where there is no LDMC in the workplace any dispute is negotiated directly between the worker and management or, more likely, the worker is given a take-it or leave-it option. There is a much small number of Labour Dispute Arbitration Committees (LDAC), which provides binding arbitration by neutral persons from outside the workplace, many of whom are trade union cadres with arbitration qualifications. The LDACs normally involve persons from local government administration. This description of the formal institutional arrangements does not capture the flavour of actual voice provided by trade unions. Unions are certainly consulted on issues that concern workers such that ‘‘the proposals of management or the trade union were referred to lower levels for discussion, and their [members’] comments and recommendations were reported back to the enterprise trade union for its consideration y However, the process of ‘‘consultation’’ with the members is more of an exercise in propaganda and persuasion than of the active participation of the membership’’ (Clarke, Lee, & Li, 2004). When a dispute arises between an individual or many employees and management, the union represents the workers, but their role in such disputes is mediation not negotiation (Clarke, 2005). Chen (2003) further clarifies unions’ dispute-resolving role into three categories – representing, mediating and pre-empting, resulting from their double identity as ‘‘both a state apparatus and the labour organisation’’. Chen’s analysis shows the state’s attitudes and actions affect unions’ representation role. When conflicts occur on an obvious infringement of individual workers rights by management, unions are willing to be on the individual worker’s side. The state is willing to represent and protect the weak as long as it does not arouse group dissatisfaction and social disturbance. When management clearly infringe collective workers’ rights, unions are much more cautious in taking action. To avoid group dissatisfaction and to ensure that this does not escalate into group violence, they instead mediate with management to retrieve the collective rights of the group. However, when the dissatisfied group of workers turn their disputes with management into organising their own fight either by violence or by organising themselves outside the union (which is against the law) unions stand on the government’s side – to preempt, control, then deal with the issue. Collective consultation was also promoted by some unions in the mid 1990s as a Trojan horse – a step on the road towards more substantive collective contracts – as well as beneficial in its own right with its emphasis on conflict-avoidance. Under such arrangements, unions are not considered
Chinese Unions: An Alice in Wonderland Dream World
233
as an equal bargaining partner and management has the final say, but such consultation provides the union with a presence in the workplace. Unions sign collective contracts with management on behalf of workers. It was the unions who initiated the practice of drawing up collective contracts via the 1992 Trade Unions Law, further reinforced by Labour Law 1994. The aim is to safeguard workers’ rights (Taylor et al., 2003). As unions were authorised by the law to represent workers, collective contracts encouraged management to take unions seriously and ‘‘discuss’’ with them work-related issues. The signing of collective contracts received intense coverage in Chinese newspapers and was hailed as ‘‘a breakthrough in China’s industrial relations’’. But due to lack of government support, the implementation was initially very limited. At the enterprise level, proper consultation and negotiation can only take place in those SOEs and joint ventures, which are financially viable and in big foreign owned enterprises who care about their reputation and are willing to abide by Chinese laws (Taylor et al., 2003; Chan, 2000). While all this seems quite impressive, it must be recognised that the content of collective contracts is normally very basic: y ‘‘the discussion on the draft collective contract with management appears to be more a process of consultation than negotiation, with the trade union deferring to management on any contentious issues y [the contract often just] reproduces the existing legal obligations of management’’ (Clarke et al., 2004). Many issues about workers’ benefits – which would tend to raise labour costs – are deliberately omitted from these contracts so as not to constrain management. In a nutshell, collective contracts are not about negotiation but rather ‘‘as self regulatory collective institutional mechanism to secure ‘harmonious’ labour relations’’ (Clarke et al., 2004; see also Warner & Ng, 1999). To revitalise scenescent SOEs, large-scale downsizing occurred after the fifteenth Central Committee-Chinese Communist Party (CC-CCP) meeting in 1997. SOEs were encouraged to further deepen their reforms by merger and acquisition, leasing and contracting out. During this reform process workers’ voice was sidelined. This neglect, coupled with official and management corruption, led to unprecedented social unrest among laid off workers. This, in turn, encouraged the state to search beyond the tame collective contract system for social dialogue involving unions, the state labour department and the enterprise associations – a tripartite system at the national level. As Clarke and Lee (2002) state: ‘‘while the Party-state has continued to use the ACFTU as an instrument for the mobilisation and control of the urban population, it has become increasingly aware that, if the ACFTU is to be effective as such an instrument, it has to articulate the
234
DAVID METCALF AND JIANWEI LI
aspirations and grievances of its members’’. Experience demonstrated and persuaded the ACFTU that the subordination of unions to management during the collective contract signing process implied they were unable to alter industrial relations on their own. Rather they need government support for any effective measure, which has the aim of representing and protecting workers. They found such support in the tripartite system previously promoted (as luck would have it) by the International Labour Office (ILO). The tripartite system started at national level in 2001 among the Ministry of Labour, ACFTU and the National Enterprises Association. The ACFTU saw it as ‘‘an instrument of ‘participation from the top’, and most importantly as a means of influencing legislation and government policies’’ (Clarke & Lee, 2002). The ACFTU intends to use the tripartite system to resume and extend the campaign for collective contracts to non-state sectors. The aim is to coordinate among the three parties to solve labourrelated issues. But several weaknesses constrain the implementation of the tripartite system if it is to achieve the goals of the ACFTU. First, while the Labour Ministry is identified as the representative of government in the system, many of the most pressing issues go beyond their administrative domain including permission to downsize. Second, the representatives of enterprise rather than employers blur the different interests of employers, employees and the enterprises. Furthermore, unions are subordinated to the Party, hence government. This subordination and lack of a strike weapon may lead to ‘‘tripartite accord’’ rather than a ‘‘social dialogue’’ arrangement (Clarke & Lee, 2002). Nevertheless Clarke and Lee put a gloss on the new arrangements: ‘‘the new tripartite system marked the recognition of the need to develop the effective representation of the interests of employers and employees’’ and suggest that this could lead to the transition to effective institutions representing the three parties. Since 1986, some workplaces (0.369 m in 2004) also have a system of staff and workers Congress. These are somewhat analogous to European Works Councils and were established to involve workers in the grass roots management of the workplace and to enhance workplace democracy. These institutions can be viewed as an attempt to redress ‘‘the inadequacy of the official trade union structure to act as grass root workers’ representative and spokesman’’ (Leung, 1988) and to strengthen workers ‘‘democratic participation in management, supposedly expressing the unity of interests of employer and employee in the development of the enterprise’’ (Clarke, 2005) and to ‘‘decentralise the organisation of workers’ power’’ (Ng & Warner, 1998). But, again, these bodies are mostly a facade. Their presence has crumbled away in state owned enterprises and they never really got a
Chinese Unions: An Alice in Wonderland Dream World
235
foothold in private or foreign owned companies. In Haikou City we were told (Case 5) that outside SOEs and ex-SOEs no workers congress had been established and the Haikou City Union Federation could not force an enterprise to establish one. Even where a congress exists on paper it may not do much. A recent case study of a brewery noted that the Workers’ Congress had not met for more than two years. Further, the Workers’ Congress are, in practice, subservient to local trade unions that convene their meetings, provide the secretariat and manage their business when the congress is in recess. This subservience was recognised in the 2001 Trade Union law that granted unions the right to organise workers to participate in management via the congress and to monitor the work of the congress (Ng & Warner, 1998). Direct voice occurs via the Enterprise Law (Chan, 2000) or because the enterprise has implemented modern human resource management. Under the Enterprise Law a worker representative can be directly involved in discussions on production-related issues like work organisation, technical improvements and rationalisation. Other direct voice schemes use an annual survey where employees comment on workplace life and the ‘one share one voice scheme’ where workers in private enterprises who own one share can make their voice heard at shareholders meetings. But, as in the west, their voice gets drowned out by those with bigger holdings, including managerial staff (Cooke, 2002). The world-wide ubiquity of suggestion schemes is evidenced by the remarkably specific statistic of the number of rational suggestions made by Chinese workers (6,610,729). Recently, many foreign companies have set up joint ventures (JVs) and wholly foreign-owned enterprises (FOEs). Together with financial and advanced technology investment, they have also brought management practices from their home countries into China, among which is HRM. Most research on HRM practices in China is based on case studies which analyze HRM practices enterprises with different ownership (Zhu & Warner, 2003; Chiu, 2003; Cooke, 2002; Benson, Debroux, & Yausa, 2000; Bjo¨rkman & Lu, 2000). Unfortunately, we cannot generalise from case studies to the population of enterprises and it is very likely that case studies suggest all too rosy a picture of the extent and depth of direct voice. Teamwork is encouraged among employees due to traditional Chinese collectivism. This practice is due more to peer or group pressure than to increasing autonomy, for performance is normally appraised on a group basis and decides the lump sum bonus for the team. The bonus allocation within groups is more or less egalitarian to nurture harmonious relations. Therefore individual rewards are related to the performance of the group, which motivates employees to cooperate and share tacit knowledge with
236
DAVID METCALF AND JIANWEI LI
each other. Warner and Braun (2002) noted that teamwork was assessed in employees’ performance appraisal to encourage group cooperation among workers. Further, market competition intensified the need for enhanced product or service quality. Quality control has become ‘‘a core and indispensable element of flexibility’’. All employees depend on each other in immediate information sharing to solve quality problems (Zhu & Warner, 2003; Benson et al., 2000). Performance appraisal procedures not only include direct management interviews with their subordinates, but are also linked with coaching and formal training and development. In some companies both management and employees are asked to fill in a formal assessment form. In others informal talks are held between the two parties to get direct feedback. Such appraisals are typically held once a year (Bjo¨rkman & Lu, 2000). In all enterprises, with or without unions, the government encourages activities such as ‘‘transparent workplace affairs’’ and ‘‘factory director’s open day’’ as a channel of direct communication between management and workers. In their case studies Zhu and Warner (2003) also find considerable use of employee involvement schemes, like suggestion boxes, after work meetings and information sharing practices, and financial participation via profit sharing and employee share ownership schemes. 2.4. How Unions are Financed Sources of union revenue are set out in Table 7. The bulk of union income comes via the 2% payroll levy, collected by union officials or local government authorities (e.g. Haikou City Labour and Employment Bureau) and distributed to workplace, local/provincial and industrial layers, and ACFTU in Beijing as set out in the table. There is no problem collecting the 2% levy from employers in the non-trading state sector (e.g. civil service, local government, schools). And, in the past, SOEs paid over this 2% levy as a matter of course. But now, many enterprises – ex-SOEs, foreign-invested firms and (especially) smaller enterprises – either delay payment or simply refuse to pay. This results – quite literally – in the hapless City union official banging on the door of the enterprise to collect the 2% levy. Alas, such moral suasion is typically unsuccessful. A potential second source of union revenue is a levy of 0.5% of the wage of the individual worker, paid to the union at the workplace. We were told that, in fact, this levy is seldom deducted. ACFTU officials stated that they are reluctant to make a fuss about such non-deductions because, if the levy was rigorously enforced, workers would start questioning what services and
Chinese Unions: An Alice in Wonderland Dream World
237
Table 7. Sources of Trade Union Revenue. Source
Element
Wage bill 2% of wage bill paid either to (i) local government authorities for state employees in non-tradeable occupations or (ii) the City or Provincial branch of ACFTU for the tradeable sector This is split as follows: 0.05 transferred to ACFTU in Beijing 0.38 transferred to the local, Provincial and industrial union 0.57 returned to workplace, e.g., to pay salary of union officials, for welfare role such as feast at Spring Festival, workers in need, and sports activity Membership subscriptions 0.5% of worker’s wage paid to union at workplace (normally this is not deducted) Revenue from enterprises owned by the local/provincial ACFTU body e.g. cinemas, culture palaces and travel agencies Local government sometimes contribute to specific activities e.g. workers in need Note and Source: To the best of our knowledge these details have never previously been set out. We are indebted to officials (see Exhibit 3 in Appendix) of Haikou Trade Union Federation and ACFTU in Beijing for their help and patience in our detective work.
representation the union is providing in return for the 0.5% deduction: ‘‘they might expect the trade union to work more effectively on their behalf’’. Given the textbook tax incidence issues it is interesting that such worries are admitted over an individual 0.5% levy but not over the firm-level 2% levy. Unions also generate income from quasi-entrepreneurial activities – owning cinemas and other property, cultural palaces and travel agents. For example, in Haikou City most activities on one street, Jiefang Lu, are owned or organised by the local ACFTU branch. Finally, local government sometimes contributes to the welfare role of trade unions – ‘‘workers in need’’ for example – so that local union branches can fulfil their traditional social role.
3. WHAT UNIONS DO NOW 3.1. Evolution In the old Soviet Union trade unions had two functions at enterprise level – ‘‘to serve the enterprise administration through enforcing workplace
238
DAVID METCALF AND JIANWEI LI
discipline, campaigning for improved productivity and administering social services provided by the enterprise, and acting as the ‘transmission belt’ for the party. The unions were not designed to represent workers’ collective interests and workers did not exert any pressure on them to do so’’ (Ashwin, 2003). Prior to the product market reforms in the 1980s unions had a similar function in China. But since the reform process started unions have also – as we saw above – been charged with representing workers. Now the ACFTU is supposed to be both an instrument of the state intended to promote the collective good of society and a labour organisation representing workers’ rights and interests (Chen, 2003). There is an obvious tension between these dual roles, which are probably irreconcilable. And now that the state is withdrawing from its paternalistic role in looking after workers (which Chen states caused the grass roots union to be ‘‘irrelevant’’) this tension is aggravated because the state ‘‘is sacrificing workers’ interests for the sake of restructuring the economic system’’. Manifestations of the sharpened tension between the dual roles include: smashing the iron rice bowl for SOE/collective workers; in the reformed system managers have power over labour such that workers cannot enforce the rights that they have; exploitative practices by foreign-invested enterprises concerning, e.g., safety and overtime. Essentially employees would like more representation, but the ACFTU is incapable of fully providing it. Consider two examples (Taylor et al., 2003). First, members can elect a local leader only if she/he is approved by the hierarchy, and members have no say in the appointment of higher officials. Thus, the trade union is an administrative agency of the party rather than a bargaining representative. Second, by Article 4 of the Trade Union Law 2001 the ACFTU must ensure that its policies and activities are consistent with the CC-CCP at national, regional and local level. Thus, the ACFTU is never allowed to oppose Party discipline or its ideological line. Essentially, as the market economy has developed trade unions have become an ‘antishock valve’ expected to function for both workers and management in the firm, and workers and the Party in society. The 1992 Trade Union Law (amended in 2001) and the 1994 Labour Law recognised that, as the state abandons its paternal labour relations role, the vacuum needs to be filled by trade unions. The ostensible aim of the 2001 law was to transform unions from their transmission belt role to that of representatives of workers. While this is formally set out in these laws, in practice unions remain castrated. They cannot represent workers if such representation conflicts with state policy, and unions cannot initiate or support any collective action. As in the old Soviet Union, the regime provides no legal framework for any kind of collective action or independent
Chinese Unions: An Alice in Wonderland Dream World
239
organisation: ‘‘What unions are expected to do is to placate discontented workers and prevent or defuse any confrontational labour action’’ (Chan, 2000), while simultaneously preventing any emergence of independent trade unions because the state would not relish the prospect of a Polish Solidarity union springing up. 3.2. Present Functions of ACFTU and its Grass Roots Branches Formally, the ACFTU now has three functions: to pre-empt independent organisations; to represent workers as individuals and sometimes, ostensibly, collectively; and to mediate between the workers and the firm in order to nip in the bud any possible collective action. These will be considered in turn. The ACFTU is a monopoly organisation such that, by law, no union is allowed to exist outside the ACFTU structure. This contrasts with the new Russia where there has been ‘‘a lifting of restrictions on the formation of independent organisations and institutions’’ (Ashwin, 2003). As Chan states: ‘‘The ACFTU is just an extended state power designed to place industrial workers under control and pre-empt any alternative labour organisations’’. As such the ACFTU monopoly violates Article 87 of the ILO Labour Convention, which guarantees workers rights to freely form and join organisations of their own choosing. There are many cases of fleeting attempts to form substitute organisations (see e.g. Howell, 1997, 2003) including the 1989 Beijing Workers Autonomous Federation, the migrant workers of Guangdong 1995 and 2002 and the Beijing taxi drivers 1998. Such (albeit short-lived) green shoots explains why the ACFTU pressed for, and got (under the 2001 revision of Trade Union Law 1992), an extension of union recognition. Where requested by the workers a union must now be established. The intention was to spread union presence across many more enterprises, especially in foreign-owned and private enterprises. As shown in Section 2, if the numbers are to be believed, this revision to the law was remarkably successful such that the ACFTU hierarchy now argue that the security of the Party-state is no longer threatened by alternative labour organisations. The Party-state desires such an outcome because it both fears enterprise-level or national ‘‘solidarity’’ and because it wishes to control labour in order to boost returns to capital to further strengthen the economic development of China. Nevertheless, the possible emergence of independent unionism remains one of the three key future challenges for the ACFTU (Section 5). Its success in forestalling such alternatives will largely turn on whether or not it is able to fulfil its representation and mediating functions.
240
DAVID METCALF AND JIANWEI LI
Although Chinese labour law now provides a right to membership and recognition, in practice – particularly in foreign-invested firms – real respresentation is virtually non-existent. Employees recognise this: successive surveys of workers yield an incredibly low fraction who would turn to their union for help with an employment problem or who think the union is doing a good job in their workplace (see e.g. Chen & Lu, 2000 and Yao & Guo, 2004). It is plausible that: ‘‘I really must get a thinner pencil. I can’t manage this one a bit; it writes all manner of things I don’t intend’’, The King (TLG, 1998, p. 131)
and that the inadequate implementation of the laws on representation will, in due course, threaten the stability of employee relations and therefore the Party-state. The Party-state has therefore recently introduced – a clever ploy? – a new law to encourage collective contracts and new tripartite mediation and arbitration institutions. Again the legal and institutional fac- ade is different to the reality: ‘‘That’s the reason the horse has anklets round its feet’’. ‘‘But what are they for?’’ Alice asked in a tone of great curiosity. ‘‘To guard against the bites of sharks’’ the Knight replied. (TLG, 1998, p. 208)
Just as a horse is unlikely to be bitten by a shark, management is – as we showed in Section 2 – little troubled by collective contracts whose implementation is voluntary and which, anyway, lack the substance of a contract, which is the outcome of proper collective bargaining. As Clarke et al. (2004) put it: ‘‘the negotiation of the collective contract is still a very formalistic procedure, with the collective contract only formulating the terms and conditions of employment in the most general terms and providing workers with few or no benefits not already prescribed by laws and regulations’’. And the tripartite mediation committees only deal with some 50,000 cases a year or perhaps 1-in-2,000 employees whereas the corresponding UK figure is 1-in-250 employees (see ACFTU, 2002). Again there seems little to concern management with the establishment of these new institutions. Indeed ACFTU officials describe the tripartite system as ‘‘preliminary’’ and ‘‘feeble’’ in its protection of workers rights and interests (Qiao et al., 2004). This might be thought of as rather an elegant manoeuvre by the Partystate. Unions have, at best, a modest representation role – certainly outside SOEs – so the state established new laws and new institutions to fill the vacuum. Both the Ministry of Labour and the ACFTU wish to be the
Chinese Unions: An Alice in Wonderland Dream World
241
primary body responsible for the regulation of labour relations. This recent emphasis on mediation and arbitration institutions reflects the ascendancy of the Ministry of Labour, responsible for these bodies, over the ACFTU. In fact neither the new laws, nor the new institutions do much to enhance representation yet they successfully displace union activity. Paradoxically the new laws and institutions stem from what unions claim as a success (Case 7), namely their involvement in tripartite (state, employers, unions) institutions at national level. The 2001 law provides more ‘rights’ for trade unions. These include (Articles 19–34): right to participate in democratic management on behalf of workers; right to collectively consult and to sign collective contracts with employers; and right to protect workers right of employment, remuneration, and occupational health and safety. New institutions include enterprise mediation committees and arbitration committees at city or provincial level. These deal almost entirely with disputes between an individual employee and management. It should be emphasised that the system of tripartism in these institutions is odd to western eyes. In enterprise mediation committees the trade union mediates between the employer and the worker (with the state in the background but normally, implicitly or explicitly, siding with the employer). In outsideenterprise arbitration committees the normal three parties are represented – unions, Labour Bureau, employers – but the union official will not necessarily side with the workers because she/he may be a manager him/herself. Unions’ representation role has developed more for individual workers (particularly in the mediation committees) than for collective action. Thus in SOEs the union remains a department of management, facilitating redeployment rather than preventing layoffs. For example, in Haikou when 80% of SOE workers were made redundant in the mid-1990s (e.g. rubber, printing, dyeing and tyres) unions were – as required by national and local regulations – consulted and had some modest success in ratcheting-up the (one-off) redundancy lump sum payment. But unions role is ‘‘to protect the interest of the whole society’’ [weihu quanshehui liyi] (Case 4) such that they essentially play a validating role – to confirm management decisions. Further, the recent slowdown in downsizing in SOEs [caiyuan] is attributable to concerns about unemployment and the burden on the social security system rather than a consequence of union pressure.
242
DAVID METCALF AND JIANWEI LI
Unions retain their welfare role. And they now sign collective contracts, but there is little or no collective bargaining or negotiation involved. Indeed, ‘‘most collective agreements consist of nothing more than a promise by management to pay the legal minimum wage and obey other conditions set by labor laws along with a union’s commitment to help boost productivity’’ (ACILS, 2004). In most privately owned new establishments unions are weak or non-existent. Collective contracts in such enterprises are the outcome of a ritual, formal process controlled by the employer. At the Labour Federation (Case 4) we were told that Provincial regulations require collective contracts but that such contracts have ‘‘no actual functions’’ [meiyou shiji zuoyong]. Instead they are simply ‘‘themes’’ or – case 7 – ‘‘principles’’ [yuanze] following minimum standards laid down by labour law. Further it was specifically stated the ‘‘unions have no role’’ [gonghui qi budao renhe zuoyong] in negotiating such contracts. Rather the collective contract provides a fig leaf of ‘‘consultation and equality hence there is no need for strikes’’ [meiyou bagong de biyao]. Anyway the union Federation (Case 5) stated that most enterprises do not even bother with such contracts. The Federation has some 500 grass roots branches in Haikou City (one quarter the number in the corresponding employers association), but fewer than 40% have collective contracts. Although it is the role of the Federation to spread such contracts, union chairs in enterprises do not wish to upset management by pushing too hard, particularly in newer enterprises where almost none have been signed. Indeed, it is likely that the union official was exaggerating his small success because the Employers Confederation (Case 6) told us that not one of their 2,000 members had a collective contract! Unions are more visible when dealing with individual disputes. Under the so-called tripartite system, disputes are initially processed at enterprise-level mediation committees. If an impasse remains the worker(s) can take their grievance to the higher-level (county, city or provincial) Labour Arbitration Committee. These arbitration committees have representatives from the state, employers and unions. In 2004, for example, there were 55,000 arbitrated disputes, of which only 6700 were collective. Individual cases successfully resolved involved safety, retirement (e.g. because of disability) and non-payment of wages. Collective disputes are normally over the non-signing or ending of collective contracts, but mostly concern procedural matters where the employer does not abide by standard regulations. Sometimes substantive issues are also brought to these committees – like late payment of wages or non-payment of social security contributions – but unions only take up collective disputes where ‘‘management is patently in the wrong’’ and which are ‘‘absolutely winnable’’ (like withholding pay or pension contributions),
Chinese Unions: An Alice in Wonderland Dream World
243
and avoid ‘‘complicated ones’’ (Chen, 2003). This institutional fac- ade is quite impressive, but the incidence of such cases is modest despite workplace unions being unable to do much for workers at enterprise level. The representation role of trade unions turns on their institutional status within the state apparatus, but not the ability to orchestrate collective action. At workplace level unions remain mostly ineffective and subordinate to management. It is plausible that the present laws and institutions – mostly a hollow shell – may need to be replaced or supplemented with new laws from a ‘‘thinner pencil’’ or the consequent lack of representation may spill over into real instability. The right to strike was revoked in 1982, because strikes were held to be detrimental to ‘‘stability’’ and ‘‘production’’. Nevertheless spontaneous labour-related demonstrations do occur, often involving many workers. The CCP refused a request from the ACFTU to legalise some strikes because: ‘‘legalisation of strikes could only induce more strikes’’. Instead the duty of the union is ‘‘together with enterprise administrations, to resolve through consultation, reasonable and resolvable demands from workers, and incidents, and restore production as soon as possible’’ (ACILS, 2004). Thus, unions’ priority is to defuse protest rather than to represent demonstrating workers: ‘‘persuading workers to withdraw from the streets should be the unions’ ultimate goal in a protest incident’’ (Chen, 2003). 3.3. Efficiency and Equity In Europe and the US there has been a depth of careful empirical work concerning the impact of trade unions on efficiency – the performance of the firm – and equity (see e.g. Addison & Schnabel, 2003). In China, trade unions cannot exert pressure on the firm to raise pay and achieve a union mark-up. Similarly they have few instruments to either lower productivity, e.g. restrictive practices, or raise it, e.g. by encouraging greater investment in physical and human capital. There is one possible exception. Trade unions have an official role in encouraging skill training and innovation, via problem-solving teams for example (see Cooke, 2005). But neither the literature nor our case study evidence suggest this is a central union activity. Thus, Chinese unions are mainly nugatory at influencing labour costs and therefore firm performance. But there may be some association between a union presence and fairness at work indicated by, for example, inequality in pay, coverage of various insurance benefits, labour regulations and policies towards women and older workers. The extent of pay inequality depends on pay levels across enterprises and workplaces and within them. Since the smashing of the iron rice pot (Leung,
244
DAVID METCALF AND JIANWEI LI
1988) pay inequality has increased (though it is below that in the US and UK, see Metcalf, Hansen, & Charlwood, 2001). First, pay setting is now decentralised to establishment or workplace level and enterprises have different approaches to market-based and performance-related pay. Second, within the enterprise the traditional low wage/high welfare system has been replaced by a huge variety of payment systems including: piece rates [jijian gongzi zhi], bonus system [jiangjin zhi], the ‘structural wage system’ [jiegou gongzi zhi], the ‘floating wage system’ [fudong gongzi zhi], and ‘post plus skills wage system’ [gangji gongzi zhi] (see e.g. Warner & Zhu, 2000). The presence of a trade union in the workplace has little influence on the extent of wage inequality. While it is true that wage inequality is normally lower in SOEs than in foreign-invested firms and that the incidence of union branches is also higher in SOEs than elsewhere in the economy, it would be wrong to conclude that (unlike in the UK for example) it is the trade unions that cause the wage inequality to be lower (see also Cooke, 2005; Khor & Pencavel, 2005). The extent of wage inequality depends almost entirely on choices by management. But unions may play some modest role at the margin. For example, Leung (1988) described how when the new more market-based pay arrangements were introduced, unions in some workplaces were able to build in some safeguards-particularly to protect the old to ensure that they did not suffer a drop in living standards. While unions have little affect on pay levels, they do influence benefits. Consider Table 8, drawn from a representative sample of over 3,000 private sector enterprises in 2002. The fraction of workers in unionised enterprises covered by medical, old-age and unemployment insurance is half as much again as the fraction in non-union enterprises. Given the overall low levels of such insurance in China, this greater likelihood of cover in unionised enterprises represents a tangible benefit, albeit available only to a small minority. Under the 1994 Labour Law, trade unions are charged with monitoring the implementation of workplace health and safety regulations (see Benson et al., 2000). These include rules concerning safety, various regulations ostensibly governing hours of work, ensuring the minimum wage is paid and overtime hours are properly compensated. It is possible that accident rates are lower and breaches of other regulations fewer where there is an ACFTU branch compared with similar non-union workplaces (unfortunately no such data exists). But the appalling safety record and manifold violation of wage and hour regulations in unionised workplaces again suggest nugatory unions. Instead, unions’ major workplace role concerns workers’ welfare. Before the market reforms ‘‘trade unions were responsible for the administration of a large part of the welfare policy of the Party-state’’ (Clarke, 2005).
Chinese Unions: An Alice in Wonderland Dream World
Table 8.
245
Medical, Old-age and Unemployment Insurance Coverage in Private Enterprises 2002. Union Enterprises Non-Union Enterprises All Enterprises
Medical insurance fee per worker per year (yuan) coverage (%) Old-age insurance fee per worker per year (yuan) coverage (%) Unemployment insurance fee per worker per year (yuan) coverage (%)
468 18
515 12
480 16
1064 25
1109 16
1075 22
304 10
675 6
393 8
Source: Research group on private enterprises in China (2004) Report on Private Enterprises in China, 14 March. The data are from a representative sample of 3,258 private enterprises. The research group on private enterprises is a joint research group of the All China Industry and Business Association and the Research Institute on China Private Enterprises. See www.ebg.org.cn
For example trade unions administered any sick pay which ‘‘involved visiting the sick and weeding out malingerers’’. Unions were also centrally involved in the allocation of housing, nursery places and vouchers for subsidised vacations; the organisation of summer camps for children, cultural and sport events; and provision of financial assistance to work colleagues who had fallen on hard times. Presently, trade unions retain some of these functions: they are no longer involved in housing allocations but – as two of our case study firms said in a dismissive way – they visit the sick and organise the sports day and the new year feast. For the future unions could usefully provide a voice for and represent rural migrant workers who are clearly at risk of exploitation (see Cooke, 2006), but unions’ workplacebased organisation inhibits any proper representation of agency workers or live-in maids.
4. CASE STUDIES 4.1. Hainan Special Economic Zone and the Case Study Organisations Haikou City is the capital of Hainan Province, China’s biggest special economic zone (SEZ). It is the political, economic and cultural centre of the province. It covers 2,300 square kilometres, with a population of over 1.6 m
246
DAVID METCALF AND JIANWEI LI
and has a tropical climate. Haikou’s development strategy is to build the city into a base for high-tech industries, a tropical seaside tourist resort, and a regional commerce and trade centre sustaining its agricultural base. In 2003, its gross domestic product (GDP) was 25 billion RMB. The annual per capita income is over 14,000 RMB (around £1,000). In the 1990s Haikou SOE employment, then some 35,000, was cut drastically. Presently there are 50,000 workplaces (of all sizes) of whom 2,000 are in the employment confederation and 500 in the union federation. Official employment is 250,000, but in fact it is much higher, boosted by migrants, family workers, etc. The companies/workplaces included in the case study were chosen carefully, balancing intensity of investigation – suggesting fewer cases – with attention to alternative forms of workplace governance – requiring more cases. Given Hainan’s tropical agricultural heritage and (by Chinese standards) long-standing status as China’s largest SEZ, we deliberately chose to focus attention on food products and on three (non-SOE) different types of governance. It must be recognised that Hainan is somewhat atypical: it always had less heavy manufacturing and lower SOE employment than many other Provinces. But, this is what makes it interesting. Details of the three case study organisations are summarised in Table 9. The three case organisations were visited in August 2004 and April 2005. Free access to each workplace was granted and discussions were held with management, unions and workers. Each visit lasted 1–2 days including getting our hands dirty (literally) on the shop floor. Fuller information about the organisations is given in Appendix B. The Haikou Agriculture, Industry and Trade [Luoniushan] Co. Ltd., is an ex-SOE that became a company listed on the Chinese stock exchange in 1992. Its major activity is pig farming but it is also involved in growing fruit and vegetables, egg production, instant coffee, toothpaste manufacture and operating two private schools in Haikou. It is only located on Hainan Island, with over 2,000 workers and sales of nearly 400 m RMB ($50 m) a year. Originally an SOE, it became a listed shareholder-owned company in 1992. The original pig farm component of the listed company still owns 5% of its shares and these will be sold to the employees in 2006. The Hainan Haiwoo Tinplate Co. Ltd is a joint venture among China, Korea and Japan. The original 2,000 investment shares have altered such that in 2005 the ownership is China (0.4), Japan (0.4) and Korea (0.2). The workplace has some 250 workers with an annual output of 100,000 tons of tinplate. The Holding Company has workplaces all over Asia. The Coconut Palm Company is the largest producer of coconut juice in the world, with 6,000 workers. It is vertically integrated because it manufactures much of its own
Chinese Unions: An Alice in Wonderland Dream World
Table 9.
247
Case Study Organisations: Governance, Voice and Payment Systems. Haikou Agriculture, Industry and Trade (Luoniushan) Co. Ltd
Hainan Haiwoo Tinplate Industry Co. Ltd
Coconut Palm Co.
Product
Major: pig farming Minor: operating two private schools; toothpaste; vegetables, fruit, eggs; instant coffee Only located in Hainan; total 2,100+ workers; 380 m yuan sales pa
Tinplate 242 workers in Haikou 100,000 tons highquality tinplate pa Other workplaces all over Asia
Major: coconut juice and by-products Minor: other fruit juice 6,000 workers
Governance
Ex-SOE, became listed share-owned company in 1992 (original pig farm still owns 5% of shares, to be sold to employees in 2005/ 2006)
Joint venture – China (30%), Japan (30%), Korea (40%) on stream in Hainan in 2000. 2005 respective shares are China 40%, Japan 40%, Korea 20% Haiwoo tinplate has many workplaces all over Asia
Originally 18 workplaces in Hainan SOE 2-producing 80% of output – entirely employer-owned now. Remaining 16 will become e’ee owned shortly
Voice and Dispute Resolution
Union: yes (2% paid) but Workers Congress is key discussion forum at company and workplace level Individual disputes only; informal resolution or worker leaves Chair of union is old fashioned transmission belt. He is Deputy General Manager (approx no. 5 in Co.)
Union: yes ‘‘but only for show’’, ‘‘irrelevant’’, ‘‘just do sports and entertainment’’, ‘‘will soon fade away’’ (2% paid) Workers Congress: no Voice is via frequent direct meetings between management and workers Individual disputes only. If necessary go to mediation and arbitration outside the company
Shareholder committee filters/ approves all mg decisions, votes by shares held Workers congress, to ‘‘validate’’ mg decisions Union 2% paid but ‘‘just welfare role’’
248
DAVID METCALF AND JIANWEI LI
Table 9. (Continued )
Pay Systems and Collective Contracts
Haikou Agriculture, Industry and Trade (Luoniushan) Co. Ltd
Hainan Haiwoo Tinplate Industry Co. Ltd
Coconut Palm Co.
Pig farm by group: target output by weight, if>target extra pay, ifotarget penalised. Also revenue component with floor (to protect workers) if pig prices drop sharply Collective contract (cc) when SOE. Now national cc based on Haikou ‘‘model contract’’ but implemented on individual basis, renewed every 5 years
Efficiency wages because pay based on Korean rates Production workers pay: 40% base by responsibility, skill, intensity 60% ‘‘floating’’ by production volume Individual contracts signed annually PA, workers ranked, worst 1 (or more) automatically dismissed
Intensive PRP Base, by occupation skill etc. – monthly PA Bonus, performance relative to sales target Dividend, by shares held, co performance Team monitored by next team on assembly line with big potential penalties When introduced in 1990s production manager said Inferior quality quitsorting Extra effort by employees and recruits No collective contracts
tinplate and packaging material. It was originally an SOE with 18 workplaces, solely on Hainan Island. After severe economic problems in the 1980s it switched, for the two largest workplaces accounting for 80% of output, to be an employee share-owned company in the 1990s. The remaining, smaller, 16 workplaces will also become employee-owned shortly. 4.2. Unions, Voice and Dispute Resolution All three workplaces have a trade union and pay the requisite 2% payroll levy over to the relevant union authorities, but provision of voice and dispute resolution is not a major feature of union activity. In the Pig Farm conglomerate the role of the workers’ congress is emphasised. The congress meets twice a year, separately in each workplace then at company level. In 2004, matters discussed included the 2003 annual report, targets for future
Chinese Unions: An Alice in Wonderland Dream World
249
output and housing. It is essentially top-down consultation because on matters like the layout of production or overtime arrangements the trade union chair emphasised that management decides unilaterally. The trade union chair is in fact a very senior manager and acts in the traditional transmission belt role. In the Tinplate company voice arrangements are entirely via frequent direct meetings – both ad hoc and formal – between management and workers [zhi jie]. The union chair is the top sales manager. We were told that the union ‘‘is only for show y irrelevant y just organises sport and entertainment y and will soon fade away’’ [shi baishe y wuguan y zhi zuzhi wenti huodong y bujiu jiang xiaoshi]. In the employee-owned Coconut Palm Company it is the shareholders committee that filters management decisions, which are then validated by the workers congress – a hangover from the company’s SOE days. The union ‘‘just plays a welfare role’’. The ACFTU grass roots branches in our three case study organisations are equally impotent in dispute resolution. The idea of a collective dispute was greeted by all parties with incredulity in each company. Individual disputes are settled informally at Pig Farm where ‘‘management will ask those who want to, to do the work’’. At Tinplate company individual disputes – including dismissals – are initially dealt with informally and subsequently could be heard by the formal in-firm mediation committee or Haikou (town level) arbitration committee. Such cases are very rare and the company won each case because it ‘‘carefully followed national and local government regulations’’. At the Coconut Palm company there were, it was said, no disputes because ‘‘all matters of pay and worktime etc., are in accordance with Chinese labour law, therefore there is no need for conflict’’. Further, because union officials are appointed by the state, they ‘‘work well with management’’ and are ‘‘unlikely to have disputes’’. 4.3. Collective Contracts and Payment Systems In the west, a major role (previously the major role) for unions is to negotiate collective contracts and to be involved in the detail of payment systems. By contrast the unions in our three cases played, at best, a peripheral role in such matters. Chinese labour law now encourages unions to sign and perhaps bargain collectively to negotiate collective contracts. Two of our case companies (Pig Farm and Coconut Palm) stated that they do have such collective contracts, but not as a result of negotiation with the trade union grass roots branch. Rather, the contracts set out the minimum standards in matters like
250
DAVID METCALF AND JIANWEI LI
minimum wages, working time and safety as required by national and local law. Such collective contracts were referred to derisorily by management at Coconut Palm as simply setting out ‘‘themes’’ but having no real content. Tinplate simply dismissed the notion of collective contracts. Instead, all three companies emphasised the importance of individual contracts. Pig Farm uses the model individual contract issued by the Haikou Labour Bureau, but with its own amendments. This contract is re-signed every 5 years. In Tinplate company the individual contract is for one year. We were told that this short duration is a deliberate stick to elicit effort, given the efficiency wages paid: ‘‘we could provide 3- or 2-year contracts but the 1-year contract yields more discipline’’. The reason that individual contracts are paramount is not hard to find. All three of our case companies have powerful versions of performancerelated pay. Our evidence mirrors the remarkable changes which have occurred in under a decade away from traditional work unit [danwei] – with its egalitarian, non-performance related pay system – towards company, team and individual performance pay with a low-base component (see e.g. Warner & Zhu 2000). Again, in our three companies, these systems were devised and implemented by management with no real input from unions or workers. Pig Farm sets team-level production (weight) targets with severe penalties for missing the target and generous rewards for exceeding it. There is also a risk-sharing revenue-based element in the pay system, but with a wage floor in case pig prices fall rapidly. It was emphasised by the union chair that the workers ‘‘only get paid if they perform’’. Tinplate Company has a company-wide performance-based system such that, on average, production workers’ pay is composed of a base amount of 40% and an output-related amount of 60%. For administration workers the fractions are 50/50, but the performance element is related to sales, causing wages to fall when the price of imported tin rises and sales fall as in 2004/ 2005. The deputy general manager who we interviewed stated that his notional pay should be 4,600 RMB per month, but because of falling sales it is only 3,000 RMB. More importantly, there is an ‘‘inverse tournament’’. Each worker is given a rigorous annual performance appraisal and the worker (or perhaps a few workers) with the lowest rating is automatically dismissed: ‘‘last performer, first out’’ [mowei taotai zhi]. At Coconut Palm the performance pay system uses salary, bonus, dividends and fines. The salary is based on a monthly individual performance appraisal. The bonus is related to company-wide sales. Quality control is achieved by teams (perhaps of only two workers) monitoring the output
Chinese Unions: An Alice in Wonderland Dream World
251
quality of the immediately previous team on the assembly line. A defective can would, for example, result in a fine of more than one day’s pay. Managers and union officials emphasised that when this system was introduced in the 1990s many workers left because they could not cope with the risk sharing and extra effort required-a nice example of the familiar sorting and effort effects of the introduction of performance pay (Lazear, 1998). The traditional wage system was introduced in 1956 and consisted of ‘‘low salary, high social welfare and high rate of employment’’ (Qiao et al., 2004). This system was in place until the mid 1980s. The evidence in these case studies shows how rapidly it has been inverted such that the norm is now performance related pay, low social welfare and insecure employment. 4.4. Miscellaneous Issues The literature (e.g. Taylor et al., 2003) suggests that there are three further issues which trouble workers and their representatives: delayed wage payments, redundancies and safety. Our three case companies were well run so – unlike elsewhere in China – there was no problem of delayed wage payments. Layoffs or redundancy (i.e. collective not individual separations) are now accepted as ‘‘normal’’ in the market economy. Pig Farm recently closed down its complete cattle operation because it was bankrupt and this closure was approved by the Provincial government and discussed with the Workers Congress whose representatives were shown the books. An attempt was made to relocate as many workers as possible to other parts of the conglomerate. Some modest compensation for those not transferred was agreed with worker representatives. Early in its operation Tinplate Company had to make 30 workers redundant because of a drop in sales. These separations were speedily agreed by the Haikou Labour Bureau. Coconut Palm has 6,000 workers, 4,000 on standard contracts and 2,000 temporary workers who mostly cut the coconuts from the trees and extract the flesh. These 2,000 provide a buffer labour force to protect the 4,000 from redundancy. Safety rules and monitoring were not seen as part of the union function in any of the three cases. For example, at Pig Farm, government regulations are detailed and rigorously enforced, and throughout the conglomerate the importance of educating workers – often through slogans and vivid posters – about safety was emphasised. Tinplate management set safety rules and enforced them via worker involvement. These three case studies provide concrete evidence of the impotence of worker representatives in general, and trade unions in particular, in the
252
DAVID METCALF AND JIANWEI LI
Chinese market economy. Voice comes either through the ritualistic representative workers congress, or directly via meetings between management and workers, but not via the grass roots trade union. Individual disputes are settled informally or via the in-firm mediation committee – often chaired by the most senior union official – or out of firm arbitration committee. But the union seldom represents the worker at such hearings. Collective contracts have no real content or go unsigned. Individual contracts abide by local and national regulations but are essentially a device to elicit effort and the union is not involved in negotiating such contracts on behalf of individuals or groups. Powerful performance-related pay systems operate in each organisation, again emphasising the individual and the group and marginalising any union involvement in negotiating or implementing such systems. And even on matters of safety and redundancy – key protective functions – unions are remote.
5. FUTURE OF ACFTU AND ITS GRASS ROOTS BRANCHES In recent years the ACFTU and its grass root branches have begun to adapt to the market economy. For example, the ACFTU has established new departments dealing with legal matters, labour relations issues that arise in foreign-invested enterprises, and the tripartite system of mediation and arbitration committees. It has also been centrally involved in the drafting of government labour laws and regulations including the 1994 law requiring the establishment of trade unions in all enterprises, and the 2001 law covering the right to organise. Thus, it has changed its internal structure, its operational style and the legislative and regulatory framework in which it operates. But – as Howell (2003) notes – its fundamental dilemma remains: is it a representative and participative organisation or, instead, an ‘‘integral cog of the party machine’’? The nature of this dilemma can be illustrated by setting out the way the ACFTU interacts with other parties and by discussing the three major challenges faced by ACFTU if it is to secure its future. The ACFTU interacts with the Party-state, employers, individual employees and its constituent grass roots branches. Most such interactions develop pragmatically – ‘‘crossing the river by feeling the stones’’ [mozhe shitou guohe]. The major challenges are as follows (Howell, 2003). First, it has to cope with the altered composition of enterprises and workers. Second, it has to decide where its priorities lie: is it primarily an agent of the
Chinese Unions: An Alice in Wonderland Dream World
253
Party-state or is its job to represent workers? Third, as a monopoly organisation can it maintain its legitimacy or will there continue to be spontaneous worker protests and attempts to form new, alternative, labour organisations? These three challenges are compounded by China’s entry into the World Trade Organisation. 5.1. Interactions between ACFTU and Other Parties The future of the union movement in China is bound up with the way in which it interacts with other employee relations players: the Party-state, employers, individual employees and the way the ACFTU treats its constituent grass roots. The Party-state is dominant. Given that the basic union unit is the workplace or enterprise, a cohesive union movement is not possible. With economic decentralisation unions are sidelined in SOEs and ex-SOEs and often irrelevant or missing in JVs or multi-nationals. This weakens the union command structure. Even though the ACFTU has been granted an organising monopoly by the state, plus the ability to sign collective contracts, it still cannot safeguard freedom of association, the right to collective bargaining or collective industrial action. In consequence, China is in breach of various ILO conventions. The ACFTU is wholly incorporated. Therefore the question for the future (Howell, 2003) is whether or not it can, instead, inch towards becoming a state-sponsored corporatist institution. In SOEs and ex-SOEs, unions are normally run by management with just worthy – but modest – tasks like monitoring working conditions, organising the feast at the time of the Spring Festival and visiting the sick. Even though unions now have a right of recognition in JVs and multinationals this often means little in practice – as we show below in the Wal-Mart case. The question here is whether, particularly where foreign capital is at stake, the Party-state will ever permit representation of workers to take precedence over management prerogatives. On present evidence this looks unlikely. Rather, any trade union presence in foreign owned companies or JVs is more likely to be desired because, in the absence of a Party organisation, the Party-state can keep an eye on foreign management through the trade union. In order to thrive unions must deliver for workers. Where unions are recognised there is little evidence that unions have achieved more generous severance terms when SOEs retrench. And in foreign-invested firms instances of unpaid overtime, late pay and child labour remain rife. Both the Party-state and the union movement are now placing more emphasis on
254
DAVID METCALF AND JIANWEI LI
helping individual employees via tripartite consultation. But the framework of new impressive institutions is not all it is cracked up to be. First, when the state is represented the Ministry of Employment and Social Security normally takes the lead role, but this Ministry is often not the key player in employment matters that may turn more on finance and international relations. Second, representatives of the enterprise are not split between members of the Chinese Employers Confederation (CEC) and ACFTU. Instead of such a pluralist approach it is simply asserted that there is a unity of interest between workers and management. Third, tripartite local issues under the mediation and arbitration bodies almost all deal with individual disputes and seldom extending/signing new contracts. Finally, tripartite consultation over national issues – the new labour laws – yield fancy sounding outcomes – collective contracts for example – but are essentially candy floss. And transformation of trade unions also requires more support for those workers presently unorganised: the bulk of non-state workers, temporary contract workers, migrant workers and the unemployed. Sadly most workers do not think that the union is the body to enhance their worklife or to protect them. They are right to be sceptical: unless unions start to deliver for workers, on both an individual and collective basis – they remain nugatory. Further, it is also clear that so-called democratic centralism whereby individual unions are incorporated into the ACFTU structure via its topdown system is not much of a model for the future. The grass roots trade unions are simply administrative bodies handling welfare and ideology and have almost no representative role. This deficiency is compounded by the financial insecurity detailed in Section 2. Presently, the union movement is subservient to the Party-state, seemingly a sideshow for most employers and unable to deliver requisite pay and due process benefits for workers. In addition the top-down union structure and financial insecurity of the movement do not suggest that the challenges for the future – discussed next – are likely to be overcome. 5.2. Challenges 5.2.1. Composition of Enterprises and Workers In the pre-reform period there was a strict hierarchy (Howell, 2003). At the top were workers in state owned enterprises who enjoyed the iron rice bowl of lifetime employment, welfare benefits, subsidised housing and high political status. Next came workers in collectively owned enterprises and finally those in street or neighbourhood committee enterprises. But in the post
Chinese Unions: An Alice in Wonderland Dream World
255
reform period enterprises and workers are more heterogeneous. As we saw in the case studies governance types now include foreign-owned, JVs between Chinese and foreign investors, privatised companies, state owned companies, smaller enterprises (especially in the coastal provinces) owned by overseas Chinese but run locally. In addition the traditional urban labour force has been augmented by many millions of migrants from rural areas. Previously, the ACFTU mainly represented urban workers in state owned enterprises – where they were what Taylor et al. (2003) called ‘assistants’ or ‘subsidiary departments’ of management. But, this is a shrinking pool of workers so this ‘assistant’ role is insufficient. It is apparent that the owners and management of many newer enterprises – particularly those that are foreign owned or JVs – are less keen on recognising the ACFTU. Organising in foreign-invested or smaller enterprises is a labour intensive, difficult (i.e. costly) task. The ACFTU is facing similar problems to UK (and other Western) unions who find gaining recognition in new enterprises hard to achieve. In August 2004, the standing committee of the National People’s Congress undertook a nationwide inspection of the implementation of the trade union laws. It is found that under 10% of the 500,000 foreign-funded enterprises registered in China have established trade union organisations. Consider the case of Wal-Mart, the world’s largest supermarket chain, which has 20,000 Chinese employees at 40 stores in 18 large cities across China. While the law states that no organisation or individual has the right to obstruct or prohibit the establishment of a trade union in the workplace, Wal-Mart insisted up-to-late 2004 that ‘‘we are not against the establishment of a trade union but we have not got any such requests from our staff’’. Unsurprisingly, employees interprets things differently. Li Haisheng, a porter, is quoted: ‘‘we are afraid of putting forward such a claim. The employment market is harsh, and no one wants to sacrifice his job for establishing a trade union here’’. Trade union officials, for example, in Shenzhen and Jiangsu, attempted to convene meetings with Wal-Mart, but were brushed aside with the excuse that managers are absent. In late 2004, under heavy pressure from the National People’s Congress and ACFTU, Wal-Mart agreed to the establishment of ACFTU branch unions in its retail stores. It will be interesting to see how effective such grass roots branches are. Uniden Electronics, Kodak, Dell and Samsung are similarly hostile to a union presence in their organisations. Kodak, for example, will not provide any place for union activities and managers are forbidden from engaging in any such activities. And even when the union is recognised and properly representing its workers, activity which threatens managerial fiat is likely to
256
DAVID METCALF AND JIANWEI LI
result in the sack, as at Sanhuan Sagam High-tech Co., a Sino-Japanese joint venture in Beijing which recently dismissed its trade union leader Tang Xiaodong. (see e.g. Washington Post, 26 April 2005): it seems non-nugatory unionism in a workplace is destined to be shortlived. Taken at face value it is surprising that 90% of foreign-funded enterprises have no union. China’s Trade Union Law stipulates two ways for the establishment of trade unions: one is a request from employees on a voluntary basis, and the other is via a request from the trade union at a higher level. Trade unions in the upper level of the national union hierarchy are authorised to send union officials to enterprises and help them establish trade unions. Enterprises have no right to interfere in or obstruct the process. Indeed, in theory, foreign-funded firms may be sued if they persist in their refusal to allow the establishment of trade unions. In practice things are different. Guo Jun, Director of the ACFTU Legislative Affairs Bureau recently stated: ‘‘Some enterprises refuse to set up trade unions, saying that they receive no such request from their employees. They also toss aside the requests from trade unions at higher levels, and resist publicising relevant policies on trade unions to their employees, which is obviously a violation of the Trade Union law’’. The reason nothing is done about this is straightforward: ‘‘Some local governments worry that the requirements on the establishment of trade unions may adversely affect the enthusiasm of foreign investors y [they] stress too much on GDP growth, neglecting the protection of employees’ legal rights and interests, which obviously violates the law. For example, Guangdong has the policy of deferment in the establishment of trade unions for foreign-funded enterprises’’. Unions face an almost impossible task in organising rural workers, temporary workers, those displaced by SOEs, employees in small-scale enterprises and migrants. One large group of such migrants are maids and those working in care homes. In an attempt to regulate and protect these workers a (state) licence is now required to practice these occupations. As in the UK, certification and licensing are playing a role in occupations where unions cannot get a foothold. The ACFTU is putting more emphasis on organising employees in rural areas (Case 7), for example by promoting more stringent workplace inspections, but whether or not this will translate into extra members is a nice question. ACFTU concerns essentially followed those of the Party-state that has also given agriculture and the rural economy higher priority in national economic development. The national ACFTU has also tried to influence policies concerning displaced workers. Their suggestions include strengthening the social security system, expanding re-employment centres and strengthening legal procedures regulating redundancies.
Chinese Unions: An Alice in Wonderland Dream World
257
5.2.2. Priorities Does the ACFTU and its branches exist for the Party-state or for the workers? The ACFTU is not the ‘voluntaristic’ trade union that we are familiar within the west. It is one of three mass organisations (the other two represent youth and women) and its staff are appointed and (partly) paid by the Party. As Howell puts it: ‘‘They function as a ‘transmission belt’ relaying party directives and policy downwards and transmitting grass roots opinion upwards’’. But now that owners and managers control enterprises, rather than the state, this transmission belt role is much less important and the tension between contradictory interests – workers and party – has become more noticeable and intense. Consider two recent examples (Howell, 2003). First, as we saw in the case studies, when SOEs or newer firms downsize the trade union is frequently unable to achieve satisfactory redundancy packages or defend workers against non-payment of pensions. Second, when treating with foreign investors the union is often unable to deal with complaints about excessive overtime and unpaid or delayed wages, partly because the local authority puts economic development above workers’ rights and sides with capital. This leads directly onto declining legitimacy (considered next). It is plausible that one reason that the constituent grass roots branches of ACFTU have downplayed their representation (or servicing) role is that, in addition to their subservient role in the Party-state, the ACFTU branches have insufficient finance and qualified people to, for example, monitor safety legislation and to pursue warranted worker grievances. 5.2.3. Declining Legitimacy The ACFTU is a monopoly organisation whose legitimacy is on the wane because it acts as if everyone has won and refuses to countenance possible conflicts of interest between capital and labour. ‘‘Everybody has won, and all must have prizes’’, Dodo (TLG, 1998, p. 26).
To regain some legitimacy unions will have to develop in two directions. First, new skills are needed including those required to implement the new collective bargaining law, which promotes bargaining and negotiation. Second, the different interests of capital and labour must be recognised, for example, in the retrenchment of SOEs and malpractice by foreign-owned firms. This legitimacy issue arises in part because the top-down structure of the ACFTU fits uneasily with the market economy. Although unions have partially adapted, for example the emphasis on collective contracts at
258
DAVID METCALF AND JIANWEI LI
enterprise level, such modernisation has been insufficient to suppress conflict and the formation of alternative representative institutions. Our contact at the Haikou Trade Union Federation (Case 5) captured the legitimacy problem nicely. He said that ‘‘the text book answer’’ [jiaokeshu da’an] is that the Party-state represents the whole people. But if there is a conflict between the whole society and sectional interests (e.g. over the need to foster foreign direct investment) the union almost always supports the interests of the wider society. He said this was the ‘‘first priority’’ [zhong zhong zhi zhong] of the union and required by law. Similarly at ACFTU headquarters in Beijing (Case 7), it was stated that ‘‘local government has only one target-economic growth y other interests are sacrificed for the sake of economic growth. Labour legislation and its implementation and the role of unions have all been constrained to serve the target of economic development’’. Further, the ‘‘stability of the state’’ is the paramount government policy and it is the unions role ‘‘to promote such stability’’ y and ‘‘the government stand on the boss’s side’’. The Haikou official noted that tension in the workplace between capital and labour had become more severe in the new market economy. As the enterprise union official is paid by, and often part of, management she/he is in ‘‘a difficult situation’’ [chujing jiannan]. The role of the union chair is ‘‘to achieve consensus, manage the workers and reflect views both ways’’ in an iterative process. In all circumstances ‘‘direct resistance is not practical’’ [Zhijie duikang shi bu xianshi de], because ‘‘stability’’ [wending] is what matters and society ‘‘cannot afford riots [strikes]’’ [jingbuqi dongdang]. This lack of legitimacy in the representation function was explained straightforwardly as a ‘‘system’’ [tizhi] problem – the union is subservient to the Party-state. Consequently, if a worker has a problem over say delayed wage payments or compensation for redundancy, the worker’s first port of call is almost always the Labour Bureau (i.e. a state agency) because ‘‘the worker knows the trade union cannot solve this problem’’ [gongren mingbai gonghui jiejue bu liao wenti]. Such representation is, he said (Case 5), not an important part of the Federation’s work. Instead the union focuses on workers in need and encouraging parties to sign collective contracts (recall that these have little substance) although such contracts only ever covered a maximum of two fifths of members. A further source of declining legitimacy is the fact that most chairs of enterprise unions are employees of the company. They risk losing their job – ‘‘fried squids’’ (Case 5) [chao youyu] – if they rupture their normal cosy relationship with management by pursuing workers’ concerns. In 2001, the ACFTU considered addressing this problem by moving away from the
Chinese Unions: An Alice in Wonderland Dream World
259
enterprise as the key organising unit towards a system of craft and industrial unions with full-time organising cadres (Case 7). The ACFTU official stated explicitly: ‘‘to raise the power of unions [the structure needs to change] to one of craft and industrial unions’’. In the event this policy did not find favour with either the Party-state or congress delegates, many of whom expressed worries about ‘‘stability’’ in the workplace. There is a nice parallel with the UK here. The central union body (ACFTU or TUC) suggests some innovative changes in union structure – e.g. industrial or craft unions in China or membership swaps to move more towards an industrial structure in UK – but the constituent union barons feel their noses would be put out of joint by such changes and therefore vote them down. Further, the Chinese authorities may have been worried that a union structure which would no longer be enterprise-based risked spreading discontent across many workplaces rather than confining it to one enterprise. Policies matter as well as structures. ACFTU officials (Case 7) told us that they wish to give violations of labour standards a much higher priority in the future, and that such a switch of emphasis would do much to protect exploited workers, especially those making internationally traded goods. Frankly, pigs might fly before such exploitation is seriously addressed. In recent years, there have been many more spontaneous worker protests including wildcat strikes, sit-ins in foreign owned factories in the special economic zones, blockades of roads and railways and street protests by laidoff SOE workers. In addition, because sufficient workers feel the ACFTU is neglecting its representation role new, alternative, forms of labour organisation emerge. Protests and substitute institutions are normally ruthlessly suppressed. For example, a court in southern China sent five workers to prison for ‘‘destroying property’’ in October 2004. The five were among 40 arrested after thousands of workers went on strike at a Taiwanese-owned shoe factory in protest over wages and working conditions (Guardian, 29 October 2004). And a strike by 10,000 workers in an attempt to set up an independent union at Uniden Electronics in Shenzhen (a Sino-Japanese joint venture) in April 2005 collapsed when the putative leaders were jailed and riot police ensured a return to normal working (Washington Post, 26 April 2005). The state is keen to avoid spontaneous action – what Hobsbawm (1976) called ‘‘collective bargaining by riot’’ – a reflection of inappropriate institutions, fearing that it could herald the emergence of a Polish Solidaritytype independent trade union. All this suggests that the traditional incorporated ACFTU role of conflict suppression is insufficient – it is no longer possible to pretend that the interests of capital and labour are always congruent.
260
DAVID METCALF AND JIANWEI LI
This corroding of the legitimacy of the ACFTU is compounded by its financial insecurity. The ACFTU gets most of its revenue from the 2% payroll levy. In the past, this levy was transferred automatically from SOEs and collective enterprises to the ACFTU. But, these state enterprises are now being replaced by private capital and it is proving much more difficult to enforce payment of the levy from private and foreign-owned firms.
6. CONCLUSION: WHAT FUTURE FOR UNIONS IN THE ALICE IN WONDERLAND DREAM WORLD? Alice: How nice it would be if only we could get through into Looking-glass House! I’m sure it’s got, oh! such beautiful things in it! Let’s pretend there’s a way of getting through into it somehow y (TLG, 1998, p. 127)
In her stunning family history Wild Swans, Jung Chan (1991) notes that during the famine (1959–1961) ‘‘telling fantasies to oneself as well as others, and believing them, was practised to an incredible degree y ‘self-deception while deceiving others’ [zi – qi – qi – ren] gripped the nation’’. If we turn the clock forward almost half a century a similar dream world holds concerning trade unions. For most officials involved in union activity, life is guided by the proverb zhi lu wei ma [point at a deer and call it a horse] – which means saying one thing and doing another is a way of life because the party believes that to do anything else would risk destabilising the system: democratic centralism, the right to belong and so many members, collective contracts (but not collective bargaining), tripartite mediation and arbitration committees and no conflict between capital and labour. Let’s pretend indeed! As the former Beijing bureau chief for the Los Angeles Times put it: ‘‘There is virtually no issue in China more significant and less well covered, than the rights of its workers’’. The plain fact is that Chinese unions are mostly nugatory. Thriving unions need, initially at least, some common interest among workers to flower. This has not developed, and is unlikely to evolve in China for a number of reasons. First, the basic union unit is the workplace or enterprise which makes it difficult to develop the rallying cry of western unions for much of the late nineteenth and twentieth century – ‘‘the rate for the job’’ or ‘‘common rule’’ (Webb & Webb, 1897). Second, the Party-state does not recognise any conflict between capital and labour. Third, even if class consciousness is enhanced by, for example, the dreadful and worsening
Chinese Unions: An Alice in Wonderland Dream World
261
safety record (in 2002 workplace accidents caused 140,000 deaths and 250,000 workers lost body parts and suffered other injuries, Compa, 2004; ACILS, 2004), growing wage and income inequality, exploitation (in the true economists’ definition) of workers – particularly by foreign-invested firms, any manifestations of common interests are likely to be suppressed by the Party-state. Fourth, even the ACFTU top officials consider the members to be entirely separate from ‘‘the union’’. Wang Zhaoguo, Chairman of the ACFTU in his Report (2003) to the National Congress states that Chinese unions must ‘‘forge closer links with workers’’ and ‘‘consistently keep close ties with the masses of workers as the lifeline of trade union work’’. Such statements confirm that members’ interests are presently a low priority. As Alice said: ‘‘I’m afraid I can’t put it more clearly’’ Alice replied, very politely, ‘‘for I can’t understand it myself y very confusing.’’ (TLG, 1998, p. 41)
Further, the spread of the market mechanism to product and labour markets means, whatever the law says, that many firms do not wish to treat with unions. Under the market system unions can thrive either if they can impose costs on the employer or if they can demonstrably provide benefits by boosting the efficiency of the enterprise. Unions are incapable of imposing costs and all the evidence suggests that managers in the growing non-SOE segment of the economy believe that the efficiency of the enterprise depends on their actions, utterly independent of any union’s role. Thus, collective contracts and tripartite institutions are treated with a shrug of the shoulders – implemented because it is the law, but not embraced. And all the while employees have a low priority. In their representative survey of over 3,000 private sector enterprises the Research Group on Private Enterprises (2004) asked: ‘‘If you want to be a good employer in the private sector, which kind of good things would you like to do?’’ There were six items and the employer could answer ‘‘Yes’’ to as many as she/he liked. Only 29% said, ‘‘yes’’ to ‘‘treating workers well’’. Unions will remain nugatory, while they rest content as a transmission belt such that they convey government policies to workers and enforce labour discipline to promote those policies, as opposed to representing workers in dealing with employers and the government. Wedderburn (1995), the distinguished labour lawyer, recently wrote: ‘‘it is a truism of labour relations everywhere that workers interests cannot be protected, let alone advanced, without legitimising and promoting collective action by them in autonomous organisations’’. This statement is only partially true. In China,
262
DAVID METCALF AND JIANWEI LI
in the last 20 years real wages have trebled (OECD, 2005). But collective action by trade unions is utterly irrelevant to this huge advance in workers’ well-being which simply mirrors the very rapid growth of productivity. On the other hand, representation, voice, greater safety and a tempering of exploitation surely do require both collective action and autonomous organisation. If the Chinese government wants the trade unions to provide an effective safety valve, then it has to allow the unions a degree of independence and permit, if not encourage, them to play a representative role. But the 2003, ACFTU Congress marked a major setback for progressive trade union developments, with the installation of a more conservative leadership. Another factor inhibiting unions’ representative role is the tension between the ACFTU and the Ministry of Labour, each of which wants to claim primacy in the regulation of labour relations. Over the last few years the Ministry of Labour has increasingly asserted its authority, at the expense of the ACFTU. In particular the government is relying more on conciliation and arbitration procedures, which are the domain of the Ministry of Labour, rather than collective representation as the means of maintaining order and stability in labour relations. The need for more effective collective representation is recognised by farseeing, thoughtful (albeit cautious) ACFTU officials. Recently Qiao et al. (2004) end their document for the World Bank with what is, in effect, a manifesto for reform, including: the need to help workers develop their collective interests, because presently unions are ‘‘the weakest ring’’ in the labour market; autonomous negotiations [i.e., though not stated, outside the ACFTU] and the right to strike; to recognise that ‘‘the current situation is strong capital versus weak labour’’ which can, for the time being, only be addressed by developing properly functioning tripartite institutions; and a much stronger labour inspectorate, to cover social security payments, labour contracts, collective agreements [but not safety!]. But such reasonable reforms do seem a long way off, so unions in China will almost certainly echo the White Queen: The rule is, jam tomorrow and jam yesterday – but never jam today, (TLG, 1998, p. 177)
and, alas, tomorrow never comes.
Chinese Unions: An Alice in Wonderland Dream World
263
ACKNOWLEDGEMENT AND DISCLAIMERS We are very grateful to the seven organisations and the individuals listed in Appendix B. They made this research possible but they are not responsible for our interpretation of their contribution. The cases are referred to as Case 1, Case 2 etc., throughout the text. All the issues were discussed with Sarah Ashwin and Sue Fernie, and we acknowledge their contribution with many thanks. We are also very grateful to the anonymous referee, Simon Clarke, Fang Lee Cooke, John Pencavel, Malcolm Warner, Bill Wedderburn, Linda Yueh and Guilan Yu for comments on an earlier draft. This is joint research – neither party could have undertaken it alone-but Metcalf is solely responsible for the story told from the documentary evidence and case study material. Jianwei Li (2004) put together an annotated bibliography of around 100 items on Chinese unions. This is available on request from Linda Cleavely (
[email protected]).
REFERENCES ACFTU. (2002). Chinese trade unions: Statistics yearbook. Beijing: China Statistics Press. ACFTU. (2003). The bluebook of Chinese trade unions on protecting the legal rights of staff and workers in 2002. Beijing: China Statistics Press. Addison, J., & Schnabel, C. (Eds) (2003). International handbook of trade unions. Cheltenham: Edward Elgar. American Center for International Labor Solidarity (ACILS). (2004). The struggle for worker rights in China. Washington, DC: ACILS. Ashwin, S. (2003). The regulation of the employment relationship in Russia: The Soviet legacy. In: D. Galligan & M. Kurkchiyan (Eds), Law and informal practices. Oxford: Oxford University Press. Benson, J., Debroux, P., & Yausa, M. (2000). In pursuit of flexibility: The transformation of labour-management relations in Chinese enterprises. In: M. Warner (Ed.), Changing workplace relations in the Chinese economy. Basingstoke and London: Macmillan Press Ltd., and New York: St. Martin’s Press. Bjo¨rkman, I., & Lu, Y. (2000). Local or Global? Human Resource Management in International Joint Ventures. In: M. Warner (Ed.), Changing Workplace Relations in the Chinese Economy. Basingstoke and London: Macmillan Press Ltd., and New York: St. Martin’s Press. Carroll, L. (1865). Alice’s adventures in wonderland. London: Macmillan. Carroll, L. (1872). Through the looking glass, and what alice found there. London: Macmillan. Carroll, L. (1865) and (1872). Centenary edition, edited by Hugh Haughton (1998). London: Penguin Classics. Chan, A. (2000). Chinese trade unions and workplace relations in state-owned and joint-venture enterprises. In: M. Warner (Ed.), Changing workplace relations in the Chinese economy. Basingstoke and London: Macmillan Press Ltd., and New York: St. Martin’s Press.
264
DAVID METCALF AND JIANWEI LI
Chan, J. (1991). Wild swans. London: Harper Collins. Chen, F. (2003). Between the state and labour: The conflict of chinese trade unions’ double identity in market reform. The China Quarterly, 176, 1006–1028. Chen, S. & Lu, H. (2000). Research on the problem of labour and employment insurance in Zhejiang province. Mimeo [in Chinese]. Child, J. (1994). Management in China during the age of reform. Cambridge: Cambridge University Press. Chiu, W. C. K. (2003). Employee stock ownership plans and organizational integration among workers in the People’s Republic of China. Human Resource Management Review, 13, 407–421. Clarke, S. (2005). Post-socialist trade unions: China and Russia. Industrial Relations Journal, 36(1), 2–18. Clarke, S., & Lee, C. H. (2002). The significance of a tripartite consultation in China. Asia Pacific Business Review, 9(2), 61–80. Clarke, S., Lee, C. H., & Li, Q. (2004). Collective consultation and industrial relations in China. British Journal of Industrial Relations, 42(2), 235–254. Compa, L. (2004). The struggle for worker rights in China. Report by The Solidarity Center, Washington, DC. Cooke, F. L. (2002). Ownership change and reshaping of employment relations in China: A study of two manufacturing companies. The Journal of Industrial Relations, 44(1), 19–39. Cooke, F. L. (2005). Human resource management, work and employment in China. London: Routledge. Cooke, F. L. (2006). Informal employment and gender implications in China: The nature of work and employment relations in the community services sector. International Journal of Human Resource Management, forthcoming. Hobsbawm, E. (1976). Labouring men. London: Weidenfeld and Nicolson. Howell, J. (1997). Looking beyond incorporation: Chinese trade unions in the reform era. Mondes en Development, 25(99), 73–90. Howell, J. (2003). Trade unionism in China: Sinking or swimming<. Journal of Communist Studies and Transition Politics, 19(1), 102–122. Khor, N., & Pencavel, J. (2005). Income disparities and income mobility in China. Unpublished Paper, Stanford University, September 2005. Lazear, E. (1998). Personnel economics for managers. New York: Wiley. Leung, W. (1988). Smashing the iron rice pot workers and unions in China’s market socialism. Hong Kong: Asia Monitor Resource Centre. Leung, W. (2002). ACFTU and union organising. China Labour Bulletin, 26 April. Li, J. (2004). Chinese unions: An annotated bibliography. Working Paper 1327, Centre for Economic Performance, London School of Economics. Metcalf, D., Hansen, K., & Charlwood, A. (2001). Unions and the sword of justice: Unions and pay systems, pay inequality, pay discrimination and low pay. National Institute Economic Review, 176(April), 61–75. National Bureau of Statistics of Labour and Social Security. (2003). China labour statistical yearbook 2003. Beijing: China Statistics Press. Ng, S. H., & Warner, M. (1998). Chinese unions and management. Basingstoke: Macmillan Press.
Chinese Unions: An Alice in Wonderland Dream World
265
Ng, S. H. & Warner, M. (2000). Industrial relations versus human resource management in the PRC: Collective bargaining ‘with Chinese Characteristics’. In: M. Warner (Ed.) Changing workplace relations in the Chinese economy. Basingstoke and London: Macmillan Press Ltd., and New York: St. Martin’s Press. OECD. (2005). China, OECD Economic Surveys, 2005/13, September. Qiao, J., Lin, Y., & Jiang, Y. (2004). China labor market policies review, advisory report for world bank. Beijing: ACFTU, July. Research Group on Private Enterprises. (2004). Fifth investigation into private enterprises in China. All China Industry and Business Association and Research Institute on China Private Enterprises. Taylor, B., Chang, K., & Li, Q. (2003). Industrial relations in China. Cheltenham, Glos, UK: Edward Elgar Publishing Limited and Northampton, MA, USA: Edward Elgar Publishing, Inc. Visser, J. (2003). Unions and unionism around the world. In: J. Addison & C. Schnabel (Eds), International handbook of trade unions (Chapter 11). Cheltenham: Edward Elgar. Warner, M., & Ng, S. H. (1999). Collective contracts in chinese enterprises: A new brand of collective bargaining under market socialism. British Journal of Industrial Relations, 32(2), 295–314. Warner, M., & Zhu, Y. (2000). The origins of Chinese ‘industrial relations’. In: M. Warner (Ed.), Changing workplace relations in the chinese economy. Basingstoke: Macmillan Press. Warner, M., & Braun, W. (2002). Strategic human resource management in western multinationals in China – The differentiation of practices across different ownership forms. Personnel Review, 31(5), 553–579. Webb, S., & Webb, B. (1897). Industrial democracy (1st ed.). London: Longmans. Wedderburn, L. (1995). Labour law and the individual: Convergence or diversity. In: W. Wedderburn, Labour law and freedom. London: Lawrence and Wishart. Yao, X. & Guo, D. (2004). Empirical analysis of industrial relations in the enterprises which changed from state-owned to private. Management World, 5, 21–40, [in Chinese]. Zhu, Y., & Warner, M. (2003). Emerging paradigms of human resource management in one of China’s special economic zones. East Asian Review, 7(March), 3–28.
266
DAVID METCALF AND JIANWEI LI
APPENDIX A: ACFTU DUAL STRUCTURE OF ORGANISATION AND ITS INTERELATION WITH THE STATE GOVERNMENT Other Ministries
INDUSTRY
GEOGRAPHIC LOCATION ACTFU National Level (Congress, executive and secretariat)
Ministry of Labour/ Personnel
PARTY Central Party Committee
NATIONAL
National industrial union level
Other Departments
Provincial Union Level
Department of Labour / Personnel
Provincial Party Committee
PROVINCIAL
City, County & Town Industrial Level City, County & Town Union Level
Bureau of Labour/ Personnel
City, County & Town Party Committee
LOCAL
Primary Union Level Note: 1. 2.
ACFTU dual structure Unions interrelations with the state
Workplace Level
WORKPLACE
Individual member Level
APPENDIX B: DETAILS OF CASE STUDY COMPANIES AND SEMI-GOVERNMENT ORGANISATIONS VISITED Jianwei Li was a senior staff member of Haikou Foreign Affairs Office located in Hainan, the largest Special Economic Zone in China when this research was undertaken. With the support of Zhang Songlin, then Deputy Party Secretary of Haikou Party Committee, in August 2004 Ms Li organised extensive visits to the three commercial (case study) firms (2 days each). All such visits included detailed interviews with senior managers and union officials (often one and the same), discussion with individual workers and much observation of the shop floor in operation; and to three semi-government organisations (1 day each) in Haikou. In addition the authors visited ACFTU headquarters in Beijing (1 day). The three case study organisations were revisited in April 2005 (1 day each). In addition, extensive documentary sources about the companies were studied. All interviews were conducted in Chinese
Chinese Unions: An Alice in Wonderland Dream World
267
with Ms Li translating. Notes of the discussions – in English and Chinese – are available on request. The organisations are as follows (see also text): 1. Haikou Agriculture, Industry and Trade (Luoniushan) Co Ltd. Conglomerate operating in food production and manufacturing. Bulk of output from pig farm. Ex-SOE now listed on stock exchange. Over 2100 workers, revenue nearly 400 m RMB. Main contacts: Mr Kong, Deputy General Manager and Chairman of Trade Union Mr Sun, Secretary to General Manager and Deputy Director of Company Office 2. Hainan Haiwoo Tinplate Industry Co. Ltd. Joint venture between China, Korea and Japan manufacturing tinplate. Came fully on-stream in 2000. 240 employees; sales income 1 billion RMB; production volume 100,000 tons high-quality tinplate. Main contacts: Mr Wong, Deputy General Manager Mr Anon, Sales Manager and Chairman of the Trade Union 3. Coconut Palm Company Employee-owned coconut (and other) juice and mineral water manufacturer, ex-SOE. Some 6,000 employees (4,000 on standard contracts, 2,000 buffer temporary workers). The biggest manufacturer of coconut juice in the world. Main contact: Mr Liang, Deputy General Manager 4. Labour, personnel and social security bureau (civil service) Haikou City Bureau, which consists of three separate branch Bureau’s viz Labour and Employment, Social Security and Personnel Bureau. Out time was mostly spent with officials in the Labour and Employment subBureau, responsible for around 0.5 million workers in Haikou. Main contacts: Mr Fu, Director, Labour and Employment Mr Ke, Director, Social Security 5. Trade Union Federation, Haikou City branch of ACFTU (trade union) Coordinate and direct grass root Haikou City workplace unions – 1,400 branches with approximately 0.1 m members. Main contact: Mr Huang, Section Chief, grass root unions
268
DAVID METCALF AND JIANWEI LI
6. Enterprise Confederation and Industrial and Business Confederation (employers) Helps Haikou private sector firms interact with the Party-state. Over 2,000 member firms. Main contact: Mr Zhou, Chairman Enterprise Confederation 7. ACFTU Headquarters, Beijing: China Institute of Industrial Relations To provide intellectual leadership and advice to ACFTU and to train grass roots union officials. Main contacts: Mr Huang, Director, Public Administration Department Mr Qiao, Deputy Director, Trade Union Studies Department
THE NATURE OF UNION RAIDING: EVIDENCE FROM BRITISH COLUMBIA, 1978–1998 Chris Riddell ABSTRACT A certification attempt involving an incumbent union is known as a raid. Very little is known about union raiding, yet a large number of workers are affected by raiding and unions continue to debate the process. This paper tests various hypotheses of the nature of union raiding using unique data on raiding attempts over the 1978–1998 period in the Canadian province of British Columbia. The principal findings are (1) raiding attempts are much more likely to succeed when the incumbent union has underachieved in collective bargaining, and are often used in such circumstances; (2) employers favor the incumbent union, are effective in influencing the outcome, but use very different tactics than in regular certification; (3) independent unions fare better in raid contests relative to the national and international unions; and (4) there is a modest amount of inter-affiliation raiding, but much of this is between AFL–CIO affiliates who apparently disregard their no-raiding agreements when operating in Canada.
Advances in Industrial and Labor Relations, Volume 15, 269–293 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15006-4
269
270
CHRIS RIDDELL
1. INTRODUCTION A certification attempt involving a union applying for the bargaining rights in a unit in which another union holds certification is known as a raid. The outcome of a raid is decided by an election in both Canada and the United States. Indeed, unlike certification attempts involving unorganized employees, the raiding process is relatively similar across the two countries. In the late 1950s, raiding among American unions attracted significant interest from researchers. At that time the focus was on overall trends in raiding, the rise of the independent union and competition between independents and the AFL–CIO. Because the previous literature is descriptive in nature, we know very little about the determinants of successful raiding, or alternatively, the factors that influence the incumbent’s chances of successfully defending. Essentially, nothing is known about raiding in Canada, not even overall trends. In addition, raiding is a more important issue than is often portrayed. For instance, while raids account for only around 5% of total certifications in British Columbia – the province under study in this paper – they account for over 35% of total workers involved in certifications. Moreover, the unique nature of the raiding process – in particular, the decision is between two unions rather than whether to be unionized – may yield new insights into what members value in their union. This paper makes a number of contributions to the literature on raiding in particular and union organizing in general. With respect to the raiding phenomenon, this paper tests three major hypotheses. First, union raids are analyzed as akin to hostile takeovers. The principal theory behind hostile takeovers is that the takeover market acts as a disciplinary device, protecting stakeholders from inefficient management. I argue that raids are more likely to succeed when they target poor–performing incumbent unions. Second, it is argued that employers, on average, will favor the incumbent union over the raiding union because the employer would not want to incur the costs of establishing a relationship with a new union. Third, the extent and nature of inter-affiliation raiding and so-called ‘no-raiding’ policies is examined. These hypotheses are tested using unique data compiled from the records of the British Columbia Labour Relations Board (BCLRB) covering the November 1978–May 1998 period. With respect to union organizing more generally, this paper makes an additional contribution which stems from the unique nature of raids – particularly Canadian raids. Unlike raids in the United States, the ‘no-union’ (i.e., decertification) option does not appear on the ballot in Canadian jurisdictions, and thus, Canadian raids involve a decision purely between
The Nature of Union Raiding
271
two unions. I exploit this to test the hypotheses that (a) workers prefer independent unions over the larger national and international (i.e., U.S.based) unions, and (b) workers prefer Canadian-based unions over U.S.based unions. Two of the most striking features of Canadian union incidence over the past five decades are the decline of U.S.-based unions in Canada, and the rise of independent unions in Canada – trends that have received virtually no attention from researchers. In 1962, two of every three union members belonged to an U.S. union, but by 1995, this ratio had fallen to about 20%.1 Conversely, while there were very few independent unions in Canada in the early 1960s, by 1995, over 20% of the 4 million plus Canadian union members belonged to an independent union. It has been argued that these trends are because of changing preferences – a movement away from U.S.-based unions and toward independent unions. However, there is no empirical support for such claims. Raids provide an opportunity to test whether the changing fortunes of international and independent unions reflect preferences on the part of workers.
2. PREVIOUS LITERATURE There is substantial literature on the determinants of certification outcomes. Heneman and Sandver (1983) review the pre-1984 studies, and more recent evidence can be found in Maranto and Fiorito (1987); Thomason (1994); Bronfenbrenner (1997) and Riddell (2004) – among many others. This large body of research, however, only examines single-union organizing campaigns of previously unorganized workers. Two other types of organizing processes exist: multi-union certification bids involving unorganized employees and multi-union certification bids involving an incumbent – a ‘raid’. A raid is a contest where an incumbent union already holds certification for the bargaining unit, and the membership must vote between the two unions. The rules for raiding are complex and will be discussed subsequently. Unlike regular certifications, raid contests are largely similar between Canada and the United Statesrica. The one key difference in raiding across the two countries is that, in Canada, raid elections are only between the two unions – the ‘no-union’ option does not appear on the ballot. The raiding literature is quite small with only a handful of published papers. The literature began with Krislov (1954, 1955) in the wake of the media attention that raiding was receiving following the expulsion of the ‘left-wing’ unions from the CIO. Krisilov used aggregate data on the total number of raids to demonstrate that raiding was in fact an important
272
CHRIS RIDDELL
phenomenon. Troy (1960) was the first to examine raiding success rates, with aggregate data covering the 1948–1958 years. Three interesting conclusions arose from Troy’s work: first, raiders win less than half of their attempts; second, the ‘no-union’ option is rarely realized (around 2% on an average); and third, independent unions appear to be more likely to successfully defend a unit against a national union and more likely to successfully raid a national unit. Chaison (1973, 1976) updated the work of Troy using data for the 1964– 1973 period. While Chaison did not look at the role of independent unions, he showed that the ‘no-union’ option was still rarely realized (around 4% on an average), and suggested that this may imply that employers substitute a strategy of favoritism toward one union for a strategy of opposing both unions. Finally, Odewahn and Scott (1989) updated Chaison’s work with aggregate data on raids over the 1974–1986 period. They noted that the ‘no-union’ option was still a rarity, but appeared to be increasing over time (around 8% on an average). They also found that independent unions appeared to have some minor advantage in raiding elections. None of these studies proposed or tested a model of the determinants of raiding success. There is also some literature on multi-union campaigns involving unorganized workers. Such certification bids involve two unions competing over a proposed bargaining unit of unorganized employees; that is, no incumbent union is present. With no history of unionization in the bargaining unit, the dynamics of multi-union certification without an incumbent are very different from that of raids; however, there is one study that is relevant here. Walker and Lawler (1986) examine the organizing efforts of two unions to gain bargaining rights of the California State University system faculty, and model the individual’s decision to vote for one union rather than the other. Thus, there is an element of competition in both raids and the campaigns analyzed in Walker and Lawler. Their results indicate that campaign effects play little role. This result is different than the recent evidence from single union certification bids where ‘rank-and-file’ campaign strategies have been shown to be important in affecting election outcomes (Fiorito, Jarley, & Delaney, 1995; Bronfenbrenner, 1997). In Walker and Lawler, the key factor was the individual’s pre-campaign expectation regarding bargaining outcomes of one union relative to the other.
3. CONCEPTUAL FRAMEWORK This paper tests a number of hypotheses about the union-raiding process. The first hypothesis examines the circumstances under which a raid is more
The Nature of Union Raiding
273
likely to be successful. Raids involve a union attempting to gain bargaining rights over a bargaining unit already represented by a union. This scenario parallels that of a firm attempting a hostile takeover of another firm. The principal hypothesis behind hostile takeovers is the market for ‘corporate control’ hypothesis, which contends that shareholder interests are best protected by unrestricted competition for the stewardship of corporate assets via a takeover market (Jensen, 1988). In other words, hostile takeovers can function as a disciplinary device, protecting shareholders from management inefficiency. I hypothesize that union raids can also act as a disciplinary device, protecting workers who want to be unionized but are suffering from an ineffective union, and that raids will be more successful when they target under-performing incumbents.2 The notion that union raiding is akin to hostile takeovers is purely viewed as an analogy, but there are a number of similarities. The basic conflict of interest between agents and stakeholders is comparable. Managers are the agents of shareholders, just as the union is the agent of the union members. Because the objectives of the agent and the stakeholder may differ, conflicts of interest can arise. In the case of hostile takeovers, one conflict of interest arises because managers require free cash flow3 in order to have flexibility in capital markets, but if managers are allowed access to free cash flow it is very difficult to ensure that they will use the resources in the interests of the shareholders, and not simply make expenditures that increase the size of the firm to an inefficient level – since managerial compensation is often tied to growth (Roll, 1990). A similar conflict of interest can arise when union leaders make decisions that are seen to be beneficial to the union as a whole, but not necessarily to each bargaining unit represented by the union. For instance, Godard’s (1997) analysis of a survey of Canadian union members indicates that unions exert a great deal of effort on union-wide activities and lobbying the government – activities that benefit the union, but perhaps not the bargaining unit – while union members feel that unions do not exert enough effort on improving union-management relationships at the workplace level – activities that benefit the bargaining unit, but perhaps not the union. Second, for an incumbent that loses, the two events are analogous. In hostile takeovers, there is considerable evidence that the incumbent management is worse off because most are displaced (Jarrell, Brickley, & Netter, 1988). In raids, the incumbent union is displaced if it loses the raid, and thus loses power and dues revenues. Third, the criticisms of the two events are comparable. The primary criticism of hostile takeovers is that they are inefficient because any gains to a given party are purely a redistribution based
274
CHRIS RIDDELL
on another party’s losses – and yet takeovers are costly (Jarrell et al., 1988). Critics of raiding make a similar argument. A raid simply transfers the right to represent a given bargaining unit (and the dues collected) from one union to another with no gain in members for the union movement as a whole.4 The second hypothesis posits that management activity during the raid election favors the incumbent union.5 Employers may favor the incumbent because, ceteris paribus, employers do not want to incur ‘switching costs’. That is, conditional on having a union, it is unlikely that employers wish to incur the costs of establishing a relationship and appropriate infrastructure with a new union. As noted above, previous authors have argued that employers may substitute from a union avoidance strategy toward a strategy of favoring one of the two unions given the rarity of the ‘no-union’ option. This case is bolstered in Canada where raids only involve a decision between the two unions, the ‘no-union’ option does not appear on the ballot. Arguments have not been previously presented for which union the employer will favor. Under the switching-cost hypothesis, employers – on average – should favor the incumbent. Given the very different nature of raid contests relative to regular certification bids – where workers are choosing between two unions – the evidence on management behavior may provide new insights into the role of the employer in the industrial relations system. The final part of the paper examines the nature of inter-affiliation raiding. The AFL–CIO and the Canadian Labour Congress (CLC) – the largest federation in Canada – have no-raiding agreements, which is an important feature of the raiding market.6 Most unions in Canada are affiliated with the CLC, and most of the international unions operating in Canada are affiliated with both the AFL–CIO and the CLC. If a raid occurs between CLC affiliates, the current CLC Constitution allows for the incumbent union to file a complaint in which case an independent ombudsperson is appointed to conduct an investigation. There are circumstances where a raid among affiliates is justified. Specifically, if the incumbent union’s own constitution allows locals to disaffiliate from the incumbent union voluntarily (i.e., no influence from another party), then the local membership can, in turn, join another union (or create an independent). If the raid is not ruled to be justified then the CLC Executive Committee votes as to whether the president should impose sanctions on the raider. Normally, ‘level 1’ sanctions are imposed along with a cease and desist order; if the raiding activity continues, full sanctions are imposed.7 According to the CLC, a raiding union under sanctions cannot be raided by other CLC affiliates without fear of sanctions. As well, unlike the
The Nature of Union Raiding
275
AFL–CIO, there are no financial penalties or expulsion options included among CLC sanctions. Conceptually, for federation policy to play a role in mitigating raiding among affiliates, three events must occur. First, the incumbent union must care sufficiently about the raided unit to file a complaint to the federation, which in turn initiates a long, costly and complicated process. Second, it must be determined that the raid was not justified. Third, the sanctions imposed must be sufficiently meaningful that the raiding union cares about the penalties incurred. These three events may not always be occurring. A number of high-profile raids have been recently launched by CLC affiliates against other CLC affiliates. For instance, in 2000, eight units of the Service Employees International Union (SEIU) in Ontario were raided by the National Automobile, Aerospace, Transportation and General Workers Union of Canada (CAW).8 The CAW argued that the raiding was justified because it simply offered support to a group of local unions that had voted in favor of disaffiliating from their parent union. In his ruling, however, the umpire indicated that while the local presidents ‘made a good faith judgment that the future of their members would be best served by disaffiliation from the SEIU and merger with the CAW’, the CAW played a much more active role in the raid. In particular, the umpire found that formal discussions between the eight locals and the CAW had been held for a considerable period of time and that a merger agreement had been reached prior to the locals’ application for disaffiliation from the SEIU. Ultimately, level 1 sanctions were imposed and, after the raiding continued, full sanctions followed. Even after full sanctions were imposed, the raiding activity continued (i.e., those raid applications still being processed were not withdrawn by the CAW).9 The case illustrates that (a) the justification process is complicated and, potentially, open to debate and (b) CLC sanctions may not be sufficiently meaningful to stop raiding activity, even after a raid has been deemed unjustified. There is U.S. evidence that no-raiding agreements have failed in preventing raiding among affiliates. As stated in Resolution 28 of the 1999 AFL–CIO Convention Resolutions, ‘although the language set forth in Article XX of the AFL–CIO Constitution permits the AFL–CIO to severely sanction an international union that purposefully violates the no-raiding commitment by imposing substantial financial penalties or directing that the offender be expelled from the AFL–CIO, such severe sanctions are rarely utilized’. Financial sanctions force the raiding union to compensate the incumbent union for expenses in pursuing the Article XX process (which is
276
CHRIS RIDDELL
very similar to the CLC procedure), lost dues, or other costs incurred as a result of the raid. Resolution 28 later states that the AFL–CIO will be more aggressive toward raiding unions in the future, a policy which may have worked to some extent in deterring raiding. Following a cease and desist order from AFL–CIO President John J. Sweeny, the United Transportation Union, claimed that it disaffiliated from the AFL–CIO to avoid the financial penalties that were likely to be imposed as a result of their on-going raids on the Brotherhood of Locomotive Engineers (BLE Newsletter, 2000). Overall, it is unclear whether federation no-raiding policies have reduced raiding activity in Canada, given the possibility of ‘justified raiding’, the long and costly dispute process and the relatively weak sanctions available. I examine two specific components of inter-affiliation raiding. First, I simply document the extent of inter-affiliation versus cross-affiliation raiding, which has not previously been done despite the fact that inter-affiliation raiding continues to be a much debated issue among unions. Second, using the case study of the Teamsters – who joined the AFL–CIO several years before joining the CLC and thus have unusual variation in the timing of joining different federations – I examine whether no-raiding policies are effective in reducing raiding activity.
4. DATA AND METHODOLOGY The principal data was compiled from the records of the British Columbia Labour Relations Board (BCLRB). In Canada, the Labour Code falls under provincial jurisdiction, and thus micro-data on union certifications must be collected at a provincial level, and is considered confidential data. It is therefore difficult to conduct analyses such as that in this paper for multiple provinces. A total of 606 raids were launched from October 1978 to May 1998, covering 57,508 workers. Over this time period, there were around 10,000 regular certifications filed covering just over 110,000 workers. Raids therefore appear somewhat trivial in terms of the number of certifications involved, but are clearly not trivial in terms of the number of workers involved. The British Columbia Labour Code allows for a raid under two broad scenarios.10 In cases where a collective agreement is not in force, unless the board consents, a raider cannot apply for certification of a unit if the incumbent has been certified for 6 months or less and/or is involved in a strike or lockout. Second, in cases where a collective agreement is in force, a raider can only apply for certification during the seventh or eighth month in each
The Nature of Union Raiding
277
year of the agreement (or any renewal of the agreement). The strike or lockout rule applies for cases where a collective agreement is in force as well. These regulations are very similar to the United States, although the window period for a raid that is launched while a collective agreement is in force in the United States is 60–90 days prior to the expiration of the contract (60–120 days in the health-care sector). The petition requirements for raids in British Columbia are similar to regular certifications: a raider must claim to have a majority of individuals in the unit signed up, and must be able to substantiate this support in order to proceed to a vote. This is different from the United States, where only 30% support is required (as is also the case for regular certification). While the petition requirements are somewhat different between British Columbia and the United States, the balloting procedures are essentially the same. Unlike certification applications involving unorganized employees – where, since 1984, there have been time limits on elections in British Columbia – no balloting restrictions are in place for raids. As a result, it is not uncommon for a raid campaign in British Columbia to drag on for a year. A few restrictions are imposed on the data based on the earlier-noted procedural details. There are 49 applications eliminated from the sample where the raid was automatically rejected without a vote owing to petition violations, either because it was not launched within the proper time frame (i.e., seventh or eighth month), the raider was out of time (i.e., raid began during seventh or eighth month, but petition requirements were not completed by the end of eighth month), or the proposed bargaining unit was deemed inappropriate for collective bargaining (in the case of a partial raid where only a part of the original bargaining unit is raided). There was one raid where the employer went out of business. These exclusions yielded a total sample of 556 raids.11 The first hypothesis posits that raids will be more successful when the incumbent union has underachieved in collective bargaining. I implement this with two tests. The first test is that raiding will be relatively more successful when the incumbent has failed to negotiate a first collective agreement.12 Successfully negotiating a first collective agreement is often viewed as a ‘later stage’ indicator of organizing effectiveness (Fiorito et al., 1995). In British Columbia, failing to negotiate a first collective agreement may not be as serious as in some other Canadian provinces or the United States because of the first contract arbitration legislation, which was originally contained in section 70 of the British Columbia Labour Code. However, negotiating a first collective agreement in a timely fashion is still likely to be seen as an important achievement by the membership for a number of
278
CHRIS RIDDELL
reasons. First, and most importantly, first contract arbitration was, historically, rarely used by the BCLRB. The BCLRB seldom used the section 70 power, not because there were few cases where unions and employer struggled to reach an agreement, but because the legislation has always been viewed by the board as a controversial and ‘last resort’ instrument.13 Even if first contract arbitration is seen by the membership as a viable solution, failing to negotiate a first collective agreement in a timely manner will likely introduce an element of uncertainty to the union–membership relationship, in which case a raid launched under such circumstances may be more likely to be successful. The second test of the disciplinary device hypothesis is that raids will be relatively more successful when the incumbent negotiated a below-average wage settlement. The industry average is used to compute this variable. To determine the previous wage settlement, I attempt to match each raid – using the employer’s name, union’s name and other information available on the certification application – to the appropriate collective agreement. In cases where a collective agreement is not in force, the most recent wage settlement is used. In cases where a collective agreement is in force, the current settlement is used. In general, for 1990–1998, this is relatively straightforward because collective agreements were available at the time of this research from the BCLRB. For earlier years, however, I had to match raids to wage settlement information from the monthly Industrial Relations Bulletin, which does not always contain every agreement negotiated in a given month. Thus, regressions are estimated with and without the wage settlement variable. To compute the wage settlement variable, an average hourly rate of all job titles within the unit was calculated, since the BCLRB records do not contain any information about the individual workers. The second hypothesis posits that the employer, ceteris paribus, will favor the incumbent due to ‘switching costs’. To test this hypothesis, I use unfair labor practices as a proxy for management interference. This includes section 3 ULPs for the 1978–1992 period and section 6 ULPs for post-1992 years. ULPs have the advantage of indicating that there was some severity to the employer behavior. In general, I only observe if an unfair labour practice (ULP) was filed against the employer, but not which outcome (raider, incumbent) the employer was trying to influence voters toward. Thus, the mean of the ULP variable by itself will not be informative regarding the employer’s preference for the incumbent. A negative coefficient, however, would imply that the employer prefers the incumbent. A positive coefficient would indicate a preference for the challenging union. If the coefficient is not statistically different from zero, there are a number of possibilities;
The Nature of Union Raiding
279
employers may not be able to influence voters, or employers may not care who wins (or indeed may have done nothing with the ULP filing being simply a union tactic). For the 1987–1998 period, information on ULPs was available from the BCLRB and was matched to the raids data using the employer’s name, union and dates of the raid and ULP filing. For the 1978– 1986 sample, the raid applications themselves indicated whether an ULP was involved. As noted, one of the results from the U.S. literature on raiding is that the independent unions appear to be holding their own. To investigate how Canadian independents perform in raid contests a set of union controls is included. Independent unions are classified into four groups: independent local unions (ILU), which are single-employer unions; provincial independent unions (PIU), which are multi-employer unions operating only in British Columbia and tend to be specialist unions such as the B.C. Nurses Union and Health Science Association; national independent unions (NIU), which operate across Canada; and the confederation of Canadian unions (CCU), which consists of a group of left-wing, former NIUs that banded together in 1983 to challenge the CLC because they felt the CLC was too internationally oriented (i.e., the CCU unions are all classified as NIUs prior to 1983). To categorize the unions, the British Columbia Labour Directory (BCLD) was used for all years because the affiliation of some unions changed over time. The annual BCLD contains a listing of all unions operating in the province including information on affiliation. It is useful to note that while independent unions appear to have died out in the United States (see Jacoby, 2001), independent unionism has steadily increased in Canada. While there were virtually no independent unions in the 1960s, over 20% of union members were represented by an independent union by the early 1990s (CALURA, 1992). Thomason (1994) and Riddell (2001) both found that Canadian independents were more likely to successfully organize a unit of unorganized workers relative to all other unions. Raids provide an alternative test of the effectiveness of independent unions because, as noted, raids in Canada purely involve a decision between two unions. Several additional controls are included in the raiding success model. The incumbent’s tenure (number of years certified) is included. Incumbents that have been the unit’s bargaining agent for relatively longer periods of time may have built up loyalty (or simply be more entrenched in the culture of the establishment). The size of the unit is also used. The literature has generally found that, in a non-raid setting, unit size is negatively related to organizing success (Heneman & Sandver, 1983), but there is no consensus as to why unit size affects organizing (Demsetz, 1993; Farber, 2001). One
280
CHRIS RIDDELL
argument is that unit size is negatively related to certification success because smaller units are more homogenous and communication is better facilitated in small groups. Raids may shed further light on the role of unit size given that the decision is between two unions as opposed to being between unionization and non-unionization. Finally, industry dummies are included in order to capture unobserved factors that are correlated with the nature of the industry, and year dummies are included to capture unobserved factors correlated with time such as changes in union mergers – which increasingly occurred over this time period – as well as changing economic conditions.
5. THE CORRELATES OF RAIDING SUCCESS The first column of Table 1 presents summary statistics. The raiding union was successful in 56% of cases with the incumbent maintaining its status in the remaining cases. If raiding success is plotted over time, there is no discernable trend – raiding success is largely constant over time, fluctuating from year to year in about a 10% point range around 56%. About 13% of raid applications involved a unit where no first agreement had been reached. These 75 cases amounted to over 3,000 workers. For those non-first agreement cases (481 applications), where previous wage settlement information was available (299 applications), 40% had wages below industry average. Around 8% of raids involved a ULP charge against the employer for attempting to influence voters – well below the incidence of such charges in regular certifications for this province, which was 25% for the 1987–1998 period (Riddell, 2001, 2004). The average tenure of incumbents is 13 years; the distribution is skewed however with a median tenure of 7 years. In general, raids tend to be launched on units that have been certified relatively recently. Raiding and union affiliation will be discussed in more detail later in the paper, but one observation is noteworthy here. In particular, the U.S.-based unions are more likely to be raided over the sample period – being the incumbent in 65% of cases, but the raider in only 40% of cases. The second and third columns of Table 1 present the results for two different specifications. Column two contains the full sample of 556 raids, and thus includes both raids launched on units where no first agreement was reached and all other raids. Column three restricts the sample to non-first agreement cases, and introduces the wage variable, which is only available in about 65% of non-first agreement cases. I have compared the descriptive statistics of those cases where wage information could be found relative to
The Nature of Union Raiding
Table 1.
281
Estimated Marginal Effects of the Change in Probability of Raiding Success.
Variable Raid successful No first collective agreement Prior/current wage settlement below industry average Unfair labor practice against employer Incumbent tenure Incumbent is U.S. union Incumbent is CLC Incumbent is ILU Incumbent is PIU Incumbent is NIU Incumbent is CCU Raider is U.S. union Raider is CLC Raider is ILU Raider is PIU Raider is NIU Raider is CCU Unit size Manufacturing Construction Primary Transportation and storage Trade Hotels and restaurants Miscellaneous private sector services Public sector Year dummies Pseudo R2 w2 statistic Log likelihood Number of observations
Mean 0.558 (0.497) 0.134 (0.341) 0.395a (0.489)
Marginal Effect Dependent variable 0.305 (0.063) – –
0.087 (0.282)
0.241 (0.112)
13.02 (11.67) 0.652 (0.477) 0.098 (0.297) 0.040 (0.196) 0.062 (0.241) 0.094 (0.292) 0.054 (0.227) 0.395 (0.489) 0.114 (0.318) 0.025 (0.157) 0.181 (0.385) 0.201 (0.401) 0.083 (0.277) 93.72 (259.23) 0.203 (0.403) 0.083 (0.277) 0.069 (0.253) 0.092 (0.290)
0.008 (0.002) Omitted category 0.049 (0.091) 0.069 (0.148) 0.211 (0.120) 0.197 (0.085) 0.230 (0.103) Omitted category 0.079 (0.081) 0.104 (0.149) 0.026 (0.109) 0.011 (0.099) 0.002 (0.124) 0.000 (0.000) 0.224 (0.086) 0.213 (0.108) 0.355 (0.086) 0.107 (0.116)
0.036 (0.187) 0.239 (0.427) 0.136 (0.343)
0.079 (0.153) 0.290 (0.094) 0.243 (0.095)
0.141 (0.349) – – – – 556
Omitted category Yes 0.18 123.53 317.38 556
Notes: Standard errors (standard deviation for means) are in parentheses. a The sample size for the wage settlement variable is 299. Statistical significance for the 10% level. Statistical significance for the 5% level. Statistical significance for the 1% level.
0.148 (0.061) 0.262 (0.136) 0.005 (0.003) Omitted category 0.022 (0.118) 0.110 (0.194) 0.253 (0.123) 0.308 (0.090) 0.334 (0.111) Omitted category 0.109 (0.121) 0.184 (0.152) 0.004 (0.107) 0.041 (0.090) 0.037 (0.128) 0.000 (0.000) 0.240 (0.109) 0.173 (0.132) 0.432 (0.088) 0.199 (0.135) 0.205 (0.176) 0.180 (0.132) 0.247 (0.120) Omitted category Yes 0.15 58.01 168.51 299
282
CHRIS RIDDELL
those cases where it could not, and there are no major differences between the two. It therefore appears unlikely that the missing cases are introducing a sample selection bias, but this remains a caveat for the column three results. The dependent variable equals to one if the raider was successful, zero if the incumbent was successful. All regressions are estimated by probit although the results are very similar for logits and linear probability models. To simplify the exposition the estimated probit coefficients have been converted into marginal effects.14 Beginning with column two and the paper’s first hypothesis, the results indicate that raids launched when an incumbent has yet to negotiate a first agreement are 30% points more likely to be successful – an enormous effect when one considers that, on average, raids are only successful 56% of the time. The results suggest that raids are more likely to be successful when used as a disciplinary device. While this result may not appear particularly surprising, it is somewhat controversial given that most unions argue there is no advantage to the raiding process. From a worker’s perspective, however, it may be the case that raiding provides a useful tool – a mechanism that provides an alternative source of representation when the current source is not performing up to the membership’s standards. Of course, this finding alone does not necessarily support the notion that raiding is a viable alternative to pursuing negotiation because the unit may face the same difficulty negotiating a first contract with the new union (i.e., the successful raider). Column three provides further evidence in support of the first hypothesis that union raiding operates as a disciplinary device. Raids launched on units where the current or previous agreement involved an average wage settlement below the unit’s industry average are 15% points more likely to be successful. Again, raiding may provide a useful service – allowing members to switch to a different union when dissatisfied with their current union. Moreover, from the union movement’s perspective, it should be considered whether a raid is preferred to a decertification, which, under no-raiding policies, is the only alternative. As well, I have compared the affiliation status of those raids involving an underachieving incumbent (both in terms of failure to reach a first contract, and in terms of wage settlements), and there is no correlation between a raid being an inter-affiliation case and the incumbent being an under-achiever. Thus, it appears that one cannot argue – unions only non-comply with no-raiding policies in cases of underachievement. The other estimates from column three are largely consistent with those from column two, and thus for brevity I focus on the estimates from column
The Nature of Union Raiding
283
two. Incidentally, the similarity of the estimates across columns two and three implies that there is no interaction between the covariates and the failure to negotiate a first contract. It appears that employers are able to influence the outcome of a raid campaign. Column two indicates that employer ULPs reduce the probability of raiding success by 24% points – a magnitude much larger than that found in the Canadian management opposition literature for regular certifications. Thus, employers appear to be less active in raiding elections than regular certification campaigns, but are still highly effective in influencing voters. Again, the ULP variable is of interest because it acts as a proxy for any management behavior: strategies that favor the incumbent and strategies that favor the raider. The negative estimate implies that, on average, employers favor the incumbent. To provide some evidence on how employers intervene in raid campaigns, Table 2 presents information on the ULPs filed over the 1986–1988 period. For the latter three years only, additional information is available on all ULPs filed including section 4(3) ULPs, which are complaints of illegal union practices of attempting to influence voters. Of the 15 complaints filed over the 3-year period, 10 were filed against the employer, 3 of which alleged illegal coercive practices by both the employer and the incumbent. There were no complaints filed against the employer and raider. Moreover, of the 10 complaints filed against the employer, only 1 was submitted by the incumbent union. While Table 2 provides only limited evidence, the results support the previous result that the employer favors the incumbent. Of particular interest is how employers intervene in raid campaigns. Table 2 indicates that general coercion practices – which is any activity that falls short of dismissal or actions against an individual such as demotion – and interference with trade union administration are the tactics of choice. There is only one case alleging illegal dismissal of employees. This is a completely different result relative to the evidence on how Canadian employers intervene in regular certification bids. Thomason (1994) and Riddell (2001, 2004) show that dismissal ULPS account for, by far, the majority of ULP complaints in regular certification campaigns. Thus, while employers are quite active in raid campaigns, it appears that they use different – in particular, ‘softer’ – tactics. Turning to the role of union affiliation, independent incumbents fare quite well in raid campaigns. All of the estimates for the union variables are evaluated relative to an international union. With respect to being the raider, no statistically significant differences are found across unions. However, with the exception of the independent local unions, a raid is
284
CHRIS RIDDELL
Table 2. Unfair Labor Practices during Raid Campaigns, 1986–1988. ULP Charge Filed By
Nature of ULP
Raider
Incumbent
General coercion General coercion General coercion General coercion Interference
Employer
Raider
Interference
Employer
Raider
Interference
Employer
Raider
Dismissals
Employer
Raider
General coercion General coercion General coercion Interference
Employer and incumbent Incumbent
Raider Incumbent Employees
Employee Raider Raider Incumbent Raider
Raider
General coercion Interference
General coercion
ULP Charge Filed Against Employer and incumbent Employer Raider Raider
Employer and incumbent Employer Raider Employer
Incumbent
Outcome of ULP
Outcome of Raid
Dismissed
Lost
Withdrawn or settled Withdrawn or settled Withdrawn or settled Meritorious (cease and desist) Withdrawn or settled Withdrawn or settled Meritorious (reinstatement) Withdrawn or settled Dismissed
Lost
Withdrawn or settled Withdrawn or settled Withdrawn or settled Meritorious (cease and desist) Dismissed
Won Lost Lost
Won Won Lost Lost Won Lost Lost Lost Lost
Won
substantially less likely to be successful if the incumbent is an independent or a CCU union (which are all former independents) with marginal effects on the order of 20% points. As well, the finding that ILUs are just as likely to win a raid as anyone else – and no less likely to lose a raid than the international or CLC unions – is interesting because the ILUs generally have substantially fewer resources than their opponents. Thus, the benefits of being an offshoot of the immediate workplace appear to at least cancel out any disadvantages from having fewer resources.
The Nature of Union Raiding
285
The results also imply that industry/occupational specialization may be a factor that union members value. As noted, essentially all of PIUs are specialist unions with most of the observations being either the B.C. Nurses Union or the Health Science Association. The evidence is therefore consistent with the finding in Fiorito et al. (1995) that unions with a specialization in an industry are more likely to successfully organize. The results for the CCU unions are not surprising. The small group of CCU unions are known to be left wing and very nationalistic; units that have chosen a CCU union are likely to contain similarly minded workers who would not be pleased with a CLC or international union even if their incumbent was underachieving. The result that NIUs are more likely to defend a unit is somewhat more intriguing given that these are all large unions with membership numbers similar to national and international unions. Overall, the results support the notion that a union member represented by an independent enjoys the freedom from the affiliation and political baggage that often accompanies the national and international unions. It should be noted that there are very few cases where one of the four independent unions (including CCU) raided another independent, and thus there is no interaction effect between the raider variables and the incumbent variables. There is one piece of survey evidence that Canadian independents provide better services than national/international unions such as allowing for greater input in union affairs, faster and more personal services, better knowledge of the company’s operations and procedures, lower dues and freedom from the political considerations of national and international leadership (see Rajin & Grigoletis, 1978). Some of these benefits are consistent with the organizing strategies discussed in Fiorito et al. (1995) and Bronfenbrenner (1997), which have been shown to increase organizing success rates. Overall, the results in this paper indicate that Canadian independent unions are doing something right. Given the rapid rise of independents in Canada and the current state of Canadian unionization – where union density has fallen by nearly 10% points over the 1980s and 1990s (Riddell & Riddell, 2004) – as well as the United States, it would be useful for future research to examine why the Canadian independent unions have been so successful. Interestingly, the incumbent’s tenure appears to be an important factor. From column two, an increase of one standard deviation from the mean in incumbent tenure reduces the likelihood of raiding success by almost 10% points. It is difficult to know what characteristic(s) incumbent tenure is capturing, but one possible explanation of the tenure relationship is that
286
CHRIS RIDDELL
union members become increasingly loyal to their union the longer it has been certified as the bargaining agent. Such loyalty is likely won through a consistent demonstration of satisfactory performance over the years. There is no correlation between unit size and raiding success. The industry dummies indicate that raids are much less likely to be successful in the blue collar, more militant, traditional areas of union strength including manufacturing, construction and the primary sector (mainly forestry firms).
6. FEDERATION POLICIES ON RAIDING I now proceed to the three hypotheses relating to federation policies on raiding. As discussed, there is some reason to question the effectiveness of federation policies on raiding activity. However, the earlier discussion was not based on direct evidence and thus, I explicitly examine the incidence of inter-affiliation raiding. To do this, I classify unions into affiliations using the BCLD and examine their overall raiding activity. Some unions changed their affiliation (in some cases multiple times), and thus it is important to track a given union over time. A union is only classified as having a given affiliation for the appropriate years. For this analysis, I use the full sample (i.e., 606 raids) because any raid that was attempted should be considered. Table 3 presents the results on raiding activity by the various affiliations operating in the province. The Canadian Federation of Labour (CFL) consisted of a group of AFL–CIO international unions who broke away from the CLC in 1982 owing to being discontent with CLC policy toward the traditional trades unions. The ‘AFL–CIO-only’ group consists of those international trades-oriented unions that broke away from the CLC in 1982, but opted out of the CFL, the Teamsters over the 1988–1992 period, and the former CFL unions after the CFL folded in 1996. A number of useful results are seen in the table. First, by comparing the second and third columns, it is clear that the ILUs, and especially the international unions affiliated with both the AFL–CIO and CLC, receive more than their share of raiding attempts. In particular, the AFL–CIO/CLC internationals were raided 314 times over the 21-year period, but only attempted 113 raids. As well, there was clearly a great deal of loyalty among the six CCU unions operating in the province: there was not a single raid among CCU unions. Raiding among CLC affiliates is reasonably high at 34% while raiding among AFL–CIO affiliates is common. However, further analysis reveals that the aggregate numbers in Table 3 are somewhat deceptive. Of the 24 raids by CLC-only unions (i.e., Canadian national unions only affiliated with the CLC), only 2 were against other
The Nature of Union Raiding
Table 3.
287
Raiding Incidence and Match-Ups by Union Affiliation, 1978–1998.
Union and Affiliation
International: AFL–CIO only International: AFL–CIO and CFL International: AFL–CIO and CLC International: Teamsters 1978–1987 National: CCU National: CLC Independent: ILU Independent: PIU Independent: NIU Total
Number of Raids as a Raider (1)
Number of Raids as an Incumbent (2)
Number of Raids Against an Affiliate (3)
Incidence of Inter-Affiliation Raiding (3)/(1) (%)
52 51
10 41
38 23
73 45
113
314
47
42
27
21
N/A
–
48 71 15 110 119 606
32 62 27 38 61 606
0 24 N/A N/A N/A 132
0 34 – – – 22
CLC-only unions. The remaining 22 were against international unions affiliated with both the AFL–CIO and the CLC. Thus, it would appear that the recent set of raids by the CAW against the SEIU is not an isolated event; there is a history of CLC national unions raiding CLC international unions. In addition, of those 23 raids by an AFL–CIO/CFL union against an affiliate, none were against another CFL union – all were against AFL–CIO unions that did not join with the CFL. There was clearly a great deal of loyalty among the international unions who formed the CFL. Moreover, of the 38 raids by unions coded as ‘AFL–CIO only’, all 38 were against AFL–CIO/CLC unions. Overall, a key result in Table 3 is that the U.S.based AFL–CIO unions have generally ignored their no-raiding policies when acting in Canada. Another key result is that affiliated unions can be loyal when a strong cause binds them together as is illustrated through the tightly bound CFL and CCU federations. The final issue examined is whether federation no-raiding policies reduce raiding activity. One possible way to address this question is to examine raiding activity over time as a function of the presence of no-raiding policies. However, there were no major changes in no-raiding policies over the sample period, and thus one can only rely on unions changing their affiliation thereby becoming covered (or un-covered) by existing federation policies. A problem with the latter is that there has been a great deal of union mergers
288
CHRIS RIDDELL
and quasi-mergers15 in the province (and across Canada), and thus it is difficult to separate merger effects from no-raiding policy effects. Moreover, there has been a steady, downward trend in raiding activity that cannot be solely attributed to merger activity. An alternative methodology is available, and is in the spirit of natural experiments. In particular, the Teamsters had multiple changes in their affiliation, allowing for an analysis of two different changes in no-raiding policy coverage without any contamination of merger effects. Table 4 presents the history of the Teamsters’ raiding activity in the province. The downward trend in raiding activity can be seen as the number of Teamsters raids per year fell by 5 during their independent years to 3 in both subsequent affiliation eras. More interestingly, the data implies a powerful effect of no-raiding policies. Raids on AFL–CIO affiliates fell sharply after the Teamsters joined the AFL–CIO, but did not change against CLC unions. After joining the CLC, raids fell to virtually zero against affiliates. However, raiding activity against independents increased Table 4.
The Raiding Activity of Teamsters, 1978–1998.
Years and Affiliation
1978–1987 Independent Raids versus independent Raids versus AFL–CIO Raids versus CLC Total raids Total raids per year 1988–1992 AFL–CIO Raids versus independent Raids versus AFL–CIO Raids versus CLC Total raids Total raids per year 1993–1998 AFL–CIO and CLC Raids versus independent Raids versus AFL–CIO Raids versus CLC Total raids Total raids per year
Raiding Activity as a Raider
Raiding Activity as an Incumbent
Total Raiding Activity
14
13
27
10 3 27 2.7
6 2 21 2.1
16 5 48 4.8
4
5
9
1 3 8 1.6
1 1 7 1.4
2 4 15 3
11
6
17
1 0 12 2
0 0 6 1
1 0 18 3
The Nature of Union Raiding
289
sharply. In general, raiding among the U.S.-based AFL–CIO remains an issue in the province as discussed above, and the case of Teamsters may not be representative of other unions. Nevertheless, the case illustrates how noraiding policies can dramatically reduce raiding activity even in the presence of an ‘open-door’ raiding policy as the Teamster had against independents.
7. CONCLUSION There has been no systematic, empirical study into the nature of the raiding process. However, raiding continues to be a fiercely debated issue among North American unions, and in fact, is an important phenomenon for individual workers: A total of 606 raids were launched over the sample period covering 57,508 workers – relative to around 10,000 regular certifications covering just over 110,000 workers. Moreover, the unique nature of raids – where individuals choose between two unions – suggests that they may provide useful insights into what workers want out of their union. The key findings are that raids are much more likely to be successful when the incumbent has under-achieved in collective bargaining. Moreover, the summary statistics indicate that raids are often used to discipline an underachieving incumbent. I also find that, on average, employers appear to favor the incumbent and are able to influence voters in that direction – but tend to do so using much softer tactics than those used in regular certification bids. Independent unions perform quite well in raiding. Independent local unions are just as likely to successfully raid another unit as any other union and just as likely to defend their unit as are the larger national and international unions. The national independent unions, ‘specialist’ provincial independent unions, and left-wing (former independent) CCU unions are more likely to successfully defend a unit relative to the national and international unions. While not conclusive, these findings provide the first hard evidence that the striking decline of U.S. unions in Canada, and increase in independent unionism may stem from preferences on the part of workers. Overall, the findings are consistent with the notion that union raiding may be an efficient mechanism that facilitates competition for union services. While many actors in the labor movement oppose raiding, perhaps considerations should be given to raiding as a useful device – and a preferable alternative to decertification. Raiding activity has systematically declined over time in the province (and also in the United States) during an era of mergers and condensed federations – which have no-raiding policies. The evidence presented in the paper is consistent with the view that no-raiding
290
CHRIS RIDDELL
policies have played a key role in declining raiding activity. The question for federations is then – without a viable raiding market – Will union members turn to decertification?
NOTES 1. See the annual reports of the Corporations and Labour Unions Returns Act (CALURA, 1992) for data and further information on Canadian union incidence. 2. Some behavior supports the disciplinary notion. The Association for Union Democracy (AUD) is an organization that assists union members with U.S. laborrelations law. One question posted at the AUD website in 2001 reads ‘‘We are not happy with our union and want to change unions. What are our options? When do we have a right to vote on a new one?’’ The response was ‘‘We get asked this question a lot. People who are dissatisfied with their union, who see that their leadership is unaccountable, who only get stonewalled when they complain or make suggestions, often come to the conclusion that they need to get a better union’’ (see www. uniondemocracy.com). This suggests that changing unions is an option considered by some union members. 3. ‘Free cash flow’ is cash flow in excess of that required to fund the firm’s projects that have positive net present values. 4. There are, of course, many differences between union raiding and hostile takeovers. For instance, unions are inherently more political organizations than firms; thus, while shareholders may simply have their dividend in mind when voting in a hostile takeover contest, union members likely have a more complicated decision to make since the benefits to unionization are more complex. As well, in hostile takeovers there can be ‘golden parachutes’ where incumbent executives receive buyouts when displaced (Jensen, 1988). Incumbent unions who lose a raid receive no such severance. To the extent that the hostile takeover market is unrestricted, the raiding market may also differ in that it is a restricted market owing to federation no-raiding agreements. No-raiding agreements imply that raids will typically involve unions affiliated with different federations. 5. As in the United States, employers in British Columbia are not allowed to influence voters in any manner. Employees, or the union, may file an unfair labor practice (sections 2 and 3 of the Code from 1978 to 1992, and sections 5 and 6 from 1993 to 1998) against the employer under such circumstances. However, employer coercion (and especially dismissals) during regular certification bids is fairly common in British Columbia, and even when an unfair labor practice is found to be meritorious, penalties to the employer are largely compensatory (Riddell, 2001). 6. The AFL–CIO policy is contained in Article XX of their Constitution and the CLC policy is in Article IV. 7. Level 1 sanctions means that the raiding union is denied any service that the CLC offers including educational functions and conferences. Full sanctions add other penalties including: after three months of the initial decision, loss of the right to vote on Executive Committee issues (assuming the union has representation on the Executive Committee), and after six months of the initial decision, any executive
The Nature of Union Raiding
291
members is removed from office and the union is denied the right to use the CLC justification and dispute procedure. 8. The details of this case are available in the umpire’s report located at the CLC’s website (see www.clc-ctc.ca). 9. A summary of the chronology of the CAW/SEIU raid and CLC decisions is available at the CLC’s website. 10. The rules regarding raiding and certification are in Part 3, sections 18 and 19 of the British Columbia Labour Code. 11. While it is debatable, the 11 unopposed raids, where the raid was granted because the incumbent did not respond to the application, are included. The incumbent may have abandoned (or wished to) the unit, in which case a ‘hostile takeover’ has not occurred and the raid should be dropped. However, it is also possible that unopposed raids are cases where the raider signed up an overwhelming majority of the unit, and thus the incumbent decided it was not worthwhile to contest the bid. I do not observe the specifics of the unopposed raids and so, given the small sample, I include them. The eight cases where the raid was rejected because the raider did not have a majority of the unit signed up are included in the sample since there is clear evidence that the raid failed. 12. For the 1978–March 1989 data, the data indicate whether an agreement was in force at the time of the raid and the dates of the agreement. After March 1989, the raid records contain no information about collective agreement status, but collective agreements are available from the BCLRB (in general from 1990 onward). To determine whether a first contract had been negotiated at the time of the raid for the 1978–1989 period, I match each raid to the British Columbia Calendar of Expiring Collective Agreements (CECA), which has information – including employer name, union name, unit size, city and industry – on all agreements expiring in a given month that cover 50 or more employees. If no agreement could be found for a given unit, then the case is coded as ‘No first collective agreement reached’. Raids tend to be in large units (average unit size of 94 in the data), but for those cases where the unit size is less than 50 the ‘No first collective agreement reached’ is coded as one for the 1978–1988 period, if no collective agreement is in place and incumbent tenure is less than or equal to 2 years, and coded as one for the 1989–1998 period, if incumbent tenure is less than or equal to 2 years and no agreement could be found at the BCLRB. A raid launched on a unit where no agreement is in place and the incumbent has been certified for 2 years or less is very likely to be a case where no first collective agreement has been reached. Based on the BCLRB’s review of first agreements over the 1978–1986 period, it is not uncommon for a newly certified unit to go 18 months or more without a first agreement. 13. In fact, the section 70 legislation – which became section 137.5 in 1987 – was only potentially available in cases where a party could show that there had been no attempt to ‘engage in serious, good faith bargaining following certification’ (BCLRB Annual Reports). Even in candidate cases, however, the BCLRB was reluctant to impose a contract. As the BCLRB stated, ‘It should be noted that the Board does not automatically impose a first collective agreement on the parties, even where a persuasive application is made’. Indeed, a BCLRB review of the 1978–1986 period found that there were about 10% of newly certified units each year that had not negotiated a first collective agreement by the 18 month point (approximately
292
CHRIS RIDDELL
600 cases in total), yet over the same period only 4 contracts were imposed by the BCLRB (BCLRB Annual Reports, 1978–1986). Moreover, a unit is open for raiding at the six month point. Major changes were made to first contract arbitration in 1996, but these changes are beyond the scope of this study. 14. The reported estimates are the change in the probability of raiding success for an infinitesimal change in each continuous independent variable and, for each dummy independent variable, the discrete change in the probability of raiding success when the dummy variable takes on values of 0 and 1. 15. That is, the situation where union X quasi-mergers with union Y, retaining its identity but in essence becoming an affiliate of union Y (examples include the unions taken over by the CAW among others).
REFERENCES British Columbia Labour Relations Board (various years). Annual Reports. Bronfenbrenner, K. (1997). The role of union strategies in NLRB certification elections. Industrial and Labor Relations Review, 50(2), 195–212. Brotherhood of Locomotive Engineer Newsletter (2000) ‘UTU withdraws from AFL-CIO.’ March. Chaison, G. (1973). The outcomes of multi-union representation elections involving incumbents. Public Personnel Management, 2(4), 435–438. Chaison, G. (1976). The frequency and outcomes of union raids. Industrial Relations, 15(1), 107–110. Corporate and Labour Unions Returns Act (various years). Annual Reports. Demsetz, R. S. (1993). Voting behavior in union representation election: The influence of skill homogeneity and skill group size. Industrial and Labor Relations Review, 47(1), 99–113. Farber, H. S. (2001). Union success in representation elections: Why does unit size matter? Industrial and Labor Relations Review, 54(2), 329–348. Fiorito, J., Jarley, P., & Delaney, J. T. (1995). National union effectiveness in organizing: Measures and influences. Industrial and Labor Relations Review, 48(4), 613–635. Godard, J. (1997). Beliefs about unions and what they should so: A survey of employed Canadians. Journal of Labor Research, 18(4), 621–649. Heneman, H. G., III, & Sandver, M. H. (1983). Predicting the outcome of union certification elections: A review of the literature. Industrial and Labor Relations Review, 36(4), 537–559. Jacoby, S. M. (2001). Unnatural extinction: The rise and fall of the independent local union. Industrial Relations, 40(3), 377–404. Jarrell, G. A., Brickley, J. A., & Netter, J. M. (1988). The market for corporate control: The empirical evidence since 1980. Journal of Economic Perspectives, 2(1), 49–68. Jensen, M. C. (1988). Takeovers: Their causes and consequences. Journal of Economic Perspectives, 2(1), 21–48. Krislov, J. (1954). Raiding among the ‘legitimate’ unions. Industrial and Labor Relations Review, 8(1), 19–29. Krislov, J. (1955). The extent and trends of raiding among American unions. Quarterly Journal of Economics, 69(1), 145–222.
The Nature of Union Raiding
293
Maranto, C., & Fioritio, J. (1987). The effect of union characteristics on the outcome of NLRB certification elections. Industrial and Labor Relations Review, 40(2), 225–240. Odewahn, C., & Scott, C. (1989). An analysis of multi-union elections involving incumbents. Journal of Labor Research, 10(2), 197–206. Rajin, G. S., & Grigoleitis, W. (1978). A study of the independent local union in Canada. Labour Studies Journal, 2(3), 220–230. Riddell, C. (2001). Union suppression and certification success. Canadian Journal of Economics, 34(2), 396–410. Riddell, C. (2004). Union certification success under voting versus card-check procedures: Evidence from British Columbia. Industrial and Labor Relations Review, 57(4), 493–517. Riddell, C., & Riddell, W. C. (2004). Changing patterns of unionization: The North American experience. In: T. A. Kochan & A. Verman (Eds), Unions in the 21st century: An international perspective. London, U.K.: Palgrave Macmillan. Roll, R. (1990). Empirical evidence on takeover activity and shareholder wealth. In: J. L. Lowenstein & S. Rose-Ackerman (Eds), Takeovers and contests for corporate control. New York: Oxford University Press. Thomason, T. (1994). The effect of accelerated certification procedures on certification success in Ontario. Industrial and Labor Relations Review, 47(2), 207–225. Troy, L. (1960). Local independent and national unions: Competitive labor organizations. Journal of Political Economy, 68(5), 487–506. Walker, J. M., & Lawler, J. J. (1986). Union campaign activities and voter preferences. Journal of Labor Research, 7(1), 19–40.
‘‘LIFTING THE VEIL’’ ON ANTIUNION ACTIVITIES: EMPLOYER AND CONSULTANT REPORTING UNDER THE LMRDA, 1959–2001 John Logan ABSTRACT For over thirty years, one of the most overt forms of employer opposition to unionization has been anti-union campaigns conducted by union avoidance consultants. As a result, both union and employer associations have attempted to influence the provisions of the LMRDA that cover consultant activities. This article provides the first comprehensive historical analysis of the LMRDA’s reporting and disclosure requirements covering employers and consultants. The first section examines consultant reporting policy from the late 1950s to the late 1970s, a period when unions filed relatively few complaints and the DOL initiated few investigations, but the consultant industry expanded significantly. Section two examines developments in the 1980s – the period of greatest congressional and judicial activity on consultant reporting since the 1950s. The final section looks at post-1980s events and examines why organized labor has persisted with its campaign to reform government policy on consultant reporting, despite its inability to make progress on the issue over the past four decades.
Advances in Industrial and Labor Relations, Volume 15, 295–332 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15007-6
295
296
JOHN LOGAN
INTRODUCTION Over the past few years, the Department of Labor (DOL) has sought to impose more stringent financial disclosure requirements on unions under the Labor-Management Reporting and Disclosure Act (LMRDA or LandrumGriffin Act) (Daily Labor Report, 2002a, 2002b, 2002c). Congress passed the LMRDA in 1959 to deal with the problem of union racketeering and corruption. In recent decades, union corruption has been much less of a problem than it was in the 1950s. But unions believe that the Bush Administration has used the law to impose burdensome and intrusive rules as crude ‘‘political payback’’ for their campaigning against the president. The new rules require unions with revenues of at least $250,000 per year to file detailed expenditures with the DOL covering all costs over $5,000. The DOL is also expanding its financial audits of unions – with the appointment of an additional 48 full-time staff dedicated to audits – though its critics argue that there is no evidence of an increase in infringement of the law (Greenhouse, 2005). At the same time, unions argue that the Bush Administration has neglected another critical aspect of the law: the provisions that require financial disclosure from employers and consultants who persuade employees to reject unionization. In contrast with the lack of evidence of union abuses, the American Federation of Labor-Congress of Industrial Organizations (AFL-CIO) claims that there exists ‘‘strong, documented evidence that employers and consultants deliberately flout the intent of the LMRDA’s reporting requirements that apply to them’’ (U.S. House of Representatives, 2002). For over 30 years, one of the most overt forms of employer opposition to unionization has been anti-union campaigns conducted by union avoidance consultants. As a result, both union and employer associations have attempted to influence the provisions of the LMRDA that cover consultant activities. Most previous studies of consultant reporting have focused on the case law and issues surrounding the scope of the so-called advice exemption of the LMRDA, which allows employers and consultants to avoid filing reports under certain circumstances (Aaron, 1960; Afra, 1980; Beaird, 1965, 1970, 1986; Braude, 1985; Bush, 1986; Case Notes, 1967, 1983, 1986; Donahue, 1962; Farmer & Powers 1960; Lawrence & Williams, 1983; McLaughlin and Schoomaker, 1979; O’Reilly, 1962–1963). A few scholars have examined the impact of consultant activities on the outcome of organizing campaigns (Kaufman & Stephan, 1995; Lawler, 1984). But this paper provides the first comprehensive historical analysis of the LMRDA’s reporting and disclosure requirements covering employers and consultants. The first section examines
‘‘Lifting the Veil’’ on Anti-Union Activities
297
consultant reporting policy from the late 1950s to the late 1970s, a period when unions filed relatively few complaints and the DOL initiated few investigations, but the consultant industry expanded significantly. In the early 1960s, the DOL established a policy that justified its inactivity on consultant reporting and disclosure for the next two decades. Section two examines developments in the 1980s – the period of greatest congressional and judicial activity on consultant reporting since the 1950s. Throughout the 1980s, organized labor attacked the ‘‘do nothing’’ policy of the Republican DOL that drastically reduced enforcement of consultant reporting, while significantly expanding its investigations of union corruption and racketeering. The final section looks at post-1980s events, including the limited actions of the Clinton DOL, and examines why organized labor has persisted with its campaign to reform government policy on consultant reporting, despite its inability to make progress on the issue over the past four decades. Organized labor’s failure to revise consultant reporting policy illustrates the deep divisions over labor law reform in the United States and the ability of employer organizations to block the modernization of labor policy. It also offers a stark illustration of how labor law has become ossified and out of step with the current reality of labor-management relations (Estlund, 2002). As the size and sophistication of the consultant industry has grown, the effectiveness of the law on consultant disclosure and reporting has diminished. The LMRDA has permitted aggressive consultant campaigns to flourish and has failed to expose their activities to public view, which was the intention of Congress in enacting the law. DOL interpretation and enforcement of the LMRDA, and the broader industrial relations context of that enforcement, has been just as important as the substance of the law. In the absence of labor law reform, successive governments have influenced consultant reporting through critical variations in the interpretation of the LMRDA, political appointments to key administrative agencies, and the allocation (or lack thereof) of resources and manpower. Republican Administrations have been more effective at using these indirect strategies to advance a ‘‘pro-employer’’ agenda on consultant reporting than have Democratic Administrations in using them to promote a ‘‘pro-union’’ agenda. In the 1980s and 1990s, the Reagan and Bush Administrations significantly narrowed the range of consultant activities that required reporting, appointed conservatives to key positions within the DOL, cut back on resources for the investigation and enforcement of consultant reporting, and increased funds for investigations into union corruption. Republican-controlled Congresses, meanwhile, have subjected the actions of Democratic DOLs to greater scrutiny than their Democratic
298
JOHN LOGAN
counterparts have subjected the actions of Republican DOLs. In the 1990s, the Republican Congress attacked the Clinton DOL’s efforts to step up enforcement of the reporting provisions of the LMRDA. The DOL’s failure to enforce rigorously consultant reporting has also contributed to the widely held impression among unionists and their allies that U.S. labor law condones aggressive employer and consultant opposition to unionization. In recent years, unions’ principal labor law problem has been the failure of the NLRA to offer stronger protection for the right to organize. But organized labor has continued to focus on the consultant reporting provisions of the LMRDA because it believes that this draws attention to the intense hostility with which most employers confront efforts by their employers to form unions.
PART ONE: THE DEVELOPMENT OF CONSULTANT REPORTING POLICY UNTIL THE LATE 1970S As unions have stressed, Congress was concerned not only with union corruption when it passed the LMRDA. It enacted the law, in part, to tackle the questionable conduct of some employers and their consultants, which may have interfered with employees’ right to organize and bargain collectively (Jacoby, 1997).1 Congress insisted that, while not strictly illegal, the consultant activities were ‘‘disruptive of harmonious labor relations’’ and should be ‘‘exposed to public view’’ (Senate Report, 1959). Henry Solano, Solicitor of Labor during the Clinton Administration, argued that by exposing consultants’ business dealings, Congress had hoped ‘‘both to deter persuader activity and to enable employeesy to know the true interests of a consultant hired to persuade employees’’ to reject unionization. Solano believed that Congress had viewed consultants as ‘‘inherently suspect and their activities as subject to abuse.’’2 By imposing financial reporting requirements, however, the provisions of the LMRDA were intended to disclose, not regulate, the activities of consultants and employers. Section 203r of the law requires consultants and employers to file reports when the employer hires a consultant to undertake activities ‘‘where an object thereof, directly or indirectly’’ is to persuade employees to reject unionization. But Section 203 created an exemption that states that employers and consultants are not required to file reports if the consultant’s activities are restricted to giving ‘‘advice’’ to the employer. The precise scope of the advice exemption became the focus of considerable controversy for the next few years, as the statute made no attempt to
‘‘Lifting the Veil’’ on Anti-Union Activities
299
define the term. Some commentators immediately recognized the ‘‘interpretive challenges’’ posed by the vague language of Section 203. The Bureau of National Affairs warned that the advice exemption was ‘‘susceptible of several different interpretationsy It is questionable whether the exemption would cover payments to a consultant who drafted anti-union lettersy to combat union organizing’’ (Bureau of National Affairs, 1959). In 1960, the DOL told employers that they were required to report ‘‘any arrangement’’ with a consultant, even if the consultant simply drafted speeches or written material (U.S. Department of Labor, 1960). Some DOL officials even suggested that employers and consultants were required to file reports if the consultant simply revised documents prepared by the employer. In 1961, the Commissioner of the Bureau of Labor-Management Reports John L. Holcombe wrote that situations where a consultant or lawyer revised a document for an employer intended to ‘‘persuade employees as to how to votey would require the filing of a reporty It follows, a fortiori, that an arrangement whereby a person undertakesy to prepare material, the direct or indirect object of which is to persuade employees [to reject unionization], would have to be reported.’’3 The DOL Solicitor’s office believed that Holcombe had gone ‘‘too far,’’ however, and Holcombe himself later backed off this broad interpretation of the law. In a letter to Solicitor Charles Donahue, Holcombe stated that he would ‘‘have difficultly personally in saying that the concept of advice does not include the revision of drafts.’’ Holcombe conceded that if the DOL were to adopt the position that revising an employer-prepared draft was ‘‘advice,’’ but that preparing an entire draft was an ‘‘activity’’ that triggered the reporting requirements, it ‘‘could be accused of hairsplitting.’’ But he contended that the department ‘‘should draw the line at some point in this shadowy area and I would be willing to draw it there.’’4 Donahue stressed that the DOL needed to adopt a ‘‘reasonable and realistic approach’’ as to what types of activities constituted advice and were therefore exempt from the reporting provisions.5 But adopting such an approach would not prove easy. Indicating the difficulty the DOL faced in distinguishing between exempt advice and reportable activities, one official stated that employers and consultants were required to file reports when it was ‘‘impossible to separate advice from activity that goes beyond advice’’ (Naumoff, 1960). The DOL’s original, broad interpretation of what constituted reportable activity did not survive long, as it soon modified its position under political pressure. The American Bar Association complained that Donahue’s reading of the statute would ‘‘seriously impinge on the attorney–client relationship.’’6 Congressman Philip Landrum (House sponsor of the LMRDA) was
300
JOHN LOGAN
advised of its interpretation and he pressed the DOL for an unequivocal position on the scope of the advice exemption. Landrum told Secretary of Labor Arthur Goldberg that it was clear that the LMRDA did not require that consultants file reports ‘‘regarding the giving of advice.’’7 The Secretary then sent a letter to Landrum endorsing a more expansive interpretation of the advice exemption. But he still maintained that question of what constituted advice, as opposed to reportable activity, was not a straightforward one. Goldberg told Landrum that while the ‘‘drafting of speeches or material for distribution to employees for persuasive purposes’’ did trigger the reporting requirements, it was ‘‘not possible in the absence of precise facts to express a determination’’ about what constituted reportable activity.8 A 1962 memorandum from Solicitor Donahue finally established the DOL’s new position on reportable activity.9 It stated that consultants and employers were required to file reports when consultants prepared material and delivered that material to employees. But they did not need to file reports when consultants simply revised or reviewed the legality of materials prepared by employers. Cases in which consultants prepared entire speeches or documents for the employer to deliver presented a more complex problem. Donahue believed that such actions could ‘‘reasonably be regarded as a form of written advice,’’ providing the employer was free to accept or reject the consultant’s materials. Although the LMRDA interpretative manual stated that the test to determine whether certain actions were reportable ‘‘cannot be mechanically or perfunctorily applied,’’ the DOL significantly restricted its interpretation of the scope of reportable activities after 1962. Donahue explained that the DOL had reviewed its original policy ‘‘in the light of Congressional intent,’’ which had revealed ‘‘no apparent attempt to curb labor relations advice in whatever setting it might be couched’’ (Donahue, 1962). Despite adopting this more restrictive policy, the DOL initiated several actions against consultants and employers for failure to report over the next several years. Between 1961 and 1968, the DOL filed 18 civil actions to enforce the consultant reporting provisions. But that number dropped precipitously over the next decade. Between 1969 and 1978, the DOL filed only one civil action to force compliance (U.S. House of Representatives, 1984a).10 Reasons Behind the Lax Enforcement of Consultant Reporting From the late 1960s to the early 1980s, unions repeatedly complained that the DOL’s enforcement of the consultant reporting requirements fell far
‘‘Lifting the Veil’’ on Anti-Union Activities
301
short of what was required by law. Several factors contributed to the lax enforcement of Section 203. First, as mentioned above, the DOL retreated from its original broad definition of ‘‘persuader activities’’ in 1962. Neither the statute nor the legislative history had illuminated the ‘‘advice–persuasion’’ distinction, so the DOL insisted that its narrow interpretation was entirely justified.11 Second, the budget of the old Bureau of LaborManagement Reports – the Office of Labor-Management Enforcement’s (LMSE) predecessor – was reduced beginning in 1964 (Beaird, 1965). It was merged with the Bureau of Pension-Welfare Reports, and the smaller staff stopped cross-checking National Labor Relations Board (NLRB) records. In the early 1960s, the DOL had voluntarily reviewed NLRB decisions and sought reports from companies that appeared to have spent large sums of money to defeat unions. Owing to a lack of resources, however, the DOL was pursuing only the ‘‘most outrageous’’ offenders by the late 1970s and it no longer reviewed NLRB decisions seeking new leads to investigate. As a result, enforcement of the consultant reporting provisions generally suffered. In 1980, under pressure from unions, the LMSE acknowledged that this had been a productive source of information and again undertook to review NLRB files; but DOL staff were vague about the extent of this activity and suggested that their efforts would be hampered by an extensive backlog of cases waiting investigation. In addition, the DOL claimed that information gleamed from NLRB cases was usually years out of date and thus of limited use.12 Third, complaints from the public about consultant violations were virtually nil for many years in the 1960s and 1970s, and since enforcement occurred only in response to specific complaints, the DOL took little action.13 Complaints to the DOL increased somewhat after the health care amendments to the NLRA in 1974 – which extended collective bargaining rights to most private health care employees – as health care proved a lucrative sector for both unions and management consultants.14 Complaints then increased sharply during 1979–1980, as organized labor focused its attention on union avoidance consultants after the defeat of the Labor Law Reform Act in 1977–1978. Fourth, most consultants and law firms simply stopped reporting in the 1970s. Many consultants transferred (or claimed to have transferred) their activities from reportable to non-reportable conduct. Several federal court decisions in the 1960s and 1970s that interpreted Section 203 as requiring reports from consultants concerning all their labor relations services if even one activity was reportable discouraged many consultants from complying at all (Bush, 1986).15 Other consultants simply refused to file reports or
302
JOHN LOGAN
misunderstood the reporting requirements. One study in the late 1970s concluded that many management lawyers engaged in reportable activity without filing reports with the DOL and others displayed a ‘‘shocking lack of knowledge’’ of their obligations under the statute (Craver, 1978). Another study of consultant activities in California found that less than one percent of employers and consultants engaging in reportable activity actually filed reports with the DOL (Lawrence & Williams, 1983a). Several consultants confirmed these conclusions. In the early 1990s, one former consultant wrote that he had never filed LMRDA reports, despite his involvement in hundreds of union campaigns, as it was easy to ‘‘slide under the scrutiny’’ of the DOL (Levitt, 1993). Fifth, two court decisions significantly increased the burden of proof on the DOL in filing any court action to compel employer or consultant reports. First, the U.S. District Court ruled that the DOL could not rely on NLRB findings to compel a report. Next, the Appellate Court ruled that the DOL could not compel an employer to admit on a report what it had denied before the NLRB.16 In addition, the DOL exerted less control over Section 203 litigation than over other forms of litigation – such as that arising from union elections – because there was no statutory requirement to sue promptly. The Department of Justice (DOJ) theoretically handled all litigation concerning Section 203 violations, and consultant cases often idled at the DOJ for months. But unions also complained about the DOL’s laggard investigations, and worried that this would result in the wholesale dismissals of cases that were finally brought to court. Many cases were not investigated until several years after they were opened and some remained on the DOL’s docket for over three years. Where the DOL refused to initiate an investigation until after the NLRB had concluded its proceedings, the case was often too old to be effectively pursued or the DOL would contend that the NLRB had given the union all the relief to which it were entitled. In response to union criticisms, the DOL claimed that when a case was referred to the Solicitor’s Office, enforcement was vigorous and determined. The problem, according to the DOL, was that only half a dozen or so cases a year were so referred. The penalty for willful violations of the reporting requirements under the LMRDA was a fine of $10,000 and imprisonment for up to one year. As a policy the DOL did not make criminal referrals in consultant reporting cases, however, and instead sought civil relief, thereby hoping to encourage reporting rather than punish noncompliance. But DOL officials claimed that they would seek criminal damages if a consultant failed to abide by a civil judgment that reporting was required (Daily Labor Report, 1984a).
‘‘Lifting the Veil’’ on Anti-Union Activities
303
The NRTWC, MPA, ‘‘Legal Advice,’’ and Union Avoidance Seminars Consultants and law firms were not the only ones providing employers with union avoidance advice. By the 1970s, a variety of employer and trade organizations were also advising employers and thus may have been subject to the reporting requirements of the LMRDA. But organizations such as the National Right to Work Committee (NRTWC) claimed that they did not need to report their activities under Section 203. In 1972, the United Auto Workers (UAW) tried unsuccessfully to compel the DOL to investigate the NRTWC and its Legal Defense Foundation. The union claimed that the Committee was furnishing counsel to or otherwise financing litigation by employees against labor organizations, and soliciting employers for contributions for this purpose; thus, the organization was required to file financial reports. But the DOL declined to take any action against the NRTWC as it generally addressed itself to the ‘‘general public’’ and not to particular employees. If the NRTWC furnished a specific packet of union avoidance materials to an employer upon the employer’s request, the DOL argued, reportability would depend on the nature of the materials. A sample letter for the employer to use in developing anti-union propaganda would be considered advice; material directed to employees and bearing the NRTWC’s name, in contrast, would be considered persuasion. After receiving advice from Solicitor of Labor Richard I. Schubert, the Secretary of Labor stated that he did not believe that ‘‘contributions by employers’’ to the NRTWC ‘‘assume the status of an agreement or arrangement’’ within the meaning of the LMRDA. Nor did the DOL believe that the LMRDA ‘‘suggests a massive reporting program such as might result if all contributors to associations, or dues paying members of employer associations, were required to report.’’17 The UAW fared no better in the courts. The D.C. Circuit Court ruled that it had no right to file a private action against the NRTWC for noncompliance, as only the DOL had the right to enforce the law.18 In contrast with its position on the NRTWC, the DOL filed suit against the Master Printers of America (MPA) in the 1970s, claiming that its ‘‘Craftsmanship’’ program constituted reportable persuader activity.19 The Craftsmanship program provided employers with leaflets containing antiunion statements and advised them to contact the MPA when faced with an organizing campaign. The MPA’s activities were not illegal, but the DOL argued that its conduct fell within the scope of the reporting provisions. The MPA was a consultant within the meaning of the LMRDA, and it had failed to file reports from 1973 through 1977, the period during which it had
304
JOHN LOGAN
received payment from employers for union avoidance services. But the Courts disagreed with the DOL’s interpretation. In 1981, the U.S. District Court for the Eastern District of Virginia ruled that an employer association that distributes anti-union monographs under an agreement with its members does not engage in persuader activity and is exempt from the reporting requirements. Foreseeing ‘‘serious First Amendment implications’’ if the DOL’s interpretation were to prevail, the Court ruled that the arrangements between the MPA and its members were not the ‘‘kind that the reporting requirementsy were intended to illuminate.’’ When the DOL did not appeal the decision, unions feared that this signaled a return to a more limited definition of ‘‘persuader activity’’ under the LMRDA (Daily Labor Report, 1981).20 The DOL also had to rule on the reporting obligations of management attorneys who explained the law to employees for fear that employers who talked directly to them would unwittingly violate the law. This was common practice during organizing campaigns at smaller companies. Reportability in these cases depended on whether the information provided to employees was ‘‘factual’’ or ‘‘highly colored’’ against unionization. Attorneys exposed themselves to the risk of reportability whenever they spoke directly to employees, but most failed to file reports. For the most part, the DOL declined to take action in such cases in the 1960s and 1970s. The DOL explained that it did not require reports ‘‘where the services rendered are intended toy instruct employees on the general subject of labor law.’’21 It also considered the rapidly expanding business of union avoidance seminars as falling outside the bounds of reportable activity because there was no written agreement between those running the seminars and those attending. Consultants used seminars to generate income and recruit new clients. Much to unions’ chagrin, the DOL appeared to consider the principal message of these seminars as ‘‘do workers good and you won’t have a union.’’ Developments on consultant reporting in the 1970s were not all bad for organized labor. After their failure to break a Senate filibuster on the Labor Law Reform Act in 1978, unions turned their attention to union avoidance consultants.22 The reason behind labor’s shift in focus was clear: unlike NLRA reform, a change in the DOL’s interpretation of consultant reporting did not require congressional approval. At the urging of the AFL-CIO, the Subcommittee on Labor-Management Relations of the House Committee on Education and Labor held eleven days of hearings on ‘‘Pressures in Today’s Workplace’’ in December 1979. Although the hearings included testimony on topics such as surveillance and privacy at the workplace, their real focus was on the activities of consultants, particularly on those of the
‘‘Lifting the Veil’’ on Anti-Union Activities
305
most notorious firms, such as Illinois-based Modern Management Methods (3M) and California-based West Coast Industrial Relations Association (WCIRA) (U.S. House of Representatives, 1979). In late 1979, the Carter DOL raised the priority of its field investigations into complaints by unions that consultants and employers were failing to file reports. Shortly after the ‘‘Pressures’’ hearings, the DOL established a targeting-prioritizing strategy for enforcement, with the aim of concentrating its resources on the most notorious consultants. To assist field offices with investigations, the DOL appointed James Vaughan, Branch of Special Investigations, as designated National Coordinator of Investigations and Herbert Raskin, Chief, Branch of Interpretations and Standards, as National Office Coordinator for Interpretations and Analysis (Daily Labor Report, 1980a). The reorganization soon paid dividends, as the DOL was investigating 150 complaints by June 1980, compared to about 30 the previous year. Much to the frustration of organized labor, however, the Carter Administration rejected a proposal to narrow the range of consultant activities covered by the advice exemption. The Solicitor of Labor believed that advice, ‘‘even when embedded in a persuasive speech or statement prepared by the advisor, nevertheless remains advice which need not be reported unless the advisor himself uses the speech or statement directly to persuade employees.’’23 Thus, consultants could still orchestrate union avoidance campaigns from behind the scenes and avoid filing reports with the DOL so long as they made no direct contact with employees. The Issue of Limited Resources Limited resources consistently hampered the DOL’s investigative capabilities. The unprecedented increase in new cases in the early 1980s, according to the DOL, was placing a ‘‘severe strain’’ on the resources of the LMSE.24 In fiscal year (FY) 1980, the LMSE received 423 consultant-related complaints, compared with around 40 in the previous year (Lawrence & Williams, 1983b). Of the 423, between 60 and 70 were dismissed due to lack of merit, and 8 or 9 were referred to litigation. The LMSE could handle only about 100 cases per year due to the small staff in its national office and in the field. DOL officials stated that they would investigate the remaining complaints insofar as its resources allowed. In 1980, only about 10 percent of the department’s 190 investigators were assigned to Section 203 enforcement because union elections were a priority due to the specific statutory investigatory timetables. The future outlook for the LMSE budget in the early 1980s, was even bleaker, with one official calling its finances a
306
JOHN LOGAN
‘‘catastrophe’’ (Bernstein, 1980). Thus, any increase in DOL activity in 1979–1980 was not the result of greater resources being devoted to consultant reporting and enforcement. According to Assistant Secretary of Labor, William Hobgood, an increase in union complaints with credible supporting documentation had stimulated the upsurge in DOL investigations. Hobgood believed that unions were increasingly sensitive to what constituted improper consultant activities and that the ‘‘scope and sophistication’’ of the union avoidance industry had increased substantially in the 1970s.25 Much of the recent increase in consultant activity was in the South, according to acting LMSE director Richard Hunsucker, but all regions of the country had been affected (U.S. Department of Labor, 1980). In light of the greater size of the consultant industry and its growing sophistication, unions argued, the DOL urgently needed to reevaluate its interpretation of the advice exemption and intensify its enforcement efforts. Unions believed that a deficiency of resources and manpower at the DOL were only part of the problem, however, and complained that the department was neglecting its duty to ensure that employers and consultants file reports. Even if vigorous enforcement were to begin under the current DOL interpretation of the LMRDA, they argued, the purpose of Congress in enacting its reporting provisions would not be met. The DOL’s construction of the statute was much narrower than the language and legislative history mandated and this had resulted in widespread evasion of the purposes of the reporting requirement. In several cases in the 1960s and early 1970s, the DOL had required consultants and employers to file reports, not only when consultants made direct contact with employees, but also when consultants had used supervisors to acquire information on the union sentiment of employees and prepared anti-literature for employers to distribute to employees.26 But that practice had ceased by the late 1970s. Some Carter DOL officials shared unions’ skepticism concerning the prevailing DOL interpretation and supported a more expansive interpretation of the reporting requirements. In 1980, Assistant Secretary of Labor for Labor-Management Relations William Hobgood stated that it was ‘‘beyond dispute’’ that consultants and employers should be filing more financial reports. Hobgood was ‘‘troubled’’ by the current position that consultants were not required to file reports if they prepared material for an employer and trained supervisors in union avoidance techniques but did not talk directly with employees. If this interpretation were stretched to its logical extreme, Hobgood explained, it would permit a consultant to ‘‘prepare and orchestrate the dissemination of an entire package of persuader material while sidestepping the reporting requirement merely by using the employer’s
‘‘Lifting the Veil’’ on Anti-Union Activities
307
name and letterhead or avoiding direct contact with employees’’ (Daily Labor Report, 1980b). Other DOL staffers were unequally unhappy with weak enforcement of the consultant reporting provisions. Herb Raskin, Chief of the Branch of Interpretations and Standards, had cooperated offthe-record with Georgetown law professor Charles Craver in the research for his article, which labeled enforcement of the reporting provision, ‘‘benign neglect personified’’ (Craver, 1978).27 The advice–persuasion distinction was clearly the central issue for unions and for the LMSE as it reevaluated its interpretations and brought new litigation against employers and consultants. The DOL’s position was of the utmost importance because it alone can enforce this part of the statute; private parties such as unions cannot bring suit against consultants and employers to compel disclosure of information.28 By the late 1970s, consultants were hiding behind employers, the reverse of the situation in the late 1950s, when the law was enacted. The DOL believed that it needed to ‘‘look hard’’ at union avoidance campaigns orchestrated by consultants behind the scenes, the deployment of supervisors as front-line troops, and ‘‘spontaneous’’ employee anti-union committees. These new tactics were, unions claimed, a perversion of the advice exemption, and many consultants did not even understand their reporting obligations. Consultants frequently excused their failure to comply with the reporting requirements by arguing that the law was complex and vague. Assistant Secretary Hobgood admitted that, in certain instances, ‘‘the wisdom of Solomon might be helpful in determining’’ whether consultants were required to file reports. Hobgood maintained that these difficult cases were rare, but he was pessimistic about the impact of a change of interpretation of the reporting rules, stating that ‘‘even if we take a hard line, any sophisticated consultant can devise a way to evade’’ reporting their activities. The detailed information on consultant activity required to determine whether consultants and employers were required to file reports meant that tremendous resources would be required to administer the law effectively. Moreover, as the LMRDA was aimed at disclosure not regulation of consultant campaigns, the DOL’s ability to stop activities that interfere with workers’ rights would always be limited.29 Thus, from the early 1960s to the late 1970s, the principal obstacles to efficient Section 203 enforcement were a lack of both resources and political will, and unions believed that the situation would only get worse under the Reagan DOL in a budget-cutting climate. More hopeful signs for unions included the DOL’s apparent willingness to reevaluate the advice–persuasion distinction, more careful targeting of its enforcement efforts (focusing on the ‘‘worst’’ offenders), and DOL initiation of new litigation against
308
JOHN LOGAN
employers and consultants. Unions sought to maintain a high volume of meritorious complaints against consultants and supply research assistance where the DOL either could not or would not investigate their complaints. The labor movement’s most important goal, however, was to achieve a more expansive definition of ‘‘persuader activities,’’ since it believed that this alone could discourage, or at least illuminate, many anti-union actions that had previously proceeded unhindered.
PART TWO: DEVELOPMENTS IN THE 1980S In the 1980s and early 1990s, the Reagan and Bush Administrations influenced government policy through a subtle reinterpretation of the rules on consultant reporting, political appointments to key DOL agencies, and the allocation of resources and manpower for the investigation and enforcement of policy on consultant reporting. Unions repeatedly condemned the ‘‘do nothing’’ policy of the Republican DOL on consultant reporting and argued that it was deliberately turning a blind eye to employers and consultants who violated the LMRDA, while simultaneously intensifying its monitoring of union financial activities. The ‘‘Dotson Memo’’ and Agency Reorganization Prior to 1982, the DOL had occasionally adopted the position that ‘‘indirect persuader activity’’ (in which consultants avoided face-to-face contact with employees) triggered the reporting requirements of the LMRDA. The DOL had filed suit against consultants and employers in two cases in 1975 and 1981 in which there was no direct contact between consultants and employees; in these cases, consultants had used supervisors to gather information on union sympathies and disseminate anti-union materials. The DOL argued that the involvement of the consultant was ‘‘sufficiently great’’ to fall within the scope of the reporting requirements.30 Both cases were settled and reports were filed.31 Unions claimed that these cases demonstrated that the DOL could achieve results without expensive and protracted litigation. Thus, if employers and consultants believed that the DOL was serious about enforcing the law, voluntary compliance would follow. Union hopes for increased enforcement were soon dashed. In March 1982, at the instruction of Assistant Secretary of Labor Donald Dotson, the DOL published a reinterpretation of the law under which indirect persuader activities were explicitly excluded. Issued without public notice or comment,
‘‘Lifting the Veil’’ on Anti-Union Activities
309
the memo stated that, with the DOL’s limited resources and large number of open consultant cases, officials must give priority to cases ‘‘involving issues with legal precedent. Examples would include direct contact involving persuader activities or information gathering or coercion by consultants.’’32 Deputy Assistant Secretary of Labor Ronald J. St. Cyr maintained that cases involving indirect consultant contact had ‘‘no legal precedent’’ and should be closed.33 Because the coordination of union avoidance campaigns through front-line supervisors was the most prevalent form of consultant activity, the DOL’s new policy effectively eliminated from reporting and disclosure a major form of anti-union campaigning. And because LMSE enforcement had been so slight since 1959, precedent supported only the narrow interpretation of Section 203 criticized by organized labor. Thus, LMSE inactivity was effectively responsible for the current lack of legal precedent, but the agency was refusing to investigate cases that lacked precedent. The Dotson memo also stated that new cases would only be initiated in response to a specific complaint from the public. Prior to 1982, the LMSE manual suggested that it could initiate investigations of employers or consultants based on complaints, NLRB records, news media reports and other resources (Daily Labor Report, 1984a). As a result of the new policy, the DOL would eschew much of its investigative power under the law and would not of its own initiative review NLRB decisions, even those that had revealed clear reportable activity. Unions complained that this ‘‘no complaint’’ directive contradicted June 1980 and September 1982 ‘‘LMSE Enforcement Strategy Documents’’ stating that NLRB liaison was an important aspect of enforcement of consultant reporting. Reinitiated in response to the ‘‘Pressures’’ hearings, NLRB liaison had generated several investigations in which LMRDA violations were found. Unions complained that some regional DOL offices would not investigate complaints if the NLRB were already doing so, arguing that the board’s case rendered LMRDA enforcement unnecessary, while other regional DOL offices would not pursue a complaint unless there was an NLRB record to rely upon. Dotson’s memorandum had the effect of closing dozens of pending ‘‘failure to report’’ cases and strictly limiting the opening of new ones. Many of the closed cases had been opened between 1979 and 1981 by the Carter DOL, but had generated little investigative activity until their closing. The post-Dotson memorandum policy reduced the pending consultant reporting caseload by about 65 percent; approximately 99 cases were closed (U.S. House of Representatives, 1984a).34 The DOL robustly defended its decision, with one official insisting that many of the discontinued cases had
310
JOHN LOGAN
consisted of ‘‘vague, contradictory, and uncorroborated opinions and suppositions’’ of employees. Dotson claimed that cases were closed because the Carter DOL had opened them on the basis of ‘‘discredited’’ legal theories. Thus, his instructions had merely ensured that the DOL pursue ‘‘straightforward established statutory theories, and theories that had been accepted by the courts.’’35 After the positive developments under the Carter Administration, unions believed that the Dotson memo indicated a clear return to the DOL’s earlier restrictive interpretations of Section 203. But they feared that the Reagan DOL intended to retreat from even the historically restrictive view of what activities triggered the LMRDA’s reporting requirements. Dotson had effectively ‘‘drawn the curtain’’ on consultant reporting by identifying the line between advice and persuasion as the line between indirect and direct conduct.36 The limited revival in DOL interest in effective enforcement during the Carter years, unions concluded, had virtually ceased under the Reagan Administration. Any increase in DOL activity in the early 1980s was due not to the commitment of department officials, but to the renewed dedication to Section 203 enforcement of unions, who had been much more aggressive in filing complaints against consultants and employers. By 1983– 1984, however, the number of complaints received by the DOL had declined sharply, much to the disappointment of the Building and Construction Trades Department (BCTD) of the AFL-CIO. In the early 1980s, the BCTD and the UAW had mass-distributed complaint forms for consultant and employer compliance to union locals, but this strategy did not have the desired result of putting pressure on the DOL by generating hundreds of new complaints. The number of new cases based on union complaints increased from 14 in 1977 to a high of 424 in 1980. But by 1983 the number had fallen to only 30 new complaints (U.S. House of Representatives, 1984a). The Reagan DOL did not only influence government policy through its restrictive interpretation of the rules on consultant reporting. It also used an internal reorganization at the DOL in 1984 to advance its conservative agenda. As part of the reorganization, Deputy Assistant Secretary of Labor Mario Lauro was placed in charge of employer and consultant reporting. Lauro had arrived at the DOL from the NLRB, where he was aide to archconservative, John Van de Water (Gross, 1995), and unions believed that he was ‘‘part and parcel of the Dotson–St. Cyr school of ideological politics.’’ According to Richard Hunsucker, Director, Office of Labor Management Standards Enforcement, Lauro was a right-wing ideologue who ‘‘could pass for Dotson’s brother’’ and had direct connections to the
‘‘Lifting the Veil’’ on Anti-Union Activities
311
White House.37 The combined effect of the DOL’s reinterpretation of the reporting requirements and its internal reorganization, which had promoted several hard-line conservatives to key positions, was to decrease further reporting by employers and consultants. Enforcement of the consultant reporting requirements had practically ground to a halt by the mid-1980s – all during a time when, according to organized labor, employers and consultants were ever more actively, boldly, and creatively fighting unionization. The 1984 House Subcommittee Hearings Even under the Reagan and Bush Administrations, organized labor made some limited progress on consultant reporting. Under the leadership of its new chairman, Representative William Clay (D-Mo), the House Subcommittee on Labor-Management Relations held hearings on the DOL’s enforcement of consultant reporting and disclosure in 1984. The hearings were more narrowly focused than the 1979–1980 ‘‘Pressures’’ hearings, which had included detailed testimony on consultants’ union avoidance activities. In February 1984, the Subcommittee examined labor’s charge that the DOL was unevenly enforcing the LMRDA’s provisions by sharply upgrading enforcement efforts against unions while neglecting consultant reporting. The Democratic majority on the Committee considered this a promising issue because they believed that, under the new interpretation of the advice exemption, the Reagan DOL was effectively challenging Congress to protect the integrity of its process from faceless, unelected bureaucrats. The Subcommittee obtained information that it believed could be used to considerable advantage in litigation against the department for non-enforcement of consultant reporting (though the AFL-CIO complained that the DOL was failing to provide the Subcommittee with adequate access over its files). The Subcommittee’s investigation was intended to hold hearings and propose legislation to strengthen enforcement of the employer and consultant reporting requirements. While they would not result in more favorable legislation, unions believed that the oversight hearings would at least generate publicity on consultant activities. The AFL-CIO recognized that Congress was not about to act on the consultant reporting in the near future, but it believed that keeping the issue alive through the hearings might set the stage for reform under a future Democratic Administration. Its principal goal was to provide the Subcommittee with ‘‘horror stories’’ of consultant campaigns that would personalize the DOL’s ‘‘egregious failure’’ to enforce the reporting provisions. The only potential drawback of this strategy was the possible expansion of the hearings to cover union reporting and
312
JOHN LOGAN
disclosure, and the AFL-CIO believed that the Reagan DOL was already using this issue to attack unions, in any case. At the hearings, House Democrats criticized the DOL’s lopsided enforcement against unions as opposed to employers and consultants. DOL officials offered a rather lackluster defense of their enforcement practices, with Richard Hunsucker, Director of the Office of Labor Management Standards Enforcement, admitting that his office had started cross-checking employer and consultant reports ‘‘the day before yesterday.’’ (On hearing his comment, the room erupted into derisive laughter.) Unions had been pushing the DOL for many years to cross-reference reports, whereby the department would ensure that an employer had filed a report where a consultant had filed a report pursuant to an agreement with an employer, and vice-versa. Hunsucker attributed the DOL’s poor performance to murky legal precedent, limited resources, and ‘‘numerous competing priorities,’’ such as probes of the misuse of union funds. Hunsucker admitted that he was troubled by the decision to close cases not based on complaints in which consultant or employer violations had already been established by agency investigators. The Subcommittee attacked this decision as ‘‘an extravagant waste of limited resources’’ (Daily Labor Report, 1984a). Hunsucker’s testimony demonstrated that LMSE staff members were no longer ‘‘disturbed’’ by the DOL’s restrictive interpretation of reportable activity, as Assistant Secretary of Labor William Hobgood had asserted in 1980 (Wall Street Journal, 1984). The Subcommittee concluded that there had been a ‘‘dramatic increase’’ in employers’ use in consultants during the previous two decades (U.S. House of Representatives, 1984b). However, several major consulting firms, whose activities had been extensively documented in the ‘‘Pressures’’ hearings had never filed financial reports and the DOL had failed to prosecute firms who violated the reporting requirements (Daily Labor Report, 1984a). The LMSE had closed several investigations against 3M despite finding numerous LMRDA violations, and West Coast Industrial Relations Associates (WCIRA) had refused to report all its clients, which was required by law if the firm had engaged in reportable activity in any single union avoidance campaign. But the DOL had also declined to take action against WCIRA. Nor had the DOL ever sought to assess the full dimensions of the consulting industry, claiming that it lacked the resources and expertise to do so, but it acknowledged that the industry appeared to have grown significantly over the past decade. At the same time as the number of consultant firms expanded and the industry grew in sophistication, the Reagan DOL cut back on resources and manpower dedicated to the investigation and enforcement of consultant
‘‘Lifting the Veil’’ on Anti-Union Activities
313
reporting and increased resources for investigation of union finances. The period between FY 1980 and FY 1984 saw a 93-percent decline in resources allocated to consultant reporting (U.S. House of Representatives, 1984b). As a result, resources devoted to enforcement were equivalent to less than one staff person-year in FY 1983. The same four-year period saw large budget and staff increases devoted to investigations of the embezzlement of union funds. In 1983, 198 employer and consultant reports were filed under Section 203 of the LMRDA; the same year over 71,000 union reports were filed (U.S. House of Representatives, 1984b). Critics of the Reagan DOL accused it of ‘‘criminalizing’’ the regulation of unions, while ‘‘systematically dismantling’’ enforcement of consultant reporting (U.S. House of Representatives, 1984a). While its dereliction of duty was not part of some grand anti-union conspiracy, the Subcommittee conceded, the DOL had failed to enforce the law in a balanced manner. Not only had it ‘‘arbitrarily’’ reinterpreted the consultant reporting provisions, the DOL had failed utterly to enforce its revised interpretation of the law. According to the AFL-CIO, the hearings simply confirmed what unions had known for some time: DOL enforcement of the LRMDA was woefully inadequate and it had engaged in a pattern and practice of abandoning its enforcement obligations generally. Since the Reagan Administration assumed office, the DOL had adopted the position that consultants and employers were not required to file reports when a consultant assisted an employer’s union avoidance efforts through the manipulation of supervisors but engaged in no direct fact-to-face contact with employees. In the DOL’s view, indirect persuader activity was not subject to the reporting requirements of the law. Unions contended that the DOL’s post-1982 interpretation of the reporting provisions had encouraged improper consultant activities and flew in the face of congressional intent and fair public policy. By the mid-1980s, they claimed, the DOL had ‘‘abdicated its responsibility’’ to enforce the law’s consultant reporting provisions, while simultaneously turning its financial and manpower resources almost entirely on audits and embezzlement investigations of unions. As a result, the AFL-CIO was having difficulty convincing its affiliates that the DOL ‘‘takes seriously its obligation to enforce’’ the LMRDA.38 The Republican Response Reagan officials and congressional Republicans rejected the Subcommittee’s conclusions and defended DOL policy on consultant reporting. Attacking the Subcommittee’s ‘‘hyperbolic report,’’ its Republican minority stressed
314
JOHN LOGAN
that the DOL had filed 12 civil cases against consultants and employers for noncompliance since 1980, compared with only four cases during the previous decade (U.S. House of Representatives, 1984b). Solicitor General Charles Fried complained that unions and their congressional allies had attributed to Congress ‘‘an intention to enact a very intrusive and possibly overboard statutey in an area that involves First Amendment and attorney–client values.’’39 NLRB Chairman (and former Assistant Secretary of Labor) Donald Dotson refuted the accusation that the DOL was lax on consultant reporting. Promoting a radically different interpretation of the reporting requirements than that pursued by Carter DOL, Dotson insisted that the previous administration had squandered scarce enforcement resources by ‘‘developing strained and convoluted theories’’ on consultant reporting. Dotson, in contrast, believed that the main thrust of the LMRDA was against union racketeering and that Congress had inserted the consultant provision as an afterthought merely to give the impression of symmetry to the legislation (Daily Labor Report, 1984b). Nor was Dotson alone in this view. When asked about DOL inactivity on consultant reporting, its Inspector General J. Brian Hyland stated that he was ‘‘trying to assess whether or not the need is still there; whether [consultant reporting] is serving any useful purpose; if anyone is reading [the reports]y Or whether it’s just some needless burden.’’ The central issue concerning the lack of action on consultant reporting was, according to Hyland, ‘‘Hey, is it hurting anyone?’’ (Bureau of National Affairs, 1984). Unions and their congressional allies believed that Dotson and Hyland had been influenced by the right-wing Heritage Foundation’s report, Mandate for Leadership: Blueprint for a Conservative Administration,40 which had recommended no funding to enforce employer and consultant reporting, though the DOL strongly denied this (Daily Labor Report, 1984a). The Kawasaki Case In 1989, the Circuit Court dealt another blow to organized labor’s consultant reporting campaign in the Kawasaki case. In 1978, the UAW started an organizing campaign at a Kawasaki Motors Corp. plant in Lincoln, Nebraska. Three years later, an Administrative Law Judge (ALJ) found that during a (rerun) representation election in 1979 the company had discharged one union activist and that supervisors had made coercive statements that the plant would close or relocate in the event of a union victory. The union subsequently lost the election by a vote of 257 to 207 (U.S. Department of Labor, 1981). After the NLRB had found the company guilty of several
‘‘Lifting the Veil’’ on Anti-Union Activities
315
unfair labor practices,41 the union filed a petition with the DOL claiming that Kawasaki and its consultant-lawyers, Tate, Brucker, & Sykes, had engaged in activities that triggered the reporting requirements of the LMRDA (Wall Street Journal, 1989).42 The UAW stated that, among other activities, the firm had devised personnel policies designed to prevent unionization and that Kawasaki had paid supervisors to discourage unionization and threaten employees attempting to organize. The DOL did not dispute the union’s allegations but declined to bring enforcement action against Kawasaki and its attorneys because it was now interpreting the LMRDA as not requiring reporting for the specific activities involved in the case.43 DOL officials referred to the ‘‘trivial nature’’ of the issues, and contended that neither the consultant nor company were required to file reports because the consultant had made no direct contact with employees and had not used a third party to discourage unionization (Daily Labor Report, 1989a).44 The preparation of personnel policies to deter unionization constituted written advice under the Section 203, the DOL concluded, as the company was free to accept or reject the material. As the law prohibited it from filing a private right of action against Kawasaki or its lawyers, the UAW filed suit against the DOL for failure to bring enforcement action against the firm and its attorneys for refusing to file reports. The UAW argued that the LMSE field office had found seven instances of reportable activity. Joined by the Workers Defense League and the Center to Protect Workers’ Rights, it accused the DOL of arbitrary and capricious behavior in its failure to enforce the reporting requirements of the LMRDA. The labor movement believed that the Kawasaki case would highlight the DOL’s non-enforcement of the consultant provisions of the LMRDA. It was the perfect vehicle to negotiate in settlement talks a new DOL approach to consultant reporting and enforcement. The AFL-CIO wanted to keep the case alive until after the 1984 presidential election because, in the unlikely event that Walter Mondale won, it could use the case to pressure the DOL to step up enforcement against consultants and employers. If Mondale beat Reagan, the AFL-CIO believed, the Kawasaki case would become a ‘‘terrific lever,’’ even if labor lost the case. If, as expected, Mondale lost the election, the outcome of the case would not make much difference to DOL policy. In addition, there was little chance of an adverse precedent coming out of the case; and by appealing an adverse decision, unions could hardly make the law in this area any worse than it was already. In the early 1980s, the UAW had almost lost the right to file suit against the DOL in the Kawasaki case. In November 1983 and again in November 1984, Judge Greene of the D.C. Circuit dismissed a lawsuit brought by
316
JOHN LOGAN
the UAW and the others on the grounds that the court had no jurisdiction to order the DOL to enforce the reporting provisions against any particular employer or consultant. Unions protested that Greene’s decision effectively granted the DOL an administrative veto over congressional enactments, and the Court of Appeals agreed. In February 1986, the Court of Appeals reversed the D.C. Circuit and held that although unions and other private parties could not challenge any particular enforcement decision, they could challenge the DOL’s interpretations of the reporting provisions that were rendered in the course of making individual enforcement decisions. Thus, the UAW was entitled to proceed with its lawsuit against the DOL. The initial decision in the Kawasaki case provided a significant victory for organized labor. In February 1988, after a five-year lawsuit brought by the UAW, the Center to Protect Workers’ Rights, and the Workers’ Defense League challenging the DOL’s enforcement policy, the Court rejected the broad interpretation of the advice exemption, finding clear legislative history that Congress was concerned with behind-the-scenes manipulations of employees by consultants.45 On remand, Judge Harold Greene of the D.C. District Court stated that contrary to the DOL’s restrictive reading of Section 203, the law was ‘‘hardly ambiguous’’ and ‘‘all persuader activities must be reported’’ (Daily Labor Report, 1988). The labor movement believed that Greene’s decision could have substantial positive ramifications, especially under a future Democratic Administration that was prepared to devote real enforcement resources to consultant reporting and disclosure. Since the early 1960s, the DOL had rarely enforced Section 203 in order to give vigor to a broad interpretation of consultant reporting, but, after the Kawasaki decision, that was now at least a possibility. The AFL-CIO immediately announced that consultant activities that had for decades ‘‘been hidden from public view must now be reported’’ (AFL-CIO, 1988). The labor movement knew that the DOL would appeal the decision but hoped that a new Democratic Administration might influence the litigation. While preparing for the appeal, the AFL-CIO explored ways to bring pressure on the DOL to comply with the ruling and asked its affiliates to monitor the behavior local DOL officials. The House Labor-Management Subcommittee also sought to put pressure on the DOL to comply with the ruling. The UAW and other groups fighting the Kawasaki case remained optimistic about the eventual outcome. They had persisted with the case for several years and poured so many resources into fighting it because they believed they might ‘‘get something’’ from the D.C. Circuit. With a sympathetic panel of judges, unions might win the case. Even with an unsympathetic panel, the case would likely get a ‘‘decent burial.’’
‘‘Lifting the Veil’’ on Anti-Union Activities
317
In August 1988, the AFL-CIO received the names of the judges who would hear the government’s appeal of the District Court decision. The Court of Appeals panel consisted of Judges Abner Mikva, Ruth Ginsburg, and Thomas Hogan (Daily Labor Report, 1989b). Unions considered it about as good a panel as they could have expected. None of the judges were Reagan appointees to the D.C. Circuit. Mikva and Ginsburg were Carter appointees; Ginsburg was on the panel that had reinstated the UAW’s lawsuit in 1986 after its original dismissal. Although Reagan had appointed Hogan to the District Court in 1983, he was well respected as a fair and competent judge and was not considered a right-wing ideologue. Unions believed that the panel strengthened their prospect of obtaining a favorable decision, but recognized that prior Supreme and Circuit Court precedent regarding deference to administrative agencies (such as the DOL) would act as a constraint on the judges. They still hoped that a Democratic victory in the 1988 presidential election would compel the DOL to enforce the law in accordance with the District Court ruling. But their hopes were dashed by George Bush’s victory, which made inevitable the continuation of the DOL’s appeal of the District Court decision. Both sides recognized the high stakes in the Kawasaki case. Six national business organizations filed friend-of-the-court briefs stating that the District Court’s decision would have a ‘‘chilling effect on the ability of employersy to obtain [union avoidance] advice and on the attorney members who provide it’’ (Daily Labor Report, 1989a).46 Oral arguments before the Court of Appeals in January 1989 offered cold comfort for either unions or management. The DOL told the Court that its reinterpretation was justified under Section 203, which specifically exempted from reporting the giving of advice by a consultant. But Judge Hogan observed that the DOL’s post-1982 interpretation of advice read out of the statute virtually all consultant activities. Unions fared little better. Judge Mikva suggested to AFL-CIO counsel Larry Gold that the outcome of the case would make no practical difference to the outcome of labor struggles and suggested that unions were seeking a statute that Congress had not passed – one that would disclose how consultants were paid in all union avoidance campaigns (Daily Labor Report, 1989b). Outcome of Kawasaki Case The Court’s final ruling proved a significant setback for organized labor. In March 1989, the U.S. Court of Appeals for the District of Columbia reversed the District Court decision and ruled that the statute was ‘‘silent or ambiguous’’ with the respect to the interpretative issues on consultant reporting.47
318
JOHN LOGAN
It found that the DOL had ‘‘rationally construed the statute,’’ and thus the Court would defer to its construction as reasonable in the context of the case. Judge Ruth Ginsburg explained: ‘‘Recognizing the Secretary’s right to shape her enforcement policy to the realities of limited resources and competing prioritiesy we reject the challenge to her ruling’’ (Daily Labor Report, 1989c). But the Court did not clearly draw the line between reportable and non-reportable conduct. Thus, consultant orchestration of union avoidance campaigns that allowed no opportunity for employers to reject their advice might still be considered reportable. But unions recognized that the decision was certain to be characterized by employers and the Bush Administration as the Court’s rigid direct contact–indirect contact line between reportable and non-reportable conduct. Thus, the DOL would devote few resources to consultant reporting regardless of the breadth of the Court’s ruling. And unions saw little purpose in seeking review of the decision either by the Court of Appeals en banc or the Supreme Court. Neither the AFL-CIO nor the UAW published press releases or otherwise initiated media attention to the Kawasaki decision. The outcome did not completely undermine labor’s efforts to reform consultant reporting policy, however. Although the decision represented a major setback for unions, it did at least mean that a future Democratic DOL was free to take a different view on the appropriate interpretation on consultant reporting and disclosure. Shortly after the Court of Appeal’s Kawasaki ruling, Acting Deputy Secretary for Labor-Management Standards Mario Lauro issued a memorandum that guided the actions of OLMS staff for the next decade. The ‘‘Lauro memorandum’’ stated that while there was ‘‘no purely mechanical test’’ for determining which consultant activities were reportable, an ‘‘usual indication’’ that the consultant was exempt from reporting was that the consultant had made no direct contact with employees and had merely provided the employer with materials that the employer was then free to accept or reject.48 By the early 1990s, unions were facing hostile employers, a hostile NLRB, and a hostile DOL that had no intention of acting against employers and consultants for failure to file financial reports under Section 203 of the LMRDA, except in the most egregious of cases.
PART THREE: DEVELOPMENTS IN THE 1990S AND 2000S After the election of Bill Clinton in 1992, consultant reporting and disclosure assumed a lesser priority for organized labor. With the Democrats in
‘‘Lifting the Veil’’ on Anti-Union Activities
319
control of the White House and Congress for the first time since the late 1970s, the AFL-CIO believed that it had a realistic opportunity to revise the NLRA. First, it tried for a second time (after its failure to overcome a Senate filibuster in July 1992) to gain legislation outlawing permanent striker replacements. After receiving (lukewarm) support from the Clinton White House, the striker replacement bill passed the House in 1993, but failed to win 60 votes in the Senate the following year (Logan, 2004). Then, under the blue-ribbon presidential Commission on the Future of WorkerManagement Relations (Dunlop Commission), organized labor attempted to gain recommendations for card check recognition, first contract arbitration and stiffer penalties for unfair management practices. At the Commission’s hearings, unions claimed that sophisticated consultant campaigns were undermining the right to organize and that employers were paying exorbitant fees to consultants. Commission members agreed that firms spend ‘‘considerable internal resources and often hire management consultantsy. at a considerable cost’’ (Commission on the Future of WorkerManagement Relations, 1994). They also acknowledged that consultants might contribute to adversarial labor relations. Once again, however, unions’ hopes for reforms that would eliminate (or drastically reduce) the role of employers and consultants in the representation process were dashed. After the Republicans won control of Congress in November 1994, the Commission’s final recommendations (which were far from organized labor’s liking) were effectively ‘‘dead on arrival’’ (Logan, 2007). After its failure to gain NLRA reform, the AFL-CIO again turned its attentions to consultant reporting, as it could pursue this issue under a Republican Congress. First, it pressed the DOL to commit greater resources and manpower to consultant reporting. Second, it argued for a reinterpretation of the advice exemption that would include a broader range of consultant activities. The DOL soon gave a higher priority to its investigations of consultant campaigns, although it provided no more money for enforcement. But it took considerably longer to take action on the advice exemption. The AFL-CIO vigorously lobbied the Clinton Administration for a reinterpretation of the advice exemption during a campaign by the International Brotherhood of Electrical Workers (IBEW) at the Baltimore Gas & Electric Co. (BGE) between 1994 and 2000. Shortly after the IBEW started its campaign in 1994, BGE hired one of the nation’s largest consultant agencies, the Malibu-based Burke Group. The union claimed that BGE had spent over $50 million contesting three rancorous NLRB elections in 1996, 1998 and 2000 (Daily Labor Report, 1998; International Brotherhood of Electrical Workers, 2000), and in 1998 it filed a complaint against the Burke
320
JOHN LOGAN
Group and BGE for failure to report under the provisions of the LMRDA. After the IBEW filed its complaint, the AFL-CIO urged the Clinton DOL not only to take action in the BGE case but also to revise the interpretation of the consultant reporting provisions to prevent this type of abuse of the advice exemption happening in the future. Both bureaucratically and politically, however, consultant reporting proved a thorny issue for the DOL. As a result of resistance from the OLMS and because DOL officials initially failed to appreciate the importance of consultant reporting, the department was reluctant to act on the issue. The 1962 Donahue memo had more-or-less fixed DOL policy on the advice exemption for four decades, and the OLMS had all-but gotten out of the business of consultant reporting. The agency now faced the worst of all worlds – being asked to adopt a politically controversial role without the guarantee of significant additional resources. Unlike the political appointees at the DOL, OLMS staff members are career civil servants, and they did not relish the prospect of being at the center of political controversy over an issue on which they had little expertise. In contrast with the current Bush Administration, which has appointed a former congressional staffer as head of the OLMS, the Clinton Administration appointed as the head of the agency a career civil servant in an attempt to insulate it from accusations of political bias (especially as Republicans viewed the agency as a captive of organized labor). From the point of view of career civil servants, consultant reporting was simply a distraction from the agency’s main responsibilities – and what congressional Republicans considered its true mission – overseeing the election of union officers, and investigating union expenditures and corruption. The agency aimed to conduct the audit of one major union per year. Its resistance to change was not driven by any specific policy disagreement, but by a more fundamental sense of bureaucratic self-preservation. Even more important than opposition from the OLMS was the broader political context of the struggle over consultant reporting. Facing accusations of bias from a hostile Republican Congress after the 1994 mid-term elections, the DOL needed to proceed cautiously. Republicans forced the DOL to devote significant resources responding to requests for internal documents, press hearings, and newspaper stories. In the 1980s, Republican DOLs had faced a Democratic Congress, but its oversight actions were less intrusive than those of the post-1994 Republican Congress. Under the leadership of Representatives Peter Hoekstra, chairman of the Subcommittee on Oversight and Investigations and John Boehner, chairman of the Subcommittee on Employer–Employee Relations, congressional Republicans subjected the actions of the DOL (and the Clinton NLRB) to minute scrutiny.
‘‘Lifting the Veil’’ on Anti-Union Activities
321
The DOL’s reluctance to act on consultant reporting was not so much a question of political will as one of limited resources. If the Labor Secretary were called to appear before the subcommittee to explain the actions of his department, it would take tremendous time and effort to respond to its questions. Simply by making accusations of political bias, therefore, a hostile Congress could do much to prevent the DOL pursuing an agenda of reform. To complicate matters, the relationship between organized labor and the DOL was far from ideal. The first Clinton Labor Secretary, Robert Reich, was a liberal and a close friend of the President, but unlike previous Democratic Labor Secretaries, he had no background in labor relations and no previous connection to the union movement. The relationship between the Clinton DOL and the AFL-CIO improved over time and once engaged on the issue, the DOL constructed a well-crafted case for action, which was essential in the conservative climate of the 1990s. The DOL finally issued a reinterpretation of the advice exemption in January 2001, a few days before Clinton left office. It stated that the ‘‘application of the ‘advice’ exemption depends on whether an activity can fairly be considered giving ‘advice,’ as opposed to engaging in direct or indirect persuasion of employees’’ (Federal Register, 2001a). Thus, employers and consultants would be required to file reports if the consultant prepared materials for the employer, if the object of that material were to persuade employees to reject unionization, even if the consultant had made no face-to-face contact with employees (Daily Labor Report, 2001a). The Clinton DOL offered several justifications for narrowing the advice exemption. First, it believed that the ‘‘textual basis’’ for the existing interpretation, which the DOL had first adopted in 1962, was questionable as it treated consultant revisions of material prepared by the employer as the same as material prepared exclusively by the consultant. The latter was not advice ‘‘in the ordinary sense’’ and in these cases the consultant was functioning ‘‘less as an advisor to the employer than as a persuader of employees.’’ Second, this narrow interpretation of reporting activity had damaged the law’s effectiveness, especially since the consultant industry had grown tremendously in size and sophistication since the late 1950s. Employers’ use of consultants had ‘‘increased significantly,’’ consultants ‘‘may contribute to harmful conflicts,’’ and current policy did not ‘‘reflect the scale or scope of consultant activity’’ (Federal Register, 2001a). The new interpretation of the advice exemption was never implemented. In its first action in the arena of labor-management relations, the Bush DOL first delayed then rescinded the new rule on employer and consultant reporting. The Bush DOL argued that the previous, more expansive
322
JOHN LOGAN
interpretation of what constituted advice, which had effectively guided the actions of successive administrations since 1962, was the ‘‘more appropriate one.’’ Criticizing the Clinton DOL for failing to consult with employer associations, the Bush DOL insisted that the revised interpretation was ‘‘not warranted or justified’’ (Federal Register, 2001b). Stressing that the Bush DOL had issued its notice to rescind the Clinton policy without input from labor organizations, in contrast, the AFL-CIO attacked the reversal as just one of a ‘‘stunning series of moves attacking worker protections’’ (Daily Labor Report, 2001b). Employers’ Response Judging from the vigor with which they have resisted any change to the interpretation of consultant reporting, employer associations appear to believe that the information disclosed by such reports could assist unions and deter employers from engaging the services of consultants. Employer organizations vigorously contested the Kawasaki case, disputed the findings of the congressional oversight hearings, and criticized the Clinton reinterpretation of the advice exemption. One critic of the Clinton policy claimed that employers would now get ‘‘less professional help with their union avoidance efforts, thus the unions’ organization efforts will be made much easier’’ (Baird, 2001). The Associated Builders & Contractors – an organization that provides union avoidance advice in the construction industry – complained that the Clinton rule would disclose ‘‘the amount of money spent on union avoidance which would then be available to unions to use during organizing campaigns.’’ The International Mass Retail Association believed that the policy would ‘‘dramatically limit the ability of employers to respond to union organizing efforts.’’49 And the Printing Industries of America believed that it would limit employers’ freedom of speech and place them in ‘‘a very bad bind’’ (Daily Labor Report, 2001b). Some union avoidance practitioners, in contrast, have claimed that the information provided by financial reports would prove of strictly limited use for unions. In the mid-1980s, one leading consultant, John Sheridan, stated that he had ‘‘grave doubts’’ that reporting would provide unions with powerful ammunition in organizing campaigns. Sheridan was ‘‘unconvinced’’ that unions could acquire the votes of employees simply by disclosing consultant fees. He believed that, at best, information on fees and services would provide an ‘‘interesting interlude’’ in the course of a heated organizing campaign. Instead, Sheridan suggested that unions were looking for a ‘‘scapegoat’’ in the face of rising organizing defeats in the 1980s (Daily
‘‘Lifting the Veil’’ on Anti-Union Activities
323
Labor Report, 1980c). Another union avoidance specialist, Woodruff Imberman, dismissed the advice offered to employers by many consultants as ‘‘hocus pocus.’’ He suggested that employers were paying huge sums of money for consultants campaigns that were, at best, probably worthless and, at worst, counterproductive (Imberman, 1980). Why have Union Persisted in their Campaign for more Vigorous Enforcement on Consultant Reporting? After almost three decades of pushing for greater government action on consultant reporting, the labor movement has achieved relatively little: congressional oversight hearings in 1979 and 1984, brief periods of heightened enforcement activity under the Carter and Clinton DOLs, a favorable ruling by the D.C. District Court that was subsequently overturned by the Court of Appeals, and, perhaps, some greater awareness among policymakers and the public concerning the obstacles facing workers attempting to organize. During the same period, most experts believe that the union avoidance industry has expanded significantly in size and sophistication. By the 1990s, over seventy percent of employers facing organizing campaigns hired outside consultants (Bronfenbrenner, 1996), and consultants are not the only ones providing union avoidance advice: law firms, industrial psychologists, employer groups, trade associations and anti-union organizations offer similar services. So why has consultant reporting remained so important an issue for unions? It has occupied their attentions for two main reasons: the impact of consultant activities on the outcome of organizing campaigns, and the strategy of using consultants as a means of drawing attention to aggressive opposition to unionization. By the 1970s, according to Alan Kistler, AFLCIO director of organization, consultants were ‘‘creating a Frankenstein monster’’ that was damaging the ‘‘entire fabric of labor-management relations’’ (Imberman, 1980). The typical employer strategy was to keep consultants hidden behind the scenes and carry out union avoidance efforts through supervisory staff. Few employers and consultants filed reports about their arrangements, fees, and expenditures for these campaigns because the DOL did not require that they be reported. This failure to expose persuader activities had adversely affected numerous organizing efforts. Unions believed that ‘‘lifting the veil’’ on consultant arrangements, including the disclosure of the terms of their agreements and the amounts of money involved, would exert a strong deterrent on the entire union avoidance industry. While reporting and disclosure would not prevent employers from
324
JOHN LOGAN
engaging in anti-union activities, it might deter ‘‘improper’’ conduct and give workers a better chance at obtaining union representation (Bernstein, 1985). In response to employer claims about unions seeking a scapegoat for their failures, unions stressed that consultant activity had increased significantly since the 1950s, and consultants had assisted employers in defeating hundreds of organizing campaigns (Bronfenbrenner, 1996; Goldfield, 1987). But unions were, in part, seeking to use consultants as ‘‘scapegoats’’ for the more deep-rooted problem of aggressive employer opposition to unionization. Unions believed that the American public and many policy-makers were not ready to hear the message that employers in general were undermining the right to organize. Thus, they needed a scapegoat to blame for the precipitous decline in union membership, and consultants were ideally placed to fulfill that role. Unions hoped that their focus on consultant activities would lead to a national debate on the need for stronger protection for the right to organize, though this strategy has enjoyed limited success, at best. Some unionists remain skeptical about the utility of consultant financial disclosure. Not only has organized labor failed to gain a more favorable policy on consultant reporting, these unionists argue, it would not make much practical difference, even if it did. Consultant reports would not be available immediately50 (or, indeed, in the same financial year, if employers and consultants resisted filing reports) and thus would be of limited use in current organizing campaigns. These unionists believe that consultant reporting is a peripheral issue to the principal policy challenge facing organized labor in the twenty-first century – how to gain stronger legal protection for the right to organize. Thus, unions needed to push for comprehensive reform of the NLRA, not devote resources to lobbying for a more favorable interpretation of Section 203 of the LMRDA. Most union officials, in contrast, believe that consultant financial information would be ‘‘extremely useful at any time.’’51 It would help in several contexts: in the event of a rerun or second election, as many large bargaining units are organized only after several attempts; when unfair labor practices had blocked the election for a substantial period of time; in future campaigns against the same employer at different locations and in campaigns involving different employers using the same consultant; and in situations in which initial bargaining rounds were protracted or in the event of subsequent decertification attempts. In particular, unions believe the financial disclosure information could neutralize the dues and assessment issue, the ‘‘outsider,’’ issue and the high union salary issue – all of which are staples of consultant campaigns. Consultants claim that unions are interested in employees only as a source of revenue, that unions are external ‘‘third parties’’
‘‘Lifting the Veil’’ on Anti-Union Activities
325
that do not share or understand the concerns of employees, and that dues money will pay for the exorbitant salaries of ‘‘fat cat’’ union officials. Financial disclosure would expose consultants as highly paid ‘‘outsiders’’ and reveal the often-large sums of money that employers spend to defeat organizing campaigns. In addition, unions could use consultant reporting not only in waging conventional organizing drives, but also in corporate campaigns and in other creative applications of pressure. Thus, unions believe that effective disclosure could help rebut consultant campaigns and impair the effectiveness of the consultant industry more generally.
CONCLUSION The polarized debate over consultant reporting shows the depth of the impasse over labor law reform in the United States. Over the past four decades, unions have struggled to revise consultant reporting policy, while employer organizations have steadfastly and successfully opposed unionsupported revisions. But a lack of reform has not meant a lack of activity on consultant reporting. For four decades, the DOL’s interpretation, enforcement, and the context of that enforcement has been just as important as the content of the law on consultant reporting. In the absence of labor law reform, both Democratic and Republican Administrations have exerted influence through critical variations in their interpretation of consultant reporting policy, their attempts to influence judicial decisions on the issue, and their allocation of resources for investigation and enforcement. Although unions are still attempting to focus public and political attention on consultants, their task is probably more difficult today than it was in the 1970s and 1980s, when consultant campaigns represented a new and, unions argued, disturbing development, which required that the DOL take vigorous action. By the 1990s, in contrast, over 70 percent of employers were recruiting union avoidance consultants and most observers viewed them as a ‘‘regular’’ aspect of recognition campaigns. As a result, unions’ arguments for tougher rules on consultant reporting perhaps do not carry the same weight as they did in previous decades. Unions have persisted with their campaign for more rigorous consultant reporting for one practical reason. Revision of DOL policy on consultant reporting does not require 60 votes in the Senate – the number required to overcome a Republican filibuster – which has proved the greatest obstacle to NLRA reform. As with campaigns to reform the NLRA, however, cohesive employer opposition has frustrated the labor movement’s efforts to revise
326
JOHN LOGAN
consultant reporting policy. At the urging of employer associations, Republican Congresses and Administrations have ensured that consultants remain free to conduct union avoidance campaigns without reporting to the DOL in most circumstances. In the 1980s, the Reagan Administration instructed the DOL to close ‘‘indirect persuasion cases’’ and instead focus its resources on union corruption and racketeering. In the 1990s, the Republican Congress resisted the Clinton DOL’s efforts to enforce more vigorously the reporting provisions of the LMRDA. Immediately after taking office in January 2001, the Bush Administration first delayed then rescinded the Clinton Administration’s last minute policy change on consultant reporting, returning to the old, restrictive interpretation of what constituted reportable activity. In part constrained by the language of the law, Democratic Congresses and Administrations, in contrast, have acted cautiously on consultant reporting, offering limited help to unions. It remains to be seen whether unions can use the issue of consultant activity to reform the LMRDA, influence reporting policy, or deter aggressive opposition to unionization.
NOTES 1. The (McClellan Committee) Hearings that led to the enactment of the LMRDA focused on the activities of one consultant, Nathan Shefferman, and his firm, Labor Relations Associates (LRA). Shefferman started his consultant work with Sears Roebuck & Co., but later built LRA into the nation’s largest union avoidance firm. Shefferman’s firm was forced out of business as a result of bad publicity from the hearings, but it produced dozens of imitators. Former LRA employees included John Sheridan of John Sheridan Associates, Herbert Melnick of Modern Management Methods and Louis Jackson of the law firm, Jackson Lewis. 2. Henry L. Solano, Solicitor, U.S. Department of Labor, letter to Hon. Seth P. Waxman, Solicitor General, U.S. Department of Justice, June 22, 2000. All manuscript sources are from the records of the AFL-CIO legal and public policy departments, Washington, D.C. 3. John L. Holcombe, Commissioner, Bureau of Labor-Management Relations, letter to Allan Topper, February 15, 1961. 4. John L. Holcombe, Commissioner, Bureau of Labor-Management Relations, Memo to Mr. Charles Donahue, Solicitor of Labor, Re: Consultant Reporting under Section 302(b), November 17, 1961. 5. Charles Donahue, Solicitor, Memo to John L. Holcombe, Commissioner, Bureau of Labor-Management Relations, ‘‘Modification of position regarding scope of ‘advice’ under Section 203r,’’ February 19, 1962. 6. F. William McCalpin, Bar Association of St. Louis, letter to John L. Holcombe, Director, Bureau of Labor-Management Reports, September 28, 1961.
‘‘Lifting the Veil’’ on Anti-Union Activities
327
7. Honorable Philip M. Landrum, U.S. House of Representatives, letter to Honorable Arthur J. Goldberg, Secretary of Labor, May 9, 1961. 8. Arthur Goldberg, Secretary of Labor, letter to The Honorable Phil M. Landrum, May 24, 1961; Philip Landrum, letter to Honorable Arthur J. Goldberg, June 7, 1961; Arthur Goldberg, Secretary of Labor, letter to The Honorable Phil M. Landrum, June 29, 1961. 9. Charles Donahue, Solicitor of Labor, memo to John L. Holcombe, Commissioner of the Bureau of Labor-Management Reports, February 19, 1962. 10. Statement of Richard G. Hunsucker on Labor Department Enforcement of Consultant Reporting Provisions of the Landrun-Griffin Act (February 8, 1984). 11. Richard F. Schubert, Solicitor of Labor, letter to Stephen I. Schlossberg, General Counsel, UAW, November 13, 1972. 12. Larry Gold, AFL-CIO, Memo to Jules Bernstein, Robert Connerton and Phillis Payne, Re: Discussion of Labor Consultants at December 11 Meeting of ABA Committee on Development of the Law of Union Administration and Procedure, December 15, 1980; Statement of Richard G. Hunsucker on Labor Department Enforcement of Consultant Reporting Provisions of the Landrun-Griffin Act (February 8, 1984). 13. William P. Hobgood, Assistant Secretary for Labor for Labor-Management Relations, Statement Before House Subcommittee on Labor Management Relations on Activities of Management Consultants (September 3, 1980). 14. Several of the biggest consultant firms of the 1970s and 1980s, including Modern Management Methods and Management Science Associates, specialized in union avoidance campaigns in the healthcare sector. 15. See, for example, Douglas v. Wirtz, 353 F. 2d 30 (4th Cir. 1965), cert. denied, 383 U.S. 909 (1966). However, this interpretation was overturned in the 1985 ruling, Donovan v. Rose Law Firm, 768 F. 2d 964 (8th Cir. 1985). According to the Solicitor General, ‘‘So long as a law firm or consult does nothing more than provide advicey it need not report those non-persuader activities for that client even though it engages in persuader activities for other clients.’’ Charles Fried, Solicitor General, U.S. Department of Justice, letter to Honorable William D. Brock, Secretary, Department of Labor, May 8, 1986. 16. Statement of Richard G. Hunsucker on Labor Department Enforcement of Consultant Reporting Provisions of the Landrun-Griffin Act (February 8, 1984). 17. Richard F. Schubert, Solicitor of Labor, letter to Stephen I. Schlossberg, General Counsel, UAW, November 13, 1972. 18. International Union, United Auto Workers v. National Right to Work Legal Defense and Education Fund, 590 F. 2d 1139 (D.C. Cir. 1978). 19. Donavan v. Master Printers of America, USDC E Va, No. CA-80-1040A, April 15, 1981. 20. The Fifth Circuit Court later ruled that the MPA must file reports when it provided copies of its monthly magazine, which frequently contained articles on union avoidance, to member employers. The DOL also sued the Illinois chapter of the MPA, after the chapter’s executive secretary had made anti-union speeches to employees. He filed a report on these speeches but failed to report the MPA’s other union avoidance services. In a similar case, the DOL recommended litigation in Milledgeville, Georgia, where municipal officials, including the police chief, helped local the union avoidance efforts of local employers. The DOL decided that the officials must report under the LMRDA.
328
JOHN LOGAN
21. Harold A. Nouzoi, Acting Associate Solicitor, letter to Gregory A. Giordano (Clark & Stant, P.C.), December 11, 1979. 22. In the 1970s–1990s, the AFL-CIO published a newsletter on management consultants, the RUB (Report of Union Busting) Sheet, based on information sent by its affiliates. 23. Carin Ann Clauss, Solicitor of Labor, Memorandum for the Secretary of Labor, June 21, 1979. 24. Statement of Richard G. Hunsucker on Labor Department Enforcement of Consultant Reporting Provisions of the Landrun-Griffin Act (February 8, 1984). 25. William P. Hobgood, Assistant Secretary for Labor for Labor-Management Relations, Statement Before House Subcommittee on Labor Management Relations on Activities of Management Consultants (September 3, 1980). 26. Carl Rolnick, LMSE Director, memo to John Murphy, LMSE Assistant Director, September 20, 1973, Re: Employer-Consultant Reporting. 27. Martin R. Ganzglass, Delson & Gordon, letter to Jules Bernstein, Re: AFLCIO Labor Consultants Project, June 15, 1979. 28. International Union, United Auto Workers v. National Right to Work Legal Defense and Education Fund, 590 F. 2d 1139 (D.C. Cir. 1978). 29. William P. Hobgood, Assistant Secretary for Labor for Labor-Management Relations, Statement Before House Subcommittee on Labor Management Relations on Activities of Management Consultants (September 3, 1980). 30. John J. Murphy, Assistant Director, memo to Mr. Carl Rolnick, Director, LMSE, Re: Employer-Consultant Reporting, September 20, 1973. 31. Dunlop v. John Sheridan Associates, Inc. No 75-C-4205 (N.D. III); Marshall v. South Hills Health Systems, No. 81-66 (W.D. Pa.). The first case involved consultant John Sheridan’s efforts to defeat an IBEW organizing campaign at GTE Lenhurt, Inc., an Albuquerque electronics firm. The second case involved Modern Management’s campaign at South Hills Health System, a Pittsburgh hospital. 32. Richard G. Hunsucker, Director, Office of Labor-Management Standards Enforcement, U.S. Department of Labor, Labor-Management Services Administration Notice No. 13-82, March 12, 1982. 33. Ronald J. St. Cyr, Deputy Assistant Secretary of Labor, Memo for Richard G. Hunsucker, Director, LMSE, Subject: Policy Guidance in Title II EmployerConsultant Reporting Cases, March 5, 1982. 34. Randy Rabinowitz and Larry Gold, memo to Stephen Schlossberg, Jules Bernstein, Leonard Page, September 21, 1983, Re: Summary Judgment Motion in UAW v. Donovan. 35. Charles M. Williamson, memo to Jim Greene, Acting Chief, BCI, January 25, 1982. 36. Larry Gold, AFL-CIO, memo to Jules Bernstein, July 12, 1983, Re: LMSA Logs in UAW v. Donovan. 37. Larry Gold, AFL-CIO, memo to Jules Bernstein, October 19, 1984. 38. Charles McDonald, Executive Assistant to the Director, Department of Organization and Field Services, June 10, 1983. 39. Charles Fried, Solicitor General, U.S. Department of Justice, letter to Honorable William D. Brock, Secretary, Department of Labor, May 8, 1986.
‘‘Lifting the Veil’’ on Anti-Union Activities
329
40. Senator Orrin Hatch’s legislative director, Robert P. Hunter, was the principal author of the 1981 Heritage Foundation report. The following year, President Reagan appointed Hunter to the NLRB. 41. 107 LRRM 1541. 42. Stephen I. Schlossberg, Washington Counsel, UAW, John A. Fillion, General Counsel, UAW, Judy Scott, Assistant General Counsel, UAW, letter to Honorable Raymond J. Donovan, Secretary of Labor, June 16, 1982. 43. Richard G. Hunsucker, Director, Labor-Management Standards Enforcement, letter to Polly A. Connelly, UAW, November 29, 1982. 44. Charles M. Williamson, memo to Jim Greene, Acting Chief, BCI, January 25, 1982. 45. United Auto Workers and Center to Protect Workers’ Rights v. Secretary of Labor (U.S.D.D.D.C.) 46. The six business organizations were the U.S. Chamber of Commerce, National Association of Manufacturers, American Society for Personnel Administration, American Retail Federation, Associated Builders and Contractors, Inc., and Master Printers of America. 47. United Auto Workers and Center to Protect Workers’ Rights v. Dole (D.C. Circuit), 678 F. Supp 4 (D.D.C. 1988). 48. Mario A. Lauro, Jr., Acting Deputy Secretary for Labor-Management Standards, Memo to OLMS staff, March 24, 1989. 49. Associated Builders and Contractors, Inc., ‘‘Labor Department Ruling Will Impact Employers’ Efforts to Seek Advice on Labor Issues,’’ January 26, 2001; International Mass Retail Association, ‘‘Labor Department Reinterprets ‘Persuader’ Reporting Rules,’’ January 2001. 50. Section 203 requires consultants and employers to file reports with the DOL within 30 days after they have made any agreement to engage in persuader activities. 51. Marsha Bergon, AFL-CIO, memo to Alan Kistler/Charles McDonald, Re: Enforcement by the Labor Department of LMRDA, April 7, 1980.
ACKNOWLEDGMENTS The author would like to thank John F. Colwell, Fred Feinstein, Deborah Greenfield, Tom Donahue, and the editors of this journal for comments on previous versions of the article.
REFERENCES Aaron, B. (1960). The Labor-Management Reporting and Disclosure Act of 1959. Harvard Law Review, 73(5), 851–907. AFL-CIO (1988). Unions gain major victory in court ruling on reports to U.S. labor department of consultant and employer activity. Press Release, February 8. Afra, J. P. (1980). Persuader activities and the management attorney. Employee Relations Law Journal, 6(1), 3–24.
330
JOHN LOGAN
Baird, C. W. (2001). It depends on what the meaning of ‘advice’ is. Ideas on liberty. The Smith Center for Private Enterprise Studies. Beaird, J. R. (1965). Reporting requirements for employers and labor relations consultants in the Labor-Management Reporting and Disclosure Act of 1959. The Georgetown Law Journal, 53(2), 267–313. Beaird, J. R. (1970). Some aspects of the LMRDA reporting requirements. Georgia Law Review, 4(4), 696–723. Beaird, J. R. (1986). Employer and consultant reporting under the LMRDA. Georgia Law Review, 20(3), 533–563. Bernstein, J. (1980). Enforcement by the Department of Labor of the Employer and Employer Consultant Reporting and Disclosure Act of 1959: A program for change. Washington, D.C.: AFL-CIO. Bernstein, J. (1985). The evolution of the use of management consultants in labor relations: A labor perspective. Labor Law Journal, 36(5), 292–299. Braude, N. (1985). The Labor Management Reporting and Disclosure Act: The extent of disclosure required under Sections 203(b) and (c). Chicago-Kent Law Review, 61(4), 751–769. Bronfenbrenner, K. (1996). Lasting victories: Successful union strategies for winning first contracts. Proceedings of the forty-ninth annual meeting. Madison, WI: IRRA, pp. 161–167. Bureau of National Affairs (1959). The Labor Reform Law 36. Bureau of National Affairs (1984). Employee Relations Weekly Interview. January 9. Bush, R. R. (1986). Consultant reporting requirements under the Labor Management Reporting and Disclosure Act of 1959: How much is enough? Labor Law Journal, 37(3), 180–186. Case Notes (1967). Two views of a labor relations consultant’s duty to report under Section 203 of the LMRDA. Michigan Law Review, 65(4), 752–761. Case Notes (1983). The liability of labor relations consultants for advising unfair labor practices. Harvard Law Review, 97(2), 529–546. Case Notes (1986). Donovan v. Rose Law Firm: Limiting the scope of attorney reporting requirements under the Labor-Management Reporting and Disclosure Act of 1959. Arkansas Law Review, 39(4), 687–703. Commission on the Future of Worker-Management Relations (1994). Fact Finding Report, U.S. Government Printing Press (May), Washington, DC. Craver, C. B. (1978). The application of the LMRDA ‘labor consultant’ reporting requirements to management attorneys: Benign neglected personified. Northwestern University Law Review, 73(4), 624–625. Daily Labor Report (1980a). Labor Department gives higher priority to investigating LMRDA reporting requirements. March 14, A-1. Daily Labor Report (1980b). Consultant reporting requirements being reviewed by Labor Department. June 20, A-9. Daily Labor Report (1980c). Addresses of labor relations consultants presented to New York University’s 33rd National Conference on Labor. June 20, D-1 Daily Labor Report (1981). Labor Department won’t appeal ruling on ‘persuader activity’ registration. August 25, A-1 Daily Labor Report (1984a). Labor Department official defends policies on enforcing LandrumGriffin consultant rules. February 9, A-13. Daily Labor Report (1984b). NLRB chairman Dotson challenges assertion that Labor Department was lax on consultants. February 10, A-8.
‘‘Lifting the Veil’’ on Anti-Union Activities
331
Daily Labor Report (1988). Court orders stricter enforcement of LMRDA reporting requirements. February 8, A-1. Daily Labor Report (1989a). Unions, employers are monitoring closely case on employer consulting activity. January 6, CC-1. Daily Labor Report (1989b). Labor Department seeks holding that employer and law firms are exempt from report requirement. January 18, A-1. Daily Labor Report (1989c). Appeals court sustains labor secretary’s narrow reading of consultant reporting provisions. March 14, A-1. Daily Labor Report (1998). IBEW, Baltimore Gas agree to hold rerun election, settle NLRB claims. June 16, AA-1. Daily Labor Report (2001a). DOL narrows LMRDA exemption shielding labor relations consultants. January 11, AA-1. Daily Labor Report (2001b). Broadened LMRDA reporting requirements for labor relations consultants rescinded. April 11, AA-1. Daily Labor Report (2002a). DOL again looking to revise unions’ annual financial forms. April 11, AA-1. Daily Labor Report (2002b). Bush called upon to push DOL to enforce Landrum Griffin Act. June 21, A-9. Daily Labor Report (2002c). DOL solicitor emphasizes push to monitor union accounting practices. August 13, C-8. Donahue, C. (1962). Some problems under Landrum Griffin. American Bar Association, Section of Labor Relations Law. Proceedings, 45–50. Estlund, C. (2002). The ossification of American labor law. Columbia Law Review, 102(6), 1527–1612. Farmer, G., & Powers, N. T. (1960). The employer reporting requirements and Section 302 of Taft-Hartley as amended. Northwestern University Law Review, 54(2), 782–802. Federal Register (2001a). Department of Labor, Office of Labor-Management Relations, Interpretation of the ‘‘advice’’ exemption in Section 203r of the Labor-Management Reporting and Disclosure Act; Notice. 66(8), January 11. Federal Register (2001b). Department of Labor, Office of Labor Management Standards, Interpretation of the ‘advice’ exemption in Section 203r of the Labor-Management Reporting and Disclosure Act; Notice. 66(70), April 11. Goldfield, M. (1987). The decline of organized labor in the United States. Chicago, IL: University of Chicago Press. Greenhouse, S. (2005). Labor dept. plans increasing scrutiny of union finances. New York Times, April 17, 26. Gross, J. (1995). Broken promise: The subversion of American labor law, 1947–1994. Philadelphia, PA: Temple University Press. Imberman, W. (1980). Hocus pocus in union avoidance. Journal of Labor Research, 1(2), 275–283. International Brotherhood of Electrical Workers (2000). IBEW vows to continue the fight at BGE. IBEW News Release, December 20. Jacoby, S. M. (1997). Modern manors: Welfare capitalism since the new deal. Princeton, NJ: Princeton University Press. Kaufman, B. E., & Stephan, P. E. (1995). The role of management attorneys in union organizing. Journal of Labor Research, 16(4), 439–454. Lawler, J. (1984). The influence of management consultants on the outcome of union certification elections. Industrial and Labor Relations Review, 38(1), 38–51.
332
JOHN LOGAN
Lawrence, A., & Williams, J. (1983a). Union busters and the Law: Consultant and employer non-compliance with reporting requirements of the Landrum-Griffin Act in California. Center for Labor Research and Education, Institute of Industrial Relations, University of California-Berkeley. Lawrence, A., & Williams, J. (1983b). Union busters and the Law: Consultants and employers non-compliance with reporting requirements of the Landrum Griffin Act in California. New Labor Review, 5, 1–23. Levitt, M. J. (1993). Confessions of a union buster. New York, NY: Crown Publishers. Logan, J. (2004). Labor’s ‘last stand’ in national politics? The political campaign for striker replacement legislation, 1991–1995. In: B. E. Kaufman & D. Lewin (Eds), Advances in industrial and labor relations (Vol. 13, pp. 191–244). Logan, J. (forthcoming, 2007). The Dunlop Commission and the impasse over labor law reform in the United States. Journal of Labor Research. McLaughlin, D. B., & Schoomaker, A. L. W. (1979). The Landrum-Griffin Act and union democracy. Ann Arbor, MI: University of Michigan Press. Naumoff, B. (1960). Reporting requirements under the Labor-Management Reporting and Disclosure Act. Fourteenth annual proceedings of the New York University conference on labor. (pp. 240–248). New York, NY: Matthew Bender. O’Reilly, J. A. (1962). Management aspects of the Labor-Management Reporting and Disclosure Act of 1959. University of Detroit Law Journal, 40(2), 240–248. Senate Report (1959). No 85-1684, 85th Congress, 2nd Session, 8–9. U.S. Department of Labor (1960). Bureau of Labor-Management Reports. Technical Assistance Aid No. 4: Guide for employer reporting. Washington, D.C.: U.S. Government Printing Press. U.S. Department of Labor (1980). Labor-Management Services Administration. LMSA Focus, 3(2). Washington, D.C.: U.S. Government Printing Press. U.S. Department of Labor (1981). Labor-Management Services Administration, Report of Investigation, Subject: Kawasaki Motors Corp., John Tate, Labor Relations Consultant, June 19. Washington, D.C.: U.S. Government Printing Press. U.S. House of Representatives (1979). House Education and Labor Committee. Report of the House Subcommittee on Labor-Management Relations, Committee on Education and Labor. Pressures in today’s workplace: Oversight hearings before the House Subcommittee on Labor-Management Relations of the House Committee on Education and Labor. Washington, D.C.: U.S. Government Printing Press (December). U.S. House of Representatives (1984a). House Education and Labor Committee, Report of the House Subcommittee on Labor-Management Relations, Committee on Education and Labor. Staff report concerning enforcement of consultant and employer reporting provisions of the Landrum-Griffin Act. Washington, D.C.: U.S. Government Printing Press (February). U.S. House of Representatives (1984b). House Education and Labor Committee, Report of the House Subcommittee on Labor-Management Relations, Committee on Education and Labor. The forgotten law – disclosure of consultant and employer activity under the LMRDA. Washington, D.C.: U.S. Government Printing Press (December). U.S. House of Representatives (2002). Subcommittee on Employer–Employee Relations, Subcommittee on Workforce Protections, Committee on Education and the Workforce. Testimony of James B. Coppess on behalf of the AFL-CIO. Washington, D.C.: U.S. Government Printing Press (April 10). Wall Street Journal (1984). Labor agency moves fast on House panel report. February 9, 1. Wall Street Journal (1989). Resisting unions: Should companies have to tell more? January 10, 1.
LABORS’ DIVIDED HOUSE: CONTEXTUAL AND THEORETICAL TERMS OF THE U.S. DIVORCE Roland Zullo ABSTRACT The author places the departure by the Change to Win Coalition from the AFL–CIO in contextual and theoretical terms. For context, it is argued that associational rights for U.S. wage-earners have historically and generally been subordinate to the rights of capital owners. As such, the rules regulating industrial relations tend to punish broad acts of wage-earner solidarity, channeling labor toward a strategy of achieving a larger share of the rewards of production through private contracts with employers. This has given birth to business unionism, a style of union representation characterized as exclusionary, neutral with regard to political party, business-like in operation, and accommodative to market capitalism. Presently, the internationalization of capitalism is challenging business unionism by exposing its contradictions and vulnerabilities. As political theory would predict, this is pressuring the AFL–CIO and affiliates to socialize labor–capital conflict. This shift, in turn, resulted in several major points of contention within the house of labor, leading to the departure of the Change to Win affiliates.
Advances in Industrial and Labor Relations, Volume 15, 333–358 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15008-8
333
334
ROLAND ZULLO
With tens of thousands of labor agreements nationwide, many negotiated at the worksite level, the most distinguishing characteristic of the U.S. industrial relations system might be its decentralized structure (Bok & Dunlop, 1970, pp. 208–222). Vertical cleavages based on a skill hierarchy are the norm in professional service industries. Airlines, an extreme example, negotiate with separate unions of pilots, stewards, mechanics, baggage handlers and ticket agents. Collective bargaining at hospitals, schools, and newspapers similarly involve layered representation demarcated along skill or occupational status. Although less common, production and maintenance crews in the same manufacturing centers occasionally belong to separate bargaining units. The U.S. labor relations system also features horizontal cleavages, where organized workers of comparable status perform integrated functions at the same locale. A visible example is the commercial construction industry, where unionized employers become signatories to multiple labor agreements, each one covering a specific trade. Against this backdrop it would appear inconsequential, maybe even internally consistent, for several large unions and some of the building trades to depart from the AFL–CIO to form a rival federation, Change to Win, in September 2005. Much has been written about the circumstances leading to the Change to Win departure from the AFL–CIO and the personalities involved. In this essay I attempt to place events in contextual and theoretical terms by reviewing how the legal development of the U.S. industrial relations system gave rise to a dominant labor strategy, business unionism, based on achieving a larger share of the rewards of production through private contracts with employers. I then outline four characteristics of business unionism and describe how the strategy faltered when the spread of market capitalism exposed its critical contradictions and vulnerabilities. Using political theory, I then explain why the failure of business unionism is pressuring labor to shift toward a strategy that socializes labor–capital conflict. In the final section I argue that this strategic shift has created tensions within the house of labor, prompting Change to Win to depart and embrace the business unionism model.
CONTEXT: ASSOCIATIONAL RIGHTS IN THE U.S. It is instructive to first place collective representation rights in the U.S. in historical context. Associational rights for American workers are the product of an ongoing tension between individual liberties ushered in by political
Labors’ Divided House
335
democracy and the suppression of collectivism among wage-earners by a regulatory regime dedicated to commerce. Origins of this tension trace to the Master and Servant Laws of England that took shape in the 14th century, when the black plague created a tight labor market for artisans and itinerant farmers. To ensure a predictable supply of labor for businesses and landowners, English royalty established legal obstacles for workers seeking to exit an economic relationship. The Master and Servant Law achieved this objective with three precepts: (1) labor markets were regulated by contracts between owners and hires; (2) these contracts were enforced by the judiciary; (3) hires in breach of a contract faced pecuniary and possibly corporeal punishment (Steinfeld, 2001; Tomlins, 2004). England exported this framework to her colonies, leaving indelible fingerprints on common law and municipal decree in the U.S. Far from allowing ‘‘free’’ labor to dictate wages and conditions, the propertied colonist strengthened their hand by enacting ordinances, such as maximum wage laws or restrictions on reemployment without certification, to inhibit the movement of skilled labor (Morris, 1946). Long-term employment contracts became another method for binding workers to employers, replacing the more odious institutions of indentured servitude and slavery as these laws were repealed in Northern states (Steinfeld, 2001, Chapter 8). But political democracy and capitalism followed parallel trajectories.1 As 19th century workers became increasingly conscious of their permanent status as wage-earners, as slavery aggravated the national conscience, and as the demand for skilled labor was ameliorated by immigration and technology, a legal theory was needed that absorbed the expanding rights that non-propertied citizens gained through universal suffrage yet suppressed concerted activity among these burgeoning wage earners.2 A puzzle sat squarely before the courts: how to preserve the Master and Servant canon of individual contracts, cherished instruments by property owners, yet prevent wage-earners from employing this doctrine in collective form? Power shifted markedly toward wage earners when contracts expressed their collective demands and were backed by concerted action. Over protests from Jeffersonian politicians who preferred ideological distance from English traditions, the U.S. courts bowed to Her Majesty by adopting the ancient theory of criminal conspiracy (Nelles, 1931, pp. 168–173; Hattam, 1993, Chapter 2). Literally taken, the criminal conspiracy doctrine outlawed combinations of workers without consideration for their actions or intentions. Initially such a sweeping application was rejected by most U.S. courts (Hattam, 1993, pp. 56–62). Peaceful strikes for higher pay or better conditions were
336
ROLAND ZULLO
often dismissed as provincial affairs, while many guilty verdicts yielded only minor fines.3 With organized labor in its infancy, progress toward harmonizing collective rights with individual liberties came in 1842 when Judge Shaw ruled that associations of workers were constitutionally permissible as long as their purpose and means were not criminal or unlawful. The landmark Commonwealth vs. Hunt legitimized labor unions, the strike, and other peaceful concerted activity, but for labor the victory was transitory. Near century’s end, as unions found leverage in the regional boycott and with the rise in national railroad strikes, the courts embraced a mutated version of the criminal conspiracy doctrine to circumscribe labors’ most effective tactics (Forbath, 1991, Chapter 3; Hurvitz, 1986; Cheyney, 1889). In a series of retrograde decisions dating from 1886 to 1895, the courts hardened around the concept that profits, accrued and prospective, were a form of property and contract a form of liberty, both protected by the constitution. To enforce what Hurtvitz terms ‘‘entrepreneurial property rights’’ and sidestep the decriminalization of union activity in Commonwealth vs. Hunt, courts began to impose compensatory and punitive damages on labor based on tort law, allowing businesses to claim compensable damages based on ‘‘malicious interference’’ by labor organizations (Hurvitz, 1986, pp. 328–332). This ‘‘civil conspiracy’’ method of restraint eventually proved too slow – most unions being able to inflict damage before a trial and verdict – so the courts moved to enjoin labor activities (Perlman, 1922, Chapter 7). A typical injunction required unions to cease their activity, and often even temporary writs were enough to knock the momentum from a union campaign. Stoking the ire of unionists, injunctions were broadly issued against peaceful strikes, ‘‘Do Not Patronize’’ leafletting, and to ban meetings among workers.4 Ironically, as 19th century wage earners gained freedoms to quit, petition their employers for better wages and conditions, or refuse to purchase goods from offending employers, combinations of wage earners doing the same could be served with injunctions, dragged into court for violating the order, fined and incarcerated (Pope, 1987). Behavior tolerated decades earlier when unions were considered a local nuisance became sharply proscribed as organized labor refined their tactics and grew national in scope. To fully appreciate judicial temperament during this era, one needs to juxtapose court treatment of labor organizations with adjudication regarding corporations and the monopolistic trusts. Although the injunction earned the enmity of labor, it was merely a mechanism for enforcing the transformational maxim that entrepreneurial property rights trumped other societal interests. Owners correctly perceived that organized labor posed a
Labors’ Divided House
337
direct threat to this agenda. The influential Knights of Labor pushed an economic plan that rings treasonous by contemporary U.S. standards: establish worker-owned producer cooperatives, socialize the railways, eliminate private banks, abolish prison labor, and so forth. In legal drives led by the railroad industry, owners of large enterprises successfully sought refuge in federal courtrooms, and in a series of cases by the U.S. Supreme Court elevated the corporation to ‘‘personhood’’ status based on a liberal interpretation of the 14th amendment.5 These elite associations of capital owners would henceforth enjoy constitutional privileges that would be denied to associations of wage earners.6 Decades later the New Deal brought sweeping reform that facilitated industrial unionism, and gave impetus to industrial relations as a field of study. Arguably however, it was during the Gilded Era, when corporations became fixed in the American economy and polity, that the line of scrimmage between labor and capital was drawn. Well into the next century, in the contested turf of American politics, labor would seek protective law and collective bargaining rights where they had some influence: the legislative process, while capital would defend against any encroachment against entrepreneurial property rights in that most elite and least democratic branch: the federal courts. Legislation endorsed by labor to humanize work, such as limits on work hours, requirements for employers to pay in legal tender (as opposed to company script), and so forth, were routinely struck down.7 Boycotts, a popular form of concerted activity, were severely suppressed.8 Even laws expressly enacted to regulate corporations were used to proscribe union activity (Kutler, 1962).9 Time and again entrepreneurial property rights, a judicial concoction, overruled societal responses to concentrated private wealth and power, including negating or adulterating protective legislation.10 Socialized tactics to redress perceived inequality were abandoned by many mainstream labor leaders in favor of tactics that focused on securing better compensation and conditions through contracts at the point of production. Even here, however, any legal guarantee to unionize and bargain collectively had to pass strict judicial review. Given the size of the industry and strategic importance to interstate commerce, railroads became the first battleground. In a pattern to be repeated in U.S. history, Congress passed the Erdman Act in 1898, which recognized railway unions and offered a dispute resolution procedure intended to address industry conflict spanning from the Great Railroad strike of 1877 to the Pullman strike of 1894. When in 1908 the Supreme Court invalidated the section of the Erdman Act that
338
ROLAND ZULLO
protected workers against discrimination for union activity, again on the basis of liberty of contract, labor promptly rejected the terms.11 It took the Great Railroad strike of 1922 (400,000 strikers estimated), before President Coolidge called on railway unions and industry leaders to propose rules to regulate union–management relations.12 In 1926 Congress incorporated these ideas in the Railway Labor Act, which survived the test of constitutionality.13 Enabling legislation in most remaining private industries followed a similar path: decades of unprotected mobilizing, punctuated by explosions of illegal concerted activity, awaiting and in part establishing the political preconditions to institutionalize unions and collective bargaining. Yet statutory rights were always less a product of philosophical apostasy than a political reflex designed to calm a restless populace and avoid radical reform. The Wagner Act’s objective, as Section 1 specifies, was to safeguard and promote commerce.14 Strikes and other forms of conflict were viewed as injurious and disruptive to the production and flow of goods. Finding the authority to regulate labor–management relations in the Commerce Clause of the U.S. Constitution, the experiment was to promote the ‘‘friendly adjustment of industrial disputes’’ through collective bargaining. The Supreme Court, which had rejected similar reasoning to invalidate the National Industrial Recovery Act in 1935,15 upheld Wagner in 193716 only after Roosevelt threatened to pack the Court with compliant appointees. In reply, the Supreme Court wasted little time restricting labors’ tactical options under the statute (Klare, 1978; Pope, 2004). One year after upholding Wagner, strikers could be permanently replaced.17 Strike variations involving control over private property (sit-downs), or that interfered with an employers’ ability to replace workers (intermittent strikes), or concerted actions that threatened profits (slowdowns), were deemed unprotected (Becker, 1994).18 A more subtle intrusion into Section 7 was the allowance for anti-union employer behavior (Stone, 1992). In straightforward language, Wagner defined it an unfair practice for employers to ‘‘interfere with, restrain, or coerce employees in the exercise’’ of their right to join a union.19 In 1941, employers were granted the right to communicate with employees about unionization based on a first amendment free speech rationale.20 Change in public sentiment toward unions due to national prosperity, strikes in the coal industry during World War II, and the industrial strike wave of 1946, brought the Taft-Hartley Act of 1947. Substantively, the major provisions of the renamed Labor Management Relations Act, especially the secondary boycott prohibitions and ‘‘free speech’’ proviso, only codified and reinforced established common law. Symbolically the
Labors’ Divided House
339
legislation signaled a rejection of the underlining theory that gave rise to Wagner; the political establishment now perceived unions as a detriment to the economy. With Taft Hartley serving as the imprimatur to interpret the statute to the benefit of industry, the courts would proceed to fill the skeletal framework with rulings that would strip the law of any semblance of symmetry. Without an unimpeachable anchor in the constitution, labors’ demand for associational rights was to yield to the demands made by protected ‘persons’ before federal courts. Thus, when one asks why the U.S. industrial relations system is decentralized; why it lacks elements of corporatism found elsewhere; why its unions are factionalized and discordant, the answer partially lies in this grudging, incremental, and often brutal treatment of collective action. Popular tactics for achieving semi-uniform standards that might flourish in a highly decentralized and unregulated democracy, such as boycotts or legislated protective law, have been repeatedly circumscribed by the courts. Labor was channeled, as Forbath (1991) concludes, into a legitimate role as agents for establishing better terms and conditions through formal labor agreements with individual employers. It was a narrow purpose, but one that could be implemented within the constraints posed by the entrepreneurial property rights doctrine. Accordingly, efforts toward sweeping reform were abandoned in favor of legislation that strengthened collective bargaining. Here, however, an encompassing philosophical statement on associational rights for workers was never established. Instead collective bargaining law evolved as a political compromise to privilege a disruptive class of wage earners at the exclusion of others.21 What exists is a patchwork of statutory rights, prone to legislative amendment or judicial tampering, with a tendency to defer to commercial interests. The absence of a cohesive, national policy on labors’ associational rights is further complicated by the exceptional provisions for classes of workers across the multifarious laws,22 and the community of interest criteria used to determine bargaining units. Diversity in regulatory bodies, rules, and in enforcement has compartmentalized the union movement. For wage-earners, the resultant body of rules presents a hard choice between dubious acts of class-based solidarity versus more certain gains through contract-based unionism.
THE RISE OF A DOMINANT LABOR STRATEGY In obedience to these pressures a federation of ‘‘job conscious’’ trade unionists emerged.23 Influential pragmatists, epitomized by the strong-willed
340
ROLAND ZULLO
Samuel Gompers persuaded workers to accept a position of collective restraint and political voluntarism in order to pursue gains through private contracts with employers (Grob, 1958; Laslett, 1986). The intent here is not to pass judgment on Gompers or his successors; nor is it to dodge the topic of union leadership by implying that legal and economic constraints predetermined their choices.24 Suffice to say, Gompers advocated a brand of unionism more limited in scope than alternatives. Yet to his credit, the American Federation of Labor (AFL) he led for 38 years attracted scores of trade unions and demonstrated longevity despite economic depressions and attacks from the organized business community. Instead the intent is to conceptualize unions as political institutions striving to function within a given socio-economic context, and in doing so explain why the AFL model has either outlived or absorbed all rivals in U.S. history. What earlier writers have pejoratively referred to as ‘‘business unionism’’ is a strategy that holds some logic in a context where federal courts and elected bodies largely serve as organs to the commercial class.25 Of the four aspects of business unionism, perhaps the most defining was the policy of organizing workers that were capable of securing labor contracts. Gompers limited charters to organizations of skilled workers, or to groups such as the miners who enjoyed strength due to the nature of their occupation. Restricting membership to workers that had power at the point of production enabled the AFL to assemble a federation of unions with enough leverage to achieve modest agreements in private negotiations with employers regardless of the law. As other scholars have pointed out, a great deal of labor relations activity occurs beneath the radar of federal and state rules. It is at this level where, without a law sanctioning unions and requiring employers to bargain, what ultimately matters is the ability for unions to inflict concentrated economic hardship on employers. Rapid industrialization following the Civil War, labor-saving technology and scientific management created large numbers of semi-skilled industrial workers. By excluding the semi-skilled from their organization, AFL leaders sought to ensure that their bargaining power would not be diluted by an adjacent, relatively replaceable workforce. Dividing wage-earners into sub-classes along craft is inherently exclusionary, which in turn necessitates protectionism. Craft, which is defined primarily by the skill and tools of members, served as a binding source of identity among AFL affiliates. Skill and tools, however, are fluid commodities that can be simplified or even eliminated by changes in technology, product markets or work design. Both can also be purchased or acquired through methods beyond the control of unions. As such, unions identifying
Labors’ Divided House
341
as craft must endeavor to protect the boundaries of their union using a combination of skill monopolization and industry regulation. Historically skill was safeguarded through the apprenticeship system, while regulation manifested as industry standards26 or broader national legislation, such as immigration restrictions.27 A second characteristic of business unionism is the philosophy of voluntarism, which has come to mean a neutral political affiliation. Voluntarism is not, however, apolitical. Gompers was aware of the connection between politics and the economic well-being of AFL members and commented regularly on public policy (Greene, 1998; Gompers, 1919, 1967; Walling, 1926). Rather, by eschewing direct relations with a single political party, voluntarism firstly asserted the view that collective bargaining should be a private affair. Under Gompers, labor was not seeking to control a major political party, nor attempting to create an independent labor party. Government was expected to repay the courtesy by abstaining from labor– management conflict. Were Gompers able to achieve government neutrality, especially relief from court interference, AFL affiliates would have benefited due to their ability to achieve modest gains through direct contract negotiations with employers in a laissez-faire environment. Secondly, to buttress private bargaining and in an additional display of distrust toward state paternalism, voluntarism meant directing resources toward legislation to support private collective bargaining as opposed to broader protective law. This stance was justified on the belief wage-earners, acting through their unions, best understood their needs. Finally, voluntarism sheltered labor against the vicissitudes of party politics. Sustaining friends in both camps of the two-party system protected the AFL against retribution from a hostile, rising party, while guarding against stale and unresponsive allegiances to a friendly party. In this sense, voluntarism provided the AFL access to Republicans and Democrats, enabling leaders to broker deals on public policy that mattered to members. A third dimension to business unionism is the willingness to structurally limit internal democracy for the sake of organizational efficiency (Mandel, 1954). Michels’ (1962) ‘‘Iron Law of Oligarchy,’’ with elements attributable to his mentor, Max Weber, provides a theoretical explanation. Organizing, bargaining, contract administration, and so forth, involves skill sets that most rank-and-filers lack. To support members and representatives, unions establish departments that provide professional and technical expertise in areas such as law and economic research. As the organization becomes more complex, knowledge of the system, contacts with influential persons, and control over internal communications give incumbents a lopsided edge over
342
ROLAND ZULLO
challengers. Union leaders, for their part, often savor their status and endeavor to protect their roles by bureaucratizing and frustrating internal democratic procedures.28 For Michels, such transgressions from democratic practice are desirable because organizations require stability, not disruptions caused by leadership turnover. Moreover, there is a practical constraint inherent to largeness: most members are apathetic or too busy to meaningfully participate in management (Lipset, 1954). Thus, even for organizations that preach democracy, the overwhelming reality is that a few members aspire to administrate, and it is these few who establish and enforce policy. The fourth and most overarching aspect of business unionism is the promotion of American-styled capitalism. Nationally, Gompers believed in labor–business partnerships (Nyland, 1998), as did his successor, William Green (Phelan, 1989). Internationally, the AFL helped export of the entrepreneurial property rights doctrine into asset-rich areas of the world. A vocal opponent of socialism, Gompers traveled widely to extol the virtues of American capitalism.29 After the merger between the AFL and CIO in 1955, George Meany and other leaders of the U.S. labor movement intensified their collaboration with the U.S. State Department to counter communism abroad (Buhle, 1999, pp. 146–148). This tact served established leaders in three ways. One, it meant an uncritical endorsement of U.S. military intervention, which massaged members’ patriotic sentiments and earned leaders’ esteem as ambassadors to state departments during times of military conflict.30 Two, the relationship with the state created a purpose for labor in international politics even as labors’ influence was waning in the domestic polity. Three, union leaders capitalized on pro-war and anti-communist fervor to harass and expunge rivals both from within and outside their organizations.31 Business unionism, in sum, sought a junior partnership with capitalism. While never fully relinquishing a labor advocacy role, Gompers’ brand of representation was nonetheless sold as a lesser evil compared with the alternative conceptions of working-class collectivism advocating a socialist political economy. Latching on to the legal tradition of liberty of contract, it stands as a pure-and-simple survival strategy in a context where entrepreneurial property rights are the centerpiece of regulatory policy. For unions, state approbation became the quid pro quo for conspiring to internationalize property rights. Consequently, the vitality of business unionism rested on the assumption of ongoing state protection for this model of trade unionism even as economic radicalism, at home and abroad, faded.
Labors’ Divided House
343
FAILED ASSUMPTIONS AND VULNERABILITIES OF BUSINESS UNIONISM The assumption of ongoing state support proved false. With the close of the Vietnam War, collapse of the Soviet Union, and global spread of market capitalism, labor experienced a rapid decline in status and political influence. The Democratic Party, in pursuit of corporate donors, distanced their relations with organized labor; while hostility from the Republican Party escalated. Visions of a labor federation with a meaningful voice in shaping federal policy vanished because organized labor is no longer instrumental to federal policy makers. In retrospect, Gompers’ quest for tripartite involvement materialized faintly under times of military uncertainty during World Wars I and II, and was tenuously extended by AFL–CIO leaders during the Cold War. The geopolitical spread of capitalism and internationalization of U.S.-based corporations since then have made labor obsolete. Hence, statute-based rules for unionizing and bargaining were not preserved and adequately enforced. Disintegration of political influence at the peak has forced labor to redirect their political strategy to the state and local government levels.32 Against employers, pressure to reverse declining membership has led to a return to traditional tactics for organizing workers, with notable new variations on the secondary boycott.33 However, an even more general tactical shift comes from the realization that, under business unionism, the rankand-file have become an underutilized source of political and economic power. While certainly the infamous purges of leftist labor leaders sapped the movement of intellectual energy, these maneuvers were a symptom of a pervasive tendency to undermine rank-and-file militancy by union administrations trying to establish a place in society for a loyal and responsible trade union movement. Centralization became necessary for organizational discipline, which was achieved by neglecting democracy and suppressing independent action at local and regional bodies. By the 1980s business unionism was suffering from an innate contradiction: heavy orchestration at the peak alienates members whose collective resolve is essential to union power. Subdue the impetuous base and forsake the source of leverage necessary to deal with employers or the state.34 Many labor organizations have recently scrambled to undo these effects by upgrading member mobilization (Tillman & Cummings, 1999). Outside labor organizations, exclusionary policies exacerbated gender, ethnic, and racial divisions among workers and led to equivocation
344
ROLAND ZULLO
on social issues important to the working class. Non-white and female wage earners correctly perceived protectionism as an extension of a legacy of discrimination against women, minorities and foreigners.35 Sexism, racism and jingoism repelled movements and thinkers that would otherwise be labors’ natural allies. Add the pro-war and anti-communist behavior that was distasteful to progressives and labor was unqualified to lead or join coalitions capable of building on New Deal social reform. Failure to achieve broad benefits for wage-earners through the political process, coupled with the gains achieved through private contracts, meant a lucrative gap between the price of union and non-union work, which national (e.g. Southern) and foreign competitors steadily exploited. Workers previously excluded from the house of labor eventually permeated occupational boundaries through jobs with non-union employers. The sectors most prone to these changes were those in the higher wage, tradable goods industries. In essence, a labor movement based largely on private contracts with owners at the point of production carries two vulnerabilities. First, by aligning the conditions of unionized wage earners to the economic fate of employers, union leaders are pressed to agree to terms during negotiations, or conspire to lobby for policy in the political arena that supports the financial viability of their employers, even though such actions may harm non-union wage-earners, other unions, or society. Solidarity, moral standing, and social responsibility are sacrificed for compensation and job security. Second, formal unionization becomes the primary vehicle by which wage-earners contract out of an employment-at-will status and improve their work conditions and compensation. While one might suspect that the condition-compensation differential between union and non-union environs should provide an incentive for non-union wage-earners to organize, comparative evidence of unionization trends suggests that such incentives are mitigated by the risks to unionize.36 When barriers to unionize are low, wage-earners are likely to organize to improve their status; when barriers to unionize are high, it means that employers can successfully invest in ‘‘union free’’ tactics to avoid the high incremental price of dealing with a unionized workforce. As Gompers even came to realize, a union movement built on a contract-based strategy depends on the state for law that protects unionization and promotes the consummation of binding agreements (Albert & Palladino, 2003; Greene, 1998). Indirectly, unions also benefit from a state policy that limits labor competition (e.g. restricts immigration) and provides for full employment.
Labors’ Divided House
345
THEORY: UNIONS AS ADAPTABLE POLITICAL INSTITUTIONS To place these developments in theoretical terms, and relate them to the Change to Win departure from the AFL–CIO, it is useful to conceptualize unions as a microcosm of the polity. Like citizens, union members are heterogeneous in their perception of the best use of collective resources. All unions have conservative and liberal protagonists who attempt to sway member-bystanders toward their organizational strategy. When there is economic growth and job security, or when management bargains with unions out of regulatory or economic compulsion, members elect leaders they believe can gain a fair share from the existing system, and conservative materialism reigns. During such periods unions establish a bureaucracy to service members and, as Michels (1962) would predict, internal democracy is ceded to organizational authority that is centered on the function of collective bargaining (Cook, 1963, Chapter 7). However for unions, Michels’ ‘‘Iron Law’’ is malleable. As insiders Golden and Ruttenberg (1942, Chapter 3) tell, radical voices gain an audience when bargaining fails to produce economic benefits or if job security is threatened. During such organizational crises, bureaucratic constraints are challenged, collaborations with management are questioned, and latent democratic tendencies among the rank-and-file are reactivated, often precipitating union leadership change at the local level before rising to the peak (Tillman & Cummings, 1999). New leaders (or old converts) urge either an independent path that consolidates assets or a path that attempts to alter the balance of power between labor and capital by recruiting sympathetic non-union groups to labors’ cause, and thus building a unified alliance against the dominant political and economic order. The latter is what Schattschneider (1960, Chapter 1) refers to as ‘‘socializing’’ conflict. For labor such a path has historically involved compacts with groups that oppose corporate hegemony or that share a common concern for wage-earners. One implication of the socialization of labor-capital conflict is that organized labor will outwardly transform into a more progressive political entity. This occurs because coalitions entail interorganizational agreements that constrain the ability for one entity to pursue policy that conflicts with interests of other coalition members. Given the reform agenda of traditional and prospective non-union coalition partners, such as civil rights groups, interfaith councils, students, environmentalists, and so forth, it follows that labor will become politically engaged toward honoring, if not adopting, the progressive agenda of these non-union coalition partners.
346
ROLAND ZULLO
One indication that labor-capital conflict is being socialized is the progressive turn of the AFL–CIO. In an abrupt reversal from their protectionist past, in 2000 the AFL–CIO announced a position favoring unconditional amnesty for undocumented aliens. Consistent with theory, the resolution did not originate from the peak, but from the base, spurred by a coalition of California unions that relied heavily on the support of organizations sympathetic to immigrants. Similarly, in 2005 the AFL–CIO endorsed a resolution brought forward by a grassroots coalition, ‘‘U.S. Labor Against the War,’’ calling for the rapid withdraw of U.S. troops in Iraq. And unlike a decade ago, outreach to the progressive non-union community is an AFL– CIO priority. To fortify their political support for worker-friendly public policy, the AFL–CIO recently launched the ‘‘Working America’’ program to communicate with non-union citizens. A second sign are labor coalitions that draw from the resources of allies during organizing and representation campaigns (Brecher & Costello, 1990; Reynolds, 2004). Public support can be critical, as was evident in the 1997 strike by the International Brotherhood of Teamsters at United Parcel Service, or the 1989 strike by the United Mine Workers at Pittston Coal (Brisbin, 2002). Hence, we witness the evolving rhetoric behind organizing drives, with the right to bargain collectively advocated for in universal terms, such as ‘‘justice’’ or ‘‘respect,’’ crafted to appeal to non-union audiences. Some unions are circumventing the legal conventions for achieving bargaining rights, rejecting NLRB supervised elections in favor of a cardcheck recognition process that leverages local community support. Once workers are formally represented, pressuring employers to agree to contract terms is less a function of strike capacity and more dependent on corporate campaigns: a broad array of tactics that exert diffuse and multi-directional forms of pressure on industry leaders (Perry, 1987). A third indicator is the organic growth of worker centers outside of the formal union structure. During the recent decades of union decline, groups like ACORN have expanded their services to marginalized workers and achieved social gains through community-based electoral pressure, agitating for living wages, affordable housing, improved schools, and so forth. Unlike most labor unions, worker centers are chartered to provide service in a specific geography, and are not employer or occupation centered (Fine, 2006). They often originate from an ethnic culture or cause, are associational and therefore draw comparatively little financial support from member dues, and their leaders tend to be racial and gender diverse. While some worker centers assist unions during organizing drives, most have a ‘‘longterm transformational social justice agenda’’ (Fine, 2006, Chapter 4).
Labors’ Divided House
347
CHANGE TO WIN: SOCIAL MOVEMENT OR BUSINESS UNIONISM? Several contemporary analysts have noted parallels between the Change to Win affiliates, especially the SEIU, and the early CIO (Fantasia & Voss, 2004; Martin, 2006). There are commonalities to draw from. Then, as now, leaders debated how to organize low income, underserved wage-earners in industries with large populations of minorities and immigrants. Then, as now, leaders differed over the use of resources, representation structure, and on the tactics for membership growth. The major CIO unions, not unlike the present day UNITE, HERE and SEIU, formed coalitions with progressive religious and secular institutions, hired left-leaning, youthful organizers, and engaged in disruptive, unconventional tactics (Keeran 1980; Zeiger, 1995, Chapter 3). These similarities have led analysts to equate the leading unions and architects of the Change to Win split as ‘‘social movement’’ in character (Fantasia & Voss, 2004, Chapter 4). Yet whether the CIO, Change to Win, or any single Change to Win affiliate embodies social movement unionism depends on how ‘‘social movement unionism’’ is defined.37 If the term is used to describe organizational tactics, such as high levels of member involvement in organizing drives, coalition building, and corporate campaigns, then social movement unionism can apply to numerous dramatic events in labor history: sit-down strikes as well as Justice for Janitors (Fantasia & Voss, 2004, pp. 127–131). On the other hand, if social movement unionism is defined as having a ‘‘social change mission,’’ attempting fundamental societal change, such as a ‘‘transformation of the capitalist mode of production’’ (Robinson, 2002), then neither the CIO nor Change to Win classifies. Only in its nascent stage did the CIO represent something radically distinct from the AFL. Beginning around 1939, four years after the founding meeting, an internecine struggle led to the rise of conservative union leaders in the major CIO affiliates (Emspak, 1984; Keeran, 1980; Zeiger, 1995). The major holdout was the UE, which had its origins in the class-oriented Trade Union Unity League (Filippelli, 1984). Lewis, President of the CIO, was an autocratic and politically conservative leader who consolidated his power by firing scores of left organizers that were responsible for much of the CIO organizing (Stepan-Norris & Zeitlin, 2003, Chapter 2). The era saw radical tactics, but not philosophical differences among federation leaders: the CIO walkout (or AFL expulsion) was over the structure and resources needed to organize industrial workers. Unlike Powderly of the Knights of Labor, who believed labor should pursue systemic reform to fundamentally change the
348
ROLAND ZULLO
wage relationship (Oestreicher, 1986), or the socialist convert Debs, Lewis, like his AFL counterpart Green, was interested in labor agreements and member growth. The debates leading to the Change to Win split were also about structure and resources, not philosophy (Hurd, 2004). Moreover, by mapping the tensions that led to the divorce it is plausible to conceptualize Change to Win as a reification of business unionism. For example, a major stress point arose from the fact that occupational groups (and their unions) do not uniformly benefit from regional coalitions. Low wage, immigrant and minority workers invoke more public sympathy and progressive charity than high wage, indigenous and majority populations (Tait, 2005). Executive Director of the Interfaith Worker Justice Network, Kim Bobo (2004), mentions strong labor–religious alliances with HERE, SEIU, UNITE and LIUNA, all Change to Win affiliates. These coalitions are feasible because they involve workers in non-tradable services that are geographically anchored, making community pressure more effective. Community-based mobilizing is less useful for unions threatened by capital expatriation. To stabilize and grow the beleaguered industrial unions require national policy, such as universal health care or a neo-mercantile trade policy. Until the political alignment at the national level changes, progressive movement on these issues is unlikely. From this perspective one can draw similarities between the decision made by the Change to Win affiliates to leave the AFL–CIO and Gompers’ reluctance to invite semi-skilled industrial workers into the AFL. The negative repercussions of influence decline at the national level have been borne predominately by the industrial unions. In relative terms, unions in nontradable services, such as health care, hotels, retail groceries, and so forth, have ascended due to their political and economic toeholds in smaller geographic regions, and their ability to benefit from sympathetic non-union entities. Departing from the federation enabled Change to Win to reroute resources previously spent by the AFL–CIO in national political campaigns, much of which was intended to slow the descent of the industrial sector into a pre-Wagner malaise. Analogous to Gompers’ exclusion of semi-skilled industrial workers a century ago, Change to Win withdrew support from the most vulnerable segments of the union movement, which presently are unions diminished by international competition and liberalized trade. Sloughing off their formal ties to the sagging industrials, Change to Win gained proportionately more resources for localized campaigns, along with greater freedom to be audacious and creative in the hunt for pockets of unrepresented workers.
Labors’ Divided House
349
Resources were indeed one issue behind the divorce. Just prior to the break, Change to Win demanded a dues rebate of 50% for unions that were organizing. Linked to this demand was a proposal originating from the SEIU calling for the forced merger of the AFL–CIOs 58 affiliates into 20 or so unions based on their dominance in specific product or service markets. Explicit in the plan was that after the restructuring unions would restrict their organizing to their respective markets. This package would have greatly benefited the SEIU, given their position in the expanding health care sector. For industrial unions which dominate in labor markets that have witnessed a dramatic retrenchment over the last decade, such terms were unacceptable. Consistent with the business unionism formula, the SEIU plan drew rigid jurisdictional boundaries that sought to segregate classes of wage-earners based on their propensity to unionize. A second source of tension relates to the progressive shift of the AFL– CIO. James P. Hoffa, President of the Teamsters, is on the advisory board of the pro-war Committee for the Liberation of Iraq. Both Hoffa and Douglas McCarron of the Carpenters have spoken with pride about their relationship with President George W. Bush, and lobbied in favor of drilling for oil in the Alaskan National Wildlife Refuge.38 SEIU leaders tend to be left-of-center on public policy, deserving praise in California for leading the fight for the AFL–CIO reversal on immigrant rights, but their commitment to progressivism is rooted in self-interest rather than ideology. In 2002, SEIU Local 1199 endorsed George Pataki over the decidedly left-leaning Carl McCall in the New York gubernatorial race in exchange for the promise of improved health benefits for members. Functional coalitions, for all their virtues, carry a cost of political interdependency. Splitting from the federation reasserts the independence of leaders such as McCarron and Hoffa by allowing them to entertain political conservatism.39 For more progressive-minded leaders like Stern, breaking from the federation means that the Change to Win unions are not obligated to assimilate the endorsements of AFL–CIO affiliates nor their non-union allies, which in turn affords their leaders more leeway to strike political deals that cater to members’ narrow economic interests. Consistent with business unionism, Change to Win signifies an abandonment of larger socio-economic goals, such as national health care or safeguarding social security, in favor of the flexibility to directly engage in political exchange with state and local elected officials. The final trace of business unionism in the AFL–CIO break is the embrace of a business culture by Change to Win. The SEIU drew inspiration for their organizing strategy from business models (Waldinger et al., 1998). Stern, of the SEIU, is outspoken in his view that unions should partner with business
350
ROLAND ZULLO
to achieve performance goals. McCarron aggressively consolidated thousands of Carpenter locals under large state-wide regional bodies to conform to changes in the building industry.40 The IBT and UFCW have historically favored larger region-wide locals over unit-level locals. Such structures might bring economies of scale for union administration, yet by definition unitlevel democracy is compromised. Often overriding rank-and-file demands to democratically elect their leaders, Change to Win affiliates have imposed trusteeships, appointments, and mergers to implement their growth plans (Erickson, Fisk, Milkman, Mitchell, & Wong, 2004; Wypijewski, 2005). By these counts, the decision by Change to Win to leave the AFL–CIO confirm the resiliency of business unionism within the U.S. context. Germinating in the late 19th century, solidifying as a doctrine by World War I, and reaching a plateau during the Cold War Era, business unionism is a gritty, pragmatic adaptation to a regulatory system that has never embraced industrial democracy. Change to Win affiliates, with comparatively greater leverage in the law, economy, and polity to grow and secure collective contracts than their AFL–CIO counterparts, have staked their camp in this familiar territory. What remains to be seen is response by the jilted federation. Will AFL–CIO affiliates continue to socialize labor-capital conflict, align solidly with progressive causes, devote attention toward national policy that benefits all wage-earners, and forge political bonds with non-union populations and sympathetic organizations? If so, will such a strategy spawn new types of radical associational representation? It is tempting to imagine the U.S. labor movement entering a cycle where the failure of business unionism to cover large segments of wage earners will foment radicalism. Associational and democratic labor organizations, such as the Knights of Labor (1869–1900), Industrial Workers of the World (1905–present) and Trade Union Educational League (1920–1928) make for colorful lore. But organizations demanding class-based systemic reform have been notoriously short-lived as wage-earners, when given a choice, gravitate toward a model that conforms to the opportunities presented by the regulatory system (Kimeldorf, 1999). Associational forms of representation in the U.S. all withered prior to surges in unionization of the Gompers’ brand. When the political-economy favored business unionism, roughly 1940 to 1970, one-third of the private sector workforce was represented and the AFL–CIO had no rivals. Steady increases in compensation and job security for much of the working class enabled union leaders to fend off shop-floor and community radicalism. So even if the U.S. is nearing the point where the absence of coverage under business unionism triggers spontaneous upsurges, a temporary adjustment in collective bargaining rights is
Labors’ Divided House
351
more likely than a repeal of the entrepreneurial property rights doctrine. If history is a useful guide, political strife will be a precondition for a revised statutory framework, and any amendments will accommodate the most disruptive classes of wage-earners. Once the law is altered these unionists will choose business unionism. The U.S. political-economy has shown little tolerance for alternatives.
NOTES 1. Perhaps the strongest linkage between political democracy and capitalism is war. War has served as a means to expand American-styled capitalism through the acquisition of necessary inputs, such as land, labor and raw materials, and by establishing political conditions for the creation of product markets. Prosecuting a war, however, requires soldiers and cooperation among industrial workers employed in the defense industry. As Keyssar (2000) outlines, suffrage rights in the U.S. were incrementally granted to various groups (e.g. non-property owners, blacks) in exchange for their participation in major U.S. wars. This legacy is reflected in the contemporary policy of offering citizenship to foreigners willing to enlist as U.S. soldiers in the Iraq conflict. 2. Perlman (1922, Chapter 1) for historic background. 3. For example: People vs. Melvin, Court of the General Sessions, N.Y. City; 2 Wheeler C.C. (N.Y.) 262 [1810] (a critical mention of the relevancy of English common law to the U.S.). In People vs. Melvin the cordwainers technically engaged in a conspiracy, yet the guilty verdict brought a fine of one dollar. 4. President of the newly founded AFL, Samuel Gompers, railed against court interference in labor affairs, and often directed his organizers to defy injunctions (Gompers, 1967, Chapter 33; Gompers, 1919, Chapter 3). 5. The 14th amendment, ratified in 1868, provided equal protection to all ‘‘persons born or naturalized in the United States,’’ and was intended to ensure that newly freed slaves enjoyed the same liberties as other U.S. citizens. The major cases granting corporations personhood based on the 14th amendment are: Santa Clara County vs. Southern Pacific Railroad, 118 U.S. 394 (1886); Pembina Consolidated Silver Mining and Milling Co. vs. Pennsylvania, 125 U.S. 181(1888); Minneapolis & St. Louis Railroad vs. Beckwith, 129, U.S. 26 (1889). Constitutional privileges for corporations expanded soon afterward to include protection against 4th amendment unreasonable search and seizure, see Hale vs. Henkel, 201 U.S. 43 (1906) and 5th amendment takings, see Pennsylvania Coal Co. vs. Mahon, 260 U.S. 393 (1922). 6. See Pope (1987) for similar arguments. See Dorchy vs. Kansas 272 U.S. 306 (1926) (striking mineworkers sought protection in the 14th amendment but were denied); Hague vs. Committee for Industrial Organization, 307 U.S. 496 (1939) (union petition for free speech and assembly protection denied because unions were not ‘‘citizens’’ of the United States as per 14th amendment). Note, however, the courts did flirt with the idea of 1st amendment protection for picketers in Thornhill vs. Alabama, 310 U.S. 88 (1940), but retreated beginning with International Brotherhood of Teamsters Union Local 309 vs. Hanke 339, U.S. 470 (1950).
352
ROLAND ZULLO
7. For a listing, see Forbath (1991) Appendix A. See Lochner vs. New York 198 U.S. 45 (1905) (a precedent for invalidating legislation that regulates the work relationship). 8. Boycotts were widely adopted by labor activists in the post-civil war era, for analysis see Schaffner (1910); Laidler (1913, Chapter 4); Wolman (1916). When in 1894 the ARU boycotted Pullman Palace Car Company products in sympathy with striking Pullman workers, the Supreme Court reasserted their commitment to interstate commerce, broke the boycott, and upheld the arrest and conviction of ARU officers, see In Re Debs 158 U.S. 564 (1895). Prohibitions were extended to consumer’s boycotts in the ‘‘Danbury Hatters’’ case Loewe vs. Lawler 208 U.S. 274 (1908). 9. Disregarding congressional intent, the courts applied the Sherman Anti-Trust Act to enjoin labor boycotts for the sake of protecting the ‘‘liberty of a trader to engage in business,’’ and thereafter exposing unions to treble damages for their actions. See Loewe vs. Lawler, 208 U.S. 274. (1908). 10. Labors’ few successes in the area of protective law were narrow regulations designed to protect women and children. For women, however, this targeted reform cut in two directions. One the one hand, the protective laws served as a toehold for legislation granting universal coverage. On the other, such laws reinforced the subordinate role of women in society and provided a justification by employers to discriminate on the basis of sex. See Kessler-Harris (2003, Chapter 7). 11. Adair vs. United States, 208 U.S. 161 (1908). 12. Coolidge was no progressive on labor matters. In 1919 he gained national recognition when as Governor of Massachusetts he suppressed a police officers strike in Boston. 13. Texas & New Orleans Railroad Company vs. Brotherhood of Railway and Steamship Clerks, 281 U.S. 548 (1930). The critical issue was recognizing a federal role in regulating labor–management relations to facilitate commerce. 14. Indeed, Wagner never established a permanent role for unions descending from the constitutional right to associate or, as labor at the time argued, the thirteenth amendment abolishing slavery (Forbath, 1991, pp. 135–141). Preserving commerce was the overriding aim. Collective bargaining was perceived as a means toward alleviating industrial conflict, which in turn would improve the function of the economic system. Once the economy recovered, the utility of unions and bilateral negotiations diminished. See Pope (2004) for similar arguments. 15. A.L.A. Schechter Poultry Corp. vs. United States, 295 U.S. 495 (1935). 16. Many employers were advised to ignore Wagner based on the court treatment of the National Industrial recovery Act. National Labor Relations Board vs. Jones & Laughlin Steel Corp. 301 U.S. 1 (1937) (upheld the constitutionality of Wagner). 17. NLRB vs. Mackay Radio & Telegraph 304 U.S. 333 (1938). 18. See NLRB vs. Fansteel Metallurgical Corp. 306 U.S. 240 (1939) (sit down strikers responding to illegal employer conduct could be terminated); International Union, UAW, AFL vs. Wisconsin Employment Relations Board 336 U.S. 245 (intermittent strikes to protest contract impasse not protected); NLRB vs. Kohler 220 F. 2d 3 (1955) (sporadic walkouts over work environment not protected). However, unions can engage in work-to-rule tactics that achieve a similar effect as a slowdown. 19. NLRA Section 8 (a) (1). 20. NLRB vs. Virginia Electric & Power Co. 314 U.S. 469 (1941).
Labors’ Divided House
353
21. Examples: The Newlands Act of 1913 offered dispute resolution mechanisms for railroad engineers; the Transportation Act of 1920 and later the Railway Labor Act of 1926 broadened coverage to include all classes of railroad employees; in a transparent nod to Southern Democrats Wagner excluded farm workers and domestic servants; Taft Hartley excluded supervisors; and although federal and municipal workers began to gain recognition rights in the 1960s, presently roughly 40 percent of public employees are denied basic collective bargaining rights (ILO, 2004). 22. Taft Hartley, for instance, required plant guards to be in a separate union from other workers. Strike exceptions for health care workers were introduced into the NLRA in 1974. Public sector law commonly carves exceptional rules for classes of employees, such as teachers or protective service personnel. 23. The phrase ‘‘job consciousness’’ is from Perlman (1928). While I share Perlman’s conclusion that the ideology of property rights shaped the U.S. union movement, I hesitate to subscribe to his notion that a social psychology of ‘‘scarcity consciousness’’ explains labors’ conservatism. Instead, as described below, I model labor unions as dynamic political institutions that work to resolve contending philosophical viewpoints. 24. I want to avoid historical determinism. Labors’ conservative turn under Gompers is partly attributable to his leadership, yet the decisions by any labor leader are constrained by the regulatory system and by the demands from organization members. 25. By ‘‘commercial class’’ I refer to the myriad of enterprises, private and nonprofit, large and small, that thrive on commercialism. I stress the term ‘‘largely’’ in deference to the democratic institutions across the polity, federally but that are especially responsive locally, that prevents a perfect alignment between politics and commercial interests. 26. The Davis–Bacon Act (1931), for example, requires contractors to compensate based on prevailing wages (local standards) for contracts using federal funding. This law severely restricts contractors that wish to undermine the local union rate for construction and renovation. 27. Gompers’ vociferous opposition to Chinese immigration is well documented, however immigration from North American nations were also opposed, see Levenstein (1968). 28. Michels’ (1962, pp. 287–292) observation that the ‘‘aristocratic tendencies’’ of U.S. union leaders mirror societal tolerance for capitalism has some traction. Unions, according to Michels, offer working class leaders the chance to share in the spoils of an economic system that apportions the lion’s share of wealth to the powerful. Hence, unions mimic business practices, securing high salaries and perks for officers, centralizing their power and control over local and regional bodies, and in some instances acting as human relations agents in the workplace. 29. In his speeches, Gompers frequently conflates capitalism with democracy, a practice that continues in U.S. politics. See Gompers (1967, pp. 175–210); Bedford (1995). 30. See Gompers (1919, Chapter 7). Gompers’ access to President Wilson during World War I appears to have protected the AFL during that period, see McCartin (1997, pp. 80–93). George Meany staunchly supported the Vietnam War, and Lane Kirkland, his successor, followed in this conservative tradition. Phillip Murray, CIO President from 1940–1952, attempted to avoid wartime conflict through tripartite negotiations between Truman, the CIO leaders and leaders of industry, but industry declined, see Buhle (1999, Chapter 3).
354
ROLAND ZULLO
31. With Gompers’ approval, the Espionage Act of 1917 was used to target the Industrial Workers of the World. The CIO decided in 1948 to bar Communists from holding office, and in 1949–1950 it expelled 11 of its affiliated unions, which were said to be Communist-dominated. Even seemingly progressive labor leaders jumped on the anti-communist bandwagon (Keeran, 1980; Reuther, 1948). See also Meyer (1992); Rosswurm (1992); Stepan-Norris and Zeitlin (2003). 32. In 2006, the AFL–CIO is attempting to establish ballot initiatives to raise the minimum wage in several U.S. states. Another example is the living wage initiative. 33. In August 2003 auto parts supplier Dana Corporation agreed to neutrality and card check after major auto producers communicated to Dana their preference for dealing with ‘‘good corporate citizens,’’ a communique´ urged by the UAW. So far the UAW action survived the secondary boycott challenge sponsored by the Right to Work Foundation because the agreement was not a condition for doing business with the auto producers. See NLRB Office of the General Council, Case 7-CC-1786. See Greenhouse (2005) for the SEIU example, and Bronfenbrenner, Friedman, Hurd, Oswald, and Seeber (1998). 34. Any effective mobilization effort, whether political or economic, requires a network of volunteer activists that are willing and capable of performing the mechanics of member mobilization. The dilemma for labor is that local activists typically distrust or even loath leaders at the peak who call upon their energy and talent with the demeanor of military commanders. Union leaders that want mobilization to happen must involve activists in the formation of the policy that necessitates the mobilization. 35. For various perspectives, see Honey (2000); Nelson (2001); Grob (1960); Mandel (1955). 36. Canada, with an economic system that is integrated with the U.S. and is subject to the same trade treaties, has managed to sustain union density levels that are near the post-WWII U.S. high. In a number of critical areas, Canadian provinces have more favorable law for unions, including card-check recognition process that the AFL–CIO is presently pursuing in the Employee Free Choice Act. 37. The earliest mention of the phrase social movement unionism that I could locate was attributed to Walter Reuther, President of the UAW, in a 1962 interview (Bok & Dunlop, 1970, p. 362). In reflecting on the early years of the UAW, Reuther mused that in ‘‘the next phase y the labor movement will become less of an economic movement and more of a social movement y concerned with the economic factors, of course, but also with the moral, the spiritual, the intellectual and the social nature of our society.’’ 38. This soured the possibility of a lasting ‘‘Teamsters and Turtles’’ coalition arising from the 1999 WTO protests in Seattle (Rose, 2004). 39. A precedent that underscores the ideological shift of the AFL–CIO, Walter Reuther opposed AFL–CIO President George Meaney’s reactionary stance on civil rights and social welfare programs and led the United Automobile Workers to withdraw from the federation in 1968. 40. Douglas J. McCarron, President, United Brotherhood of Carpenters & Joiners of America. Testimony before the Subcommittee on Employer–Employee Relations, U.S. House of Representatives, June 25, 1998.
Labors’ Divided House
355
REFERENCES Albert, P. J. & Palladino, G. (Eds). (2003). The Samuel Gompers Papers. The American Federation of Labor at the Height of Progressivism, 1913–1917 (Vol. 9). Urbana, IL: University of Illinois Press. Becker, C. (1994). Better than a strike: Protecting new forms of collective work stoppages under the National Labor Relations Act. University of Chicago Law Review, 61(2), 351–421. Bedford, J. (1995). Samuel Gompers and the Caribbean: The AFL, Cuba, and Puerto Rico 1898–1906. Labors’ Heritage, 6(4), 4–25. Bobo, K. (2004). Religion-labor partnerships: Alive and growing in the new millennium. In: D. Reynolds (Ed.), Partnering for change: Unions and community groups build coalitions for economic justice. Armonk, NY: M.E. Sharpe. Bok, D. C., & Dunlop, J. T. (1970). Labor and the American community. New York, NY: Simon and Schuster. Brecher, J., & Costello, T. (1990). Building bridges: The emerging grassroots coalition of labor and community. New York, NY: Monthly Review Press. Brisbin, R. A. (2002). A strike like no other strike: Law and resistance during the Pittston Cola Strike of 1989–1990. Baltimore, MD: John Hopkins University Press. Bronfenbrenner, K., Friedman, S., Hurd, R. W., Oswald, R. A., & Seeber, R. L. (1998). Organizing to win: New research on union strategies. Ithaca, NY: Cornell University Press. Buhle, P. (1999). Taking care of business: Samuel Gompers, George Meany, Lane Kirkland, and the tragedy of American labor. New York, NY: Monthly Review Press. Cheyney, E. P. (1889). Decisions of the courts in conspiracy and boycott cases. Political Science Quarterly, 4(2), 261–278. Cook, A. H. (1963). Union democracy: Practice and ideal. Ithaca, NY: Cornell University. Emspak, F. (1984). The breakup of the CIO. In: M. Zeitlin & H. Kimeldorf (Eds), Political power and social theory (Vol. 4). Greenwich, CT: JAI Press Inc. Erickson, C. L., Fisk, C., Milkman, R., Mitchell, DJ. B., & Wong, K. (2004). Justice for janitors in Los Angeles and beyond: A new form of unionism in the twenty-first century? In: P. V. Wunnava (Ed.), The changing role of unions: New forms of representation. Armonk, NY: M.E. Sharpe. Fantasia, R., & Voss, K. (2004). Hard work: Remaking the American labor movement. Berkeley, CA: University of California Press. Filippelli, R. L. (1984). UE: An uncertain legacy. In: M. Zeitlin & H. Kimeldorf (Eds), Political power and social theory (Vol. 4). Greenwich, CT: JAI Press Inc. Fine, J. (2006). Worker centers: Organizing communities at the edge of a dream. Ithaca, NY: Cornell University Press. Forbath, W. E. (1991). Law and the shaping of the American labor movement. Cambridge, MA: Harvard University Press. Golden, C. S., & Ruttenberg, H. J. (1942). The dynamics of industrial democracy. New York, NY: Harper & Brothers Publishers. Gompers, S. (1919). Labor and the common welfare. New York, NY: E.P. Dutton & Company. Gompers, S. (1967). Seventy years of life and labor: An autobiography. New York, NY: Augustus M. Kelley Publishers. Greene, J. (1998). Pure and simple politics: The American federation of labor and political activism, 1881–1917. New York, NY: Cambridge University Press.
356
ROLAND ZULLO
Greenhouse, S. (2005). Janitors’ Drive in Texas Gives Hope to Unions. New York Times, November 28. Grob, G. N. (1958). The knights of labor and the trade unions, 1878–1886. Journal of Economic History, 18(2), 176–192. Grob, G. N. (1960). Organized labor and the Negro worker, 1865–1900. Labor History, 1(2), 164–176. Hattam, V. C. (1993). Labor visions and state power: The origins of business unionism in the United States. Princeton, NJ: Princeton University Press. Honey, M. (2000). Black workers remember: An oral history of segregation, unionism and the freedom struggle. Berkeley, CA: University of California Press. Hurd, R. (2004). The failure of organizing, the new unity partnership, and the future of the labor movement. Working USA: The Journal of Labor and Society, 8 (1) September: 5–25. Hurvitz, H. (1986). American labor law and the doctrine of entrepreneurial property rights: Boycotts, courts, and the judicial reorientation of 1886–1895. Industrial Relations Law Journal, 8(3), 307–361. ILO. (2004). Annual survey of violations of trade union rights. Geneva: International Labor Organization. Keeran, R. (1980). The communist party and the auto workers unions. Bloomington, IN: Indiana University Press. Kessler-Harris, A. (2003). Out to work: A history of wage earning women in the united states. Oxford, UK: Oxford University Press. Keyssar, A. (2000). The right to vote: A contested history of democracy in the United States. New York, NY: Basic Books. Kimeldorf, H. (1999). Battling for American labor: Wobblies, craft workers and the making of the union movement. Berkeley, CA: University of California Press. Klare, K. E. (1978). Judicial deradicalization of the Wagner Act and the origins of modern legal consciousness, 1937–1941. Minnesota Law Review, 62(3), 265–340. Kutler, S. I. (1962). Labor, the Clayton Act, and the Supreme Court. Labor History, 3(1), 19–38. Laidler, H. W. (1913). Boycotts and the labor struggle: Economic and legal aspects. New York: John Lane Company. Laslett, J. H. M. (1986). Samuel Gompers and the rise of business unionism. In: M. Dubofsky & W. V. Tine (Eds), Labor Leaders in America. Urbana, IL: University of Illinois Press. Levenstein, H. A. (1968). The AFL and Mexican immigration in the 1920s: An experiment in labor diplomacy. Hispanic American Historical Review, 48(2), 206–219. Lipset, S. M. (1954). The political process in trade unions: A theoretical statement. In: M. Berger, T. Abel & C. H. Page (Eds), Freedom and control in modern society. Toronto, ON: D. Van Nostrand Company, Inc. Mandel, B. (1954). Gompers and business unionism, 1873–1890. Business History Review, 28(3), 264–275. Mandel, B. (1955). Samuel Gompers and the Negro workers, 1886–1914. Journal of Negro History, 40(1), 34–60. Martin, A. (2006). Why does the new labor movement look so much like the old one? Putting the 1990s revitalization project in historical context. Journal of Labor Research, 27(2), 163–185. McCartin, J. A. (1997). The struggle for industrial democracy and the origins of modern American labor relations, 1912–1921. Chapel Hill, NC: University of North Carolina Press.
Labors’ Divided House
357
Meyer, S. (1992). Stalin over Wisconsin: The making and unmaking of militant unionism, 1900–1950. New Brunswick, NJ: Rutgers University Press. Michels, R. (1962). Political parties: A sociological study of the oligarchical tendencies of modern democracy. New York, NY: Collier Books. Morris, R. B. (1946). Government and labor in early America. New York, NY: Harper & Row Publishers. Nelles, W. (1931). The first American labor case. Yale Law Journal, 41(2), 165–200. Nelson, B. (2001). Divided we stand: American workers and the struggle for black equality. Princeton, NJ: Princeton University Press. Nyland, C. (1998). Taylorism and the mutual-gains strategy. Industrial Relations, 37(4), 519–542 Oestreicher, R. (1986). Terrance Powderly, the knights of labor, and artisanal republicanism. In: M. Dubofsky & W. V. Tine (Eds), Labor leaders in America. Urbana, IL: University of Illinois Press. Perlman, S. (1922). A history of trade unionism in the united states. New York, NY: Macmillan. Perlman, S. (1928). A theory of the labor movement. New York, NY: Macmillan. Perry, C. R. (1987). Union corporate campaigns. Philadelphia, PA: University of Pennsylvania. Phelan, C. (1989). William Green: A biography of a labor leader. Albany, NY: State University of New York Press. Pope, J. G. (1987). Labor and the constitution: From abolition to deindustrialization. Texas Law Review, 65(6), 1071–1136. Pope, J. G. (2004). How American workers lost the right to strike, and other tales. Michigan Law Review, 103(3), 518–553. Reuther, W. P. (1948). How to beat the communists. In: H. M. Christman (Ed.), Walter P. Reuther: Selected papers (pp. 19–30). New York, NY: Pyramid Books. Reynolds, D. B. (2004). Partnering for change: Unions and community groups build coalitions for economic justice. Armonk, NY: M.E. Sharpe. Robinson, I. (2002). Does neoliberal restructuring promote social movement unionism? U.S. developments in comparative perspective. In: N. Bruce (Ed.), Unions in a Globalized Environment: Changing Borders, Organizational Boundaries, and Social Roles. Armonk, NY: M.E. Sharpe. Rose, F. (2004). Labor-environmental coalitions. In: D. Reynolds (Ed.), Partnering for change: Unions and community groups build coalitions for economic justice. Armonk, NY: M.E. Sharpe. Rosswurm, S. (1992). The CIO’s left-led unions. New Brunswick, NJ: Rutgers University Press. Schaffner, M. A. (1910). Effect of the recent boycott decisions. Annals of the American Academy of Political Science, 36(2). The Settlement of Labor Disputes: 23–33. Schattschneider, E. E. (1960). The semisovereign people: A realist’s view of democracy in America. New York: Holt Rinehart and Winston. Steinfeld, R. J. (2001). Coercion, contract, and free labor in the nineteenth century. Cambridge, England: Cambridge University Press. Stepan-Norris, J., & Zeitlin, M. (2003). Left out: Reds and America’s industrial unions. Cambridge, England: Cambridge University Press. Stone, K. W. V. (1992). The legacy of industrial pluralism: The tension between individual employment rights and the new deal collective bargaining system. University of Chicago Law Review, 59(2), 575–644. Tait, V. (2005). Poor workers’ unions: Rebuilding labor from below. Cambridge, MA: South End Press.
358
ROLAND ZULLO
Tillman, R. M., & Cummings, M. S. (Eds) (1999). The transformation of U.S. unions: Voices, visions and strategies from the grassroots. Boulder, CO: Lynne Rienner Publishers, Inc. Tomlins, C. (2004). Early British America, 1585–1830. In: D. Hay & P. Craven (Eds), Masters, Servants and Magistrates in Britain and the Empire (pp. 1562–1955). Chapel Hill, NC: University of North Carolina Press. Waldinger, R., Erickson, C., Milkman, R., Mitchell, D. J. B., Valenzuela, A., Wong, K., & Zeitlin, M. (1998). Helots no more: A case study of the justice for janitors campaign in Los Angeles. In: K. Bronfenbrenner, S. Friedman, R. W. Hurd, RA. Oswald & R. L. Seeber (Eds), Organizing to Win: New Research on Union Strategies. Ithaca, NY: Cornell University Press. Walling, W. E. (1926). American labor and American democracy. New York, NY: Harper & Brothers Publishers. Wolman, L. (1916). The boycott in American trade unions. Baltimore, MD: Johns Hopkins Press. Wypijewski, J. A. (2005). Is this really an ‘insurgency’ to shake up the labor movement? Counterpunch Newsletter, July 22. Online at: http://www.counterpunch.org/ jw07222005.html Zeiger, R. H. (1995). The CIO: 1935–1955. Chapel Hill, NC: University of North Carolina Press.
CHANGE TO WIN: CAN STRUCTURAL REFORM REVITALIZE THE AMERICAN LABOR MOVEMENT? Jack Fiorito, Paul Jarley and John T. Delaney ABSTRACT The U.S. labor movement is in decline and a crisis of national leadership has emerged over conflicting prescriptions for labor’s revival. Union leaders have seemingly established consensus on the need for change, but disagree about the nature of needed reform, and methods for accomplishing meaningful changes that might address the long-term crisis. This paper strives to inform and advance debates on these issues. Two national union surveys conducted in 1990 and 1997 provide the primary evidentiary base. Given their critical role in this study, measures from the surveys and certain aspects of the surveys are scrutinized. These surveys span the ‘‘Sweeney Insurgency’’ and the early years of the Sweeney AFL-CIO administration. Although both surveys have supported previous cross-section based studies, no published work has expressly focused on the change and stability within national unions or the longitudinal potential these data collectively provide. Using this potential to reexamine relations between union structures, strategies, and performance, this paper seeks to establish an evidentiary base to inform the current debate about union
Advances in Industrial and Labor Relations, Volume 15, 359–391 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0742-6186/doi:10.1016/S0742-6186(06)15009-X
359
360
JACK FIORITO ET AL.
reforms and their likely consequences. In addition, suggestions for future research on unions and approaches to studying unions are offered.
American national unions and the AFL-CIO have been in a deepening crisis for over four decades and have recognized this ‘‘fact’’ for at least two decades. In 1995, matters boiled over into the first successful electoral insurgency in the AFL-CIO’s then 40-year history. For a time at least, most national union leaders expressed hope that John Sweeney had the answer to labor’s plight. But as the subsequent ten years unfolded unions did not rebound and continued union decline stimulated discontent in the ranks. In response, the AFL-CIO Executive Council’s spring 2005 meeting considered fairly dramatic restructuring proposals (AFL-CIO Executive Council, 2005). Although incumbent President Sweeney’s allies prevailed on key votes, there was clearly a sense of urgency in the minds of some key national union leaders that was not assuaged by the Executive Council’s actions at its spring Las Vegas meeting (Cleeland, 2005; Greenhouse, 2005). This erupted in the insurgency and secession of several major affiliates just prior to and just after the AFL-CIO convention in Chicago in the summer of 2005, resulting in the rival Change to Win Federation’s (CWF) formation. This may rank among the most critical and dramatic ‘‘labor’’ developments in many years. Somewhere John L. Lewis might be thinking: ‘‘It’s about time!’’ What went wrong? Clearly a number of factors worked against organized labor in recent decades including increasing globalization, structural change in labor markets, government indifference or hostility to unions, and declining worker instrumentality perceptions to name just a few. But it is fair to say that the Sweeney administration swept to power on a platform based on structural and financial changes within the Federation to boost organizing, more prominence for unions in politics, and numerous staff changes at the AFL-CIO. At the broadest level, the Sweeney-led Federation followed a strategy intended to spur union membership through internal union reform and political activity designed to restore workers’ rights to organize. It is only a slight oversimplification to say that the Sweeney Administration’s union reform strategy stressed cultural change. It used the bully pulpit and some financial incentives to encourage unions to switch from a servicing to organizing orientation that places more responsibility into the hands of rank-and-file members, freeing both human and financial resources for increased organizing. Under the banners of ‘‘organizing to change’’ and ‘‘changing to organize’’ it encouraged affiliates to devote 30% of their budgets to organizing and touted a set of ‘‘rank-and-file’’ intensive organizing
Can Structural Reform Revitalize the American Labor Movement?
361
methods that appear to be very effective in creating more union workplaces (Bronfenbrenner, 1997, 2001). The problem has been in ‘‘changing to organize.’’ Systematic evidence is elusive, but it appears that while national unions generally responded positively to the renewed emphasis on organizing (Fiorito, Jarley, Delaney, & Kolodinsky, 2000), the effort fell far short of desired levels (Fiorito & Jarley, 2003; Hurd, 2004). Union political actions have also fallen short, with no significant improvement in national labor policy, continued Republican control of Congress, and a Bush Administration elected and reelected. Failure breeds dissatisfaction and the recent insurgency argued for a significant change in direction, one that places politics second to building workplace presence. Labor’s reformers stress a workplace first strategy that would consolidate the labor movement into a small number of largely industrial based unions better able to coordinate bargaining and make serious investments in organizing workers (see Restoring the American Dream: Building a 21st Century Labor Movement than Can Win; and Unite to Win: A 21st Century Plan to Build a New Strength for Working People, both on the SEIU website at www.seiu.org). They argue that American unions collectively have enough resources to achieve a turnaround, but that they are too fragmented to take advantage of their strength in core industries and that many unions are too small to employ the strategies necessary to succeed. To wit, the CWF unions argued for a much stronger AFL-CIO, one that would use strong incentives to force union restructuring, create new unions in growth sectors of the economy, and direct national campaigns against large multi-national employers. The insurgents argued that the labor movement must put structure first. Unless unions consolidate and rationalize to build a greater presence in American workplaces they will never generate the necessary political clout and innovation, or leverage enough resources to reinvigorate the movement. This approach is similar to the one pressed by Lewis 70 years earlier, though his insurgency occurred in a very different context. At that time, many unskilled workers were ripe for unionization and the CIO responded with a structure designed to organize them. The CIO’s centralized decision making and top-down organizing style created several new industrial unions and adopted many organizing innovations, such as the widespread use of sit-down strikes. Its ‘‘flying squadrons’’ provided the organizational structure and strategy needed to mobilize workers, win recognition, and gain contracts. Today, the reformers are arguing that a ‘‘back to the future’’ approach will produce the same benefits. While acknowledging that many factors play a role in determining labor’s fate, we too believe that union resurgence requires bold action targeted at
362
JACK FIORITO ET AL.
internal structural reform, and have spent the better part of two decades exploring the causes of national union structure and its consequences for union effectiveness. To understand better how unions function, change and perform, we conducted surveys of national unions in both 1990 and 1997. In this paper, we provide an overview of those surveys, their methods and measures, and present comparative and dynamic results for models of union innovation and performance. These findings help clarify the issues involved in ‘‘changing to organize,’’ and in general support some of the key points made by the insurgents in the recent debate on the future of the AFL-CIO, the CWF, and their affiliated unions.
TWO SURVEYS OF AMERICAN NATIONAL UNIONS Broadly speaking, the two surveys conducted in 1990 and 1997 share a common focus on national unions and their organizational characteristics, including innovation. The 1990 National Union Survey (NUS) gathered detailed information on leaders’ perceptions of their organizations in a broad spectrum of national unions. Topics included goals, strategy, ideology, environmental scanning, bureaucracy (both in terms of structuring of activities, or rationalization, and in terms of the locus of authority and decision-making, or centralization), innovation, and effectiveness. To the extent possible, the survey borrowed or adapted measurement instruments that had been developed mainly to study employing organizations (see Price & Mueller, 1986). In a way, the NUS was a ‘‘perceptual census of unions’’ as seen by their top officers and staff, and in many regards, the first such undertaking within the U.S. The U.S. Bureau of Labor Statistics (BLS) for many years collected superficial data biennially, including names and titles of officials, addresses and telephone numbers, and numbers of members and locals, with irregular supplements focusing in more depth on a particular issue such as the industrial distribution of unions’ members. These data were published in its Directory series (e.g., U.S. Bureau of Labor Statistics, 1980). In the early Reagan years the BLS survey was a casualty of downsizing, but the idea was quickly picked up by the Bureau of National Affairs, continued, and recently annualized (e.g., Gifford, 1997). There have been other notable forbearers to the NUS, including a survey by Knoke in the 1980s (Knoke, 1990), and more piecemeal surveys on particular topics, such as strategic planning (e.g., Stratton & Brown, 1989). There also have been efforts to examine a broad range of union characteristics via archival sources (e.g., Fiorito & Hendricks,
Can Structural Reform Revitalize the American Labor Movement?
363
1987; Masters, 1997). But the NUS was unique in its combination of topical and sampling breadth in a single survey. In addition to the NUS instrument, which translated into a lengthy (25–30 min) computer-assisted telephone interview even after some painful item – and topic-trimming, the survey included a request that interviewees provide a copy of their union’s constitution (most did), which was subsequently content coded. The content coding focused on representative structures and some of the same issues addressed in the interview, such as centralization. At the same time, data from government and other sources on union finances, organizing activity, and union environments were developed. The result is a rich data base that supported research on the determinants of innovation (Delaney, Jarley, & Fiorito, 1996), the role of union structures, strategies and innovation in organizing effectiveness (Fiorito, Jarley, & Delaney, 1995), the determinants of union administrative and governance structures (Jarley, Fiorito, & Delaney, 1997), and similar topics. An initial report (Delaney, Fiorito, & Jarley, 1991) on the survey method and results provides details on the NUS. The Survey of Union Information Technology (SUIT) was largely a follow-up on the NUS. In a nutshell, it compressed some of the key issues addressed in the NUS into a small fraction of the SUIT, and devoted considerable space to emerging issues of information technology (IT) use in unions. The focus was based on our belief that innovation is crucial for union success, and within the various areas where unions can innovate, IT stands out as one of rapid change and central importance. More details on the SUIT survey method and descriptive results are provided in Fiorito, Jarley, and Delaney (1998). Fig. 1 provides a snapshot comparison of the NUS and SUIT. As shown, the SUIT used a mail survey in contrast to the telephone interviews of the NUS. For both surveys, published data were a useful starting point for constructing lists of ‘‘target’’ respondents, i.e., the sample frame. In both instances, the response rate in terms of unions was quite good, although the ‘‘true’’ response rates in terms of individuals contacted were more modest. External validity is a critical issue in any survey but especially so in surveys of perceptions where self-serving bias is an issue, and where one wishes to draw inferences about realities. When NUS data have been compared to other sources, support for the NUS’ validity has been found. For example, in Fiorito et al. (1995), NUS respondents’ perceived organizing effectiveness correlated strongly with NLRB win rates (r ¼ 0.71) and moderately with membership growth (r ¼ 0.31). Comparisons of NUS innovation rankings
364
JACK FIORITO ET AL. Issue Survey Name Topics emphasized Year conducted Survey methods
NUS National Union Survey Union strategy, structure, innovation, performance 1990 (Summer) Telephone, mail pre-contact
Length "Target List" Target types Other notables
Approx. 25-minute interview Gifford, AFL-CIO staff Top officers and staff Donahue endorsement Constitutions content coded
Sample frame N Unions responding N Response bias Pct. of members represented Avg. union membership
150 111 (74%) Large, AFL-CIO affiliates 95% 192,456
120 75 (63%) "Wired," structured 75% 166,939
Individuals contacted Usable responses Respondent types
Approx. 670 275 (41%) 66% staff, 34% officers
Approx. 600 93 (16%) 65% staff, 35% officers
SUIT Survey of Union IT Union information technology use 1997 (Summer-Fall) Mail, telephone follow-up Seven pages (8.5"x11") Gifford, web sites Top officers and staff No endorsement
Summary of Responses and Nonresponses by Survey
SUIT Response
No Yes Total
Fig. 1.
NUS Response No Yes Few 53 17 58 111
Total 75
Comparison of NUS and SUIT Surveys.
with anecdotal evidence provided similar support for the NUS. Assessments of external validity for the SUIT also have been encouraging (Fiorito, Jarley, & Delaney, 2000; Fiorito, Jarley, & Delaney, 2002). More on these topics will be said below. Measures At the outset, it should be noted that this paper places a premium on comparability between the two surveys. In some instances, that has required using subscales of the full NUS scales with an associated penalty in terms of reliability. In other instances, items from different but related scales in the NUS were combined to form new scales that can be examined in both the NUS and SUIT. Similarly, the SUIT provides the potential for a much more detailed assessment of IT issues than undertaken below (see e.g., Fiorito
Can Structural Reform Revitalize the American Labor Movement?
365
Organizational Effectiveness
Three-item scale summing respondent ratings of effectiveness in organizing, bargaining, and overall.
Organizing Effectiveness
Single item for organizing effectiveness.
Innovation
Two-item scale for innovation.
Rationalization
Five-item scale for rationalization or structuring of activities.
Decentralization
Three-item scale for the extent to which decisions are made at lower organizational levels (e.g., members or locals rather than national headquarters).
Strategic Scope
Three-item scale to assess the breadth of issues the union addresses and the breadth of methods the union uses to accomplish its goals.
Size
Number of union members represented.
Membership Heterogeneity
Index where maximum heterogeneity occurs where the union's membership is 50% women and 50% minorities.
IT Use
Composite of two multi-item scales for IT use. Subscales represent, alternatively, the number of types of IT used, or "IT Forms," and the number of IT applications.
Organizing Commitment
A single item: "Compared to other unions, this union is very committed to organizing."
Membership Growth
Relative change in membership based on 1990 and 1997 survey responses.
Fig. 2.
Measures: Summary Definitions. Note: See text for details on measures.
et al., 2000, 2002), but since the NUS did not address IT in any depth, our present emphasis on comparability dictates only a cursory treatment of IT issues here. Fig. 2 provides a brief description of measures used in this chapter. Union Effectiveness. Three Likert-style items asked respondents to rate their union’s effectiveness in ‘‘organizing new members,’’ ‘‘serving its current members,’’ and ‘‘achieving its goals (overall).’’ The first item on organizing is used as a single item to represent organizing effectiveness, and all three items are combined to form a composite scale for overall union effectiveness. In the NUS, these three items were part of a larger six-item scale addressing additional areas of union performance. Innovation. Two items are combined to form a scale for innovation. The first item stated ‘‘This union is very advanced in the application of new methods
366
JACK FIORITO ET AL.
to organize and serve its members.’’ The second stated ‘‘Polls or surveys are used to help in organizing and serving current members.’’ These wordings are those used in the SUIT, and differ slightly from those used in the NUS. The two items were selected for the SUIT based on face validity and their correlations with longer scales for innovation and innovative tactics (respectively) in the NUS. Rationalization. Rationalization (or ‘‘structuring of activities’’) is measured with a five-item scale. Each item is a four-point Likert-style scale representing a sub-dimension of the larger rationalization concept. The subdimensions represented are standardization, coordination, formalization, communication, and environmental scanning. These items were summed to yield the composite rationalization scale, with higher values indicating a higher level of rationalization. The items are very similar in wording across the two surveys. In the NUS, however, environmental scanning was conceptualized as a distinct construct and other items to tap it were included. Owing to construct overlap and length limitations in the SUIT, we combined the single item with the others to form the present rationalization measure. Decentralization. In both the NUS and SUIT versions, three items are combined to measure the extent to which decisions are made at lower organizational levels (e.g., members or locals rather than national headquarters). The NUS items each dealt with a different area of decision-making, namely bargaining, organizing, and ‘‘overall.’’ The SUIT items likewise each dealt with one of these issues. The scale anchors were simplified considerably for the shorter SUIT survey, however. This is discussed below in greater detail. Strategic Scope. Three items included in each survey were designed to assess the breadth of issues the union addresses and the breadth of methods the union uses to accomplish its goals. In the NUS, these items were part of a 10-item array intended to address union strategies more broadly. For the SUIT, we selected items that evidenced the strongest correlations with larger scales developed from the NUS. The three items from the SUIT are: (1) ‘‘This union devotes significant resources to all of the issues that concern our members, on or off the job;’’ (2) ‘‘In this union, providing good representation to our current members is more important than organizing new members’’ (reverse-scored); and (3) ‘‘Collective bargaining is our main concern, and all else is secondary’’ (reverse-scored). The wording of comparable items from the NUS is almost identical. Size. Union size is a single item that asked respondents to report their union’s ‘‘number of members.’’ Both the SUIT and the NUS posed virtually the same question.
Can Structural Reform Revitalize the American Labor Movement?
367
Membership Heterogeneity. Two items, percent of members that are female, and percent of members that are minorities, were combined to create our measure of membership heterogeneity. Following Delaney et al. (1996), we created an index where maximum heterogeneity occurs in the case where the union’s membership is 50% women and 50% minorities. For some comparisons below, we use the individual items on percent female and percent minority. IT Use. Two multi-item scales for IT use were initially constructed. The first combines 17 pairs of items. The first part of each paired set of items asked whether the respondent used a particular form of IT (e.g., word processing, e-mail), planned to use it in the next two years, or did not use it or have plans to use it. If the respondent indicated that the union currently used the IT form, he/she was asked to answer the second part of the paired set of items, which asked about the extent to which the IT form was used in a fourpoint Likert-style format. We combined these pairs of items to form 17 consolidated items, coded as: 1 ¼ Not used and no plans to use; 2 ¼ Not used but will be in next two years; 3 ¼ Used very little; 4 ¼ Used a moderate amount; 5 ¼ Used quite frequently; 6 ¼ Used extensively. These 17 consolidated items were summed and averaged to form an ‘‘IT Forms’’ scale. The second multi-item scale for IT use focused on applications. Eighteen pairs of items asked about areas in which IT was applied (e.g., to track organizing efforts, producing newsletters). Here too, each pair of items was consolidated, and all 18 items were summed and averaged to form a composite scale we call ‘‘IT Applications.’’ In the present study, we combined responses to both scales to form an overall ‘‘IT Use’’ scale. It is available only for the SUIT data. Organizing Commitment. This measure is the response to a single item included only in the SUIT. The item stated ‘‘Compared to other unions, this union is very committed to organizing.’’ Although this item has no NUS counterpart, it clearly is of contemporary interest. External Validity and Reliability Evidence External validity of the NUS has been addressed at some length in previously published studies (e.g., Fiorito et al., 1995). Table 1 provides a summary of evidence on the more recent SUIT’s external validity. Notice, for example, that two indicators of IT use from the SUIT correlate moderately with independent indicators of IT use, and that the SUIT respondents’ assessments of effectiveness in organizing and ‘‘overall’’ correlate with indicators of membership growth. Further, as we have noted elsewhere for the
368
JACK FIORITO ET AL.
Table 1.
Correlations for Independent and SUIT-Based Measures. Descriptives
Mean
Independent measures Web site (1 ¼ yes) E-mail addr. (1 ¼ yes) Fax number (1 ¼ yes)
Change in net worth Change in op. income Change in membership Change composite Membership growth, 1990–1997
Correlations with SUIT-Based Measures SD
IT Forms
IT Applications
0.55 0.19
0.50 0.39
0.47 0.11
0.35 0.16
0.92
0.27
0.29
0.39
Organizing Effectiveness
Organizational Effectiveness
0.20 0.24 0.26 0.26 0.33
0.28 0.30 0.45 0.37 0.29
140.8 16.5 11.8 0.1 4.1
250.4 53.9 30.8 0.9 32.5
Note: Existence of an e-mail address and fax number are based solely on these being listed in Gifford’s (1997) directory. Existence of a web site is based on Gifford’s directory supplemented by a web-based search of union-related web sites for links to and web sites of unions. Change measures are percentage changes for 1979–1993 from Masters and Atkin (1997). The Change Composite is the sum of standardized forms of the three change measures (a ¼ .83). Membership Growth is a percentage change based on reports in the NUS and SUIT (and is therefore not truly independent). po0.10 or better, two-tailed test. N’s for correlations involving IT measures range from 54 to 61. N for the correlations using Masters and Atkin’s data is 18. N’s for the correlations with Membership Growth are 53 to 54.
NUS (Fiorito, Jarley, and Delaney, 1993), evidence from within the survey showing leaders rate their union’s organizing effectiveness significantly lower than their union’s overall effectiveness seems to lend credibility to the data, and a similar pattern holds in the SUIT. It would be foolish to confuse perceptual data with reality, but our attempts to test the validity of the NUS and SUIT suggest that the data support inferences about reality. Related to the issue of external validity is the issue of reliability, which acts as a constraint on validity. As noted earlier, the instruments we adapted for use in the NUS were often ‘‘trimmed’’ of items in order to balance time demands on respondents with the scope of issues investigated. Given the
Can Structural Reform Revitalize the American Labor Movement?
369
direct relation between instrument length (number of items) and internal consistency (Cortina, 1993), this implies lower reliabilities than in the original instruments. Another factor possibly reducing reliability is the adaptation process itself. In so far as the instruments were generally developed to measure characteristics of employing organizations, one might expect that the process of transferring the items to unions reduces internal consistency. That is, items or concepts with clear meaning in one context may be somewhat ambiguous in another. For these reasons, the low to moderate internal consistency figures (Cronbach alpha, a) in Table 2 are not overly discouraging. It should be noted that the figures shown for the NUS measures generally understate the reliability of these scales as used in previous work. For urposes of this paper and Table 2, parsimonious versions of NUS instruments – ones that rely on just a few central items – are presented to enhance comparability across the two surveys. For example, the three-item Table 2. Internal Consistency and Interrater Reliability Statistics for Selected NUS and SUIT Variables. Variable
Organizational effectiveness Organizing effectiveness Innovation Rationalization Decentralization Size (membership) Heterogeneity Strategic scope IT use (types and applications) Organizing commitment
No. Items
3 1 2 5 3 1 2 3 34 1
Cronbach’s Alpha
Interrater Reliability (R2)
NUS
SUIT
NUS
SUIT
0.48 na 0.42 0.56 0.57 na 0.49 0.58 na na
0.69 na 0.70 0.59 0.51 na nc 0.32 0.94 na
nc nc nc 0.53a 0.56b nc 0.88 nc na na
0.70 0.66 0.51 0.41 0.54 0.99 0.85 0.31 0.72 0.57
Note: NUS-based interrater reliability information is from Jarley et al. (1997). Cronbach alphas shown here for NUS-based measures are based on composites constructed for comparability with SUIT measures, and are generally much lower than those based on more items used in our previous work based on the NUS. Shown for R2 from one-way ANOVAs specifying the variance among individual responses as a function of respondents’ union affiliations where multiple respondents reported for their union. N’s for the SUIT-based analyses range from 16 to 27. N’s for the NUS-based analyses are approximately 200. NA, Not applicable; NC, Not calculated. Significant at 90% confidence level or better. a For ‘‘administrative rationalization.’’ b For ‘‘centralized control.’’
370
JACK FIORITO ET AL.
organizational effectiveness scale from the NUS shows an a of 0.48 in Table 2. The six-item version used in previous work (Fiorito et al., 1993) has a ¼ 0.70 (or so). An interesting question for ‘‘measurement types’’ might be why the same three items yield a much higher figure in the SUIT, but for the moment we can only note that there were some slight differences in wording and response scales, that the two surveys used different methods (telephone versus mail), and that the NUS data usually involve the aggregation of individual responses to a ‘‘union-level’’ response, whereas the SUIT data usually take a single respondent’s answers as the union-level response. This last point calls attention to another aspect of reliability, interrater agreement. In most cases for the NUS, and some cases for the SUIT, union-level data are based on multiple responses (typically an average). Given this, how much do the respondents from within each union agree with each other? Interrater agreement presumably indicates homogeneity within the organization and perhaps a more objective underlying reality. That is, if two officials of the same union report very different assessments of their union’s organizing performance, that may say that the reality is highly subjective – it depends greatly on one’s perspective. One official may focus on the bottom line of membership growth, and be unduly influenced by gains from mergers in assessing organizing effectiveness, while the other focuses exclusively on the union’s success in organizing unorganized workers (as intended with this question). Very different responses could result. But in general, higher interrater agreement is desirable in that it suggests greater validity. The last two columns in Table 2 provide evidence on this issue from the NUS and SUIT. Not surprisingly, the greatest agreement arises for the most objective constructs, membership levels and membership heterogeneity (diversity in terms of gender and minority status of members). The R2 values for these indicate that most variance among individual respondents’ answers is accountable in terms of which union they describe. For some of the more perceptual constructs, on the other hand, it appears that the answer to questions depends greatly on the individual respondent’s perspective, which may indicate heterogeneity within the organization, that the questions involve a highly subjective matter, or other troubling measurement possibilities such as deliberate misrepresentation by some respondents. Although the interrater reliability values do not engender great confidence in the validity of some measures, we should note that this issue is often neglected entirely: An individual’s report is taken implicitly as accurately representing his/her organization. Values of 0.50 or so – some of the lower values here – are not unusual where researchers do assess interrater agreement.
Can Structural Reform Revitalize the American Labor Movement?
371
CHANGE IN AMERICAN UNIONS: NUS VS. SUIT COMPARISONS As noted earlier, a key purpose of this paper is to make comparisons over time. In order to make comparisons we took some liberties in formulating parsimonious versions of longer NUS scales (and consequently sacrificing reliability), converting scales with more anchor points to uniform four-point scales, and perhaps most importantly, comparing responses to somewhat different questions. To alert the reader to the importance of the latter, we will review the most questionable case of this sort. In the NUS, our measure of decentralization (actually centralization, but reverse-scored here for comparability) was based on a concept of control formulated by Tannenbaum and Kahn (1958), which focused on which of various groups within the union (e.g., top officers vs. rank-and-file members) exercised authority over various issues (e.g., organizing, bargaining). In the SUIT, we relied on Likert-style scales of agreement asking about rank-and-file influence on bargaining and organizing, and a third item stating ‘‘Locals (or similar bodies) really decide what this union does.’’ The correlation between the two measures is a modest 0.25 (po0.11 in a two-tailed test, shown as nonsignificant in Table 3). Given that we expected a positive correlation, a one-tailed test is appropriate, and Table 3. Correlations and Paired-Comparison t-tests for ‘‘Comparable’’ NUS and SUIT Variables. Variable
Organizational effectiveness Organizing effectiveness Innovation Rationalization Decentralization Strategic scope Size (membership) Percent female Percent minority
NUS & SUIT r
SUIT-NUS Mean
t
N
0.38 .41 0.26 0.15 0.25 0.46 0.97 0.89 0.75
0.37 0.60 0.17 0.26 0.12 0.08 8884 3.58 6.53
4.11 4.77 1.41 3.19 1.02 1.12 0.77 2.25 2.45
54 55 56 53 44 54 58 43 33
Note: We urge readers to view the paired differences for ordinal scales (all measures except Size, Percent female, and Percent minority) with particular caution. To make comparisons possible, many NUS-based measures were converted from 5- or 6-point scales to 4-point scales. Further, differences in item wordings to varying degrees limit comparability. Given these problems and that ordinal scales have no absolute meaning, caution is warranted in drawing inferences. Significant at 90% confidence level or better, two-tailed test.
372
JACK FIORITO ET AL.
hence we could pronounce this correlation ‘‘significant’’ at the 90% confidence level. Why is it not a stronger correlation? Four main possibilities come to mind: First, the correlation could be limited (or ‘‘attenuated’’) by the fact that the items and associated response anchors used to form the two composite scales differ considerably. Second, the correlation could be limited by the unreliability of both scales – using the standard correction formula and the a’s from Table 2, the observed score correlation of 0.25 translates into a ‘‘true-score’’ correlation of 0.46. Third, the respondents may differ somewhat between the two surveys. Although the overall proportions of staff and officers are virtually the same, the type of staff who were willing to fill out the two surveys probably differed. In the SUIT, we expressly sought IT expertise as well as more traditional staff (e.g., research directors and organizing directors). Finally, seven years passed between the two surveys, and it is quite possible that some unions have undergone real change in decision-making during this time. Unfortunately, it is not possible to know unequivocally how much of the difference between 0.46 (assuming the unreliability correction is valid) and a perfect 1.0 is due to the other three possibilities or still other (but probably less important) factors that might attenuate the correlation. Fortunately, comparability of wording is less of an issue for the remaining variables in Table 3, where items used in the SUIT follow the wording and format of NUS items more closely, or in the cases of membership levels and demographics, deal with more objective matters. Thus the positive correlations shown in the first column of figures in Table 3 suggest: (1) Some degree of test–retest reliability and/or (2) Some degree of stability in the characteristics shown, especially if corrections for unreliability are considered. If there has been real change in national unions over this time, what is the nature of that change? That’s the issue we attempt to get at in the second and third columns of figures in Table 3 by conducting paired-comparison tests for average levels of the variables shown. Before considering those figures and their meanings, a further caution is warranted. As noted above, in striving for comparability we took liberties in comparing items of varying similarity and varying original scale anchors. This may limit the correlations, as noted, but in comparing variables’ levels, even more caution is in order. Consider the case of temperature. The Fahrenheit and Celsius scales clearly measure the same thing. They differ in the values they record, but since one is an exact linear transformation of the other, their correlation is a perfect 1.0 but for rounding error. If we were comparing the level of
Can Structural Reform Revitalize the American Labor Movement?
373
temperatures with measures taken at two different times with one time reported on each scale, we might well find a significant difference due solely to the use of different scales, although after converting one set of measures to the other scale, we should observe no difference in levels if the temperatures were in fact the same. But Celsius and Fahrenheit are interval level measures. It is not clear the same would hold for ordinal level measures such as those used for most of the variables shown in Table 3. Accordingly, we urge readers to consider the differences in means and the associated t-values with added caution for the perceptual variables in Table 3. That said, what do the comparisons of variable levels over time suggest? First, the data on membership levels and demographics corroborate data from other sources (see U.S. Bureau of Labor Statistics 2005 for current figures): U.S. unions lost members (though not, on average, a significant number in statistical terms) while their memberships became (on average) increasingly constituted of women and minorities. Second, in the more hazardous comparisons of perceptual variables, it appears that U.S. unions saw themselves as less rationalized and less effective in organizing and overall in 1997 as compared to 1990, but the evidence does not support any conclusions of differences in perceived innovation, decentralization, or strategic scope. Although it is tempting to interpret the decline in perceived organizing effectiveness as a response to continued organizing failings over this time and/or a reflection of raised expectations in light of a renewed emphasis on organizing, there are no clear and readily available ‘‘real world’’ benchmarks against which to compare most perceptual variables. It is tempting however, to speculate that decline in perceived organizing effectiveness in the early 1990s along with continued membership declines were part of the same discontent that led to Sweeney’s ousting of Lane Kirkland in 1995. Has there been a real change in rationalization, or ‘‘structuring of activities’’? Recall that the rationalization measure consists of five items, representing distinct dimensions of standardization, formalization, coordination, communication, and environmental scanning (the extent to which specific individuals are assigned responsibility for monitoring environmental changes). Analysis at the item level (not shown) suggests that the change in the overall rationalization variable stems from declines in coordination and environmental scanning. We could speculate that environmental scanning since 1990 was diffused throughout unions rather than assigned to specific individuals, but we have no independent evidence to support this speculation. And, as yet we have no clear explanation for the perceived decrease in coordination that is more plausible than the possibility that the
374
JACK FIORITO ET AL.
difference is a methodological artifact. If ‘‘real,’’ could it reflect the impact of union mergers? For now, we will have to leave these matters as potentially interesting questions for further study. Which Unions Are Innovators? The first section of Table 4 presents a rank-order listing of unions in terms of their innovation ranking in the SUIT. Also shown is the union’s rank-order in terms of IT use, and the rank-order in terms of innovation based on items from the NUS. The positive correlation (r ¼ 0.26, po0.10) between the innovation rankings from the two surveys noted earlier and at the bottom of Table 4 suggests that, in general, those unions that were innovation leaders in 1990 tended to hold that status in 1997 as well. (Note that the correlation in innovation scores corrected for unreliability is 0.48.) At the same time, Table 4 makes obvious that some shuffling in the ranks took place, as illustrated dramatically by comparing the SUIT and NUS innovation rankings for the first two unions, LIUNA and HERE. Both unions surged from the middle of the pack to the forefront over this seven-year interval. Overall, these data lend support to the validity of the innovation scores, as the top-ranking unions on both the innovation and IT use score (a particular form of innovation, as suggested by its 0.46 correlation with the innovation score) are often those mentioned as innovation leaders. Of course, there are some notable absences in the first section of Table 4, but in part this is due to nonresponses for the SUIT. As shown in the third and last section of the table, at least four of the top-ranking innovators from the NUS could not be ranked on the SUIT innovation score due to a lack of data.
MODELS OF INNOVATION AND EFFECTIVENESS Research has paid some attention to innovation and its impact on effectiveness in national unions (e.g., Delaney et al., 1996; Fiorito et al., 1993, 1995, 2002; Shostak, 1991, 1999; Jarley, Harley, & Hall, 2002; Templer & Solomon, 1988). Although ‘‘new,’’ does not necessarily mean ‘‘improved,’’ and a few traditional unions may benefit little from change, the literature on unions is replete with calls for changing union tactics, strategies and structures, rather than simply doing traditional activities better (e.g., Heckscher, 1988, p. 23) and our own empirical research suggests innovation is an important determinant of union success, especially in organizing (Fiorito et al., 1995). Given the immediate focus on comparative results
Can Structural Reform Revitalize the American Labor Movement?
Table 4.
375
Top-Ranking Innovators or IT Users Based on SUIT Responses.
Union
Top 10 Innovation scores Laborers (LIUNA) Hotel and Restaurant (HERE) Teachers (AFT) Machinists (IAM) Treasury Employees (NTEU) Writers Guild, East (WGE) Graphic Artists Guild (GAG) Independent Pilots (IPA) Electrical Workers (IBEW) Office & Professional (OPEIU) Top 10 IT scores (if not shown above) National Education (NEA) Television and Radio (AFTRA) Boilermakers (BBF) Letter Carriers (NALC) Postal Workers (APWU) Communications (CWA) Selected Unions with No SUIT Data Food and Commercial (UFCW) Bricklayers (BAC) State, County, Mun. (AFSCME) Service Employees (SEIU)
SUIT Rank Innovation (of 73)
SUIT Rank IT (of 73)
NUS Rank Innovation (of 111)
1 2
8 21
49 54
3 4 5
na 2 12
13 8 15
6
na
na
7
na
na
8 9 10
7 6 28
na 85 na
17
1
10
59
3
84
21 49 18 24
4 5 9 10
48 14 33 26
na
na
2
na na
na na
3 6
na
na
9
Note: For SUIT Innovation score and SUIT IT scores, r ¼ 0.46 (po0.10), between NUS and SUIT Innovation scores, r ¼ 0.26 (po0.10), and between NUS Innovation score and SUIT IT score, r ¼ 0.11 (p>0.10).
376
JACK FIORITO ET AL.
between the NUS and SUIT surveys, a detailed literature review is not attempted. Rather, the main purpose here is to delineate some of the key elements of those models in order to set the stage for examination of comparative results. One central idea is that characteristics of unions and their environments influence the extent to which they innovate. This basic concept is elaborated more fully in Fiorito and Jarley (1992), Fiorito et al. (1993), and Delaney et al. (1996). Just to cite one example of the linkages identified in the literature, size (e.g., number of members) connotes resources, and given that some innovations require substantial investments, one should expect that size will be positively associated with innovation. A second basic idea is that innovation, in turn, influences organizational performance (Fiorito et al., 1993, 1995; Shostak, 1991). Caution is advisable in transplanting business strategy concepts to unions, but nonetheless, it would seem that Barney’s (1997) notion of ‘‘organization’’ as a source of sustainable competitive advantage, or Liebenstein’s (1966) notion of ‘‘X-efficiency,’’ has relevance. Although unions do not generally compete with each other, innovation offers potential for unions to improve their services to members, their political and public relations efforts, their performance in bargaining, and their ability to organize new members. Hence while head-to-head competition between unions is atypical, innovation offers a potential source of competitive advantage when they do. More importantly, innovation offers unions a potential advantage in competing with employers over the form of workplace governance (i.e., employer unilateralism vs. joint union-management determination). An example here is the development (or perhaps redevelopment) and use of ‘‘salting’’ tactics whereby union members seek employment in nonunion firms for the purpose of organizing those firms. ‘‘Salts’’ have considerably greater access to potential members than do nonemployee organizers, enabling them to reduce one of the employer’s greatest advantages over unions, access to the workforce. Salting is one narrow form of innovation. The two items available from both surveys, dealing with ‘‘new methods’’ and ‘‘polls or surveys’’ for organizing new members and servicing current members, are clearly broader, although still limited in scope. We emphasize that these items are merely an effort to sample from the broader domain of possible innovation items that represent innovation as a broad organizational construct (Damanpour, 1991). These particular items were part of a larger scale in the NUS, but retained for the SUIT because of their centrality in the longer innovation scale, i.e., their strong item-total score correlations.
Can Structural Reform Revitalize the American Labor Movement?
377
Below, hypotheses reflect these basic ideas, which cumulatively constitute models of union innovation and effectiveness. Note, however, that the models outlined below are incomplete (one might say ‘‘parsimonious’’) versions of models developed more fully in previous literature (e.g., Delaney et al., 1996; Fiorito et al., 1995). We do not include measures of the environment in this paper in part because of the difficulties in developing consistent environment measures across a span of seven years. In addition, environment measures often performed poorly in prior efforts to explain variation in union structures, innovation, or outcomes. This was especially evident in efforts to explain perception-based dependent variables such as those emphasized here. Measures such as environmental complexity do show some ability to explain differences in centralization of decision-making and measures of latent demand for union representation do appear to influence organizing effectiveness, but in general these measures fail to explain a significant degree of the variation in either union structures or innovation (see Delaney et al., 1996; Fiorito et al., 1995; Jarley et al, 1997). Inadequate measurement of the environment remains a key potential confounding explanation for results across these studies, but it is also possible that much of the variance in structures, innovation and even outcomes in unions, organizations that are almost uniformly in decline and threatened with extinction, is more a product of their own capacity or willingness to implement change rather than of environmental imperatives (Delaney et al., 1996, p. 609). Hypotheses Causes of Innovation. The arguments and results in the Delaney et al. (1996) study suggest that innovation will be greater in large unions and unions with higher levels of rationalization (or structuring of activities). In addition, although Delaney et al. obtained weak results for variables representing centralization and democracy, the arguments for a positive effect of decentralization are appealing and supported in the broader literature on innovation (Damanpour, 1991), and a positive relation is expected. In addition, although not directly examined in Delaney et al. (1996), a positive relation between innovation and strategic scope is anticipated. Delaney et al. also found a positive relation between membership heterogeneity and innovation. It seems likely that membership heterogeneity correlates with strategic scope however, and thus the latter may act as a proxy for the former. In addition, Fiorito and Jarley (1992) found a positive link between strategic scope and a particular form of innovation (associate memberships).
378
JACK FIORITO ET AL.
These hypotheses regarding innovation measures are examined in both the NUS and SUIT, and for an IT dependent variable using the SUIT data. Consequences of Innovation. Innovation should be linked to union effectiveness in organizing and more generally. Fiorito et al. (1995) showed that innovation in national unions was linked to a variety of indicators of organizing effectiveness, including both perceptual and relatively objective measures (i.e., NLRB certification win rates and membership growth rates). Fiorito, Stepina, Jarley, Delaney, and Knudstrup (1997) found a similar linkage between perceptions of innovation and perceptions of effectiveness in the minds of union leaders, where effectiveness was operationalized to include reference to organizing, politics, bargaining, servicing members, advancing the interests of all workers, and an ‘‘overall’’ concept of effectiveness. For similar reasons and based on previous findings (Fiorito et al., 2002), a positive relation between IT use and effectiveness indicators is anticipated, but these links can only be examined in the SUIT data. Following arguments (and to some extent results) in Fiorito et al. (1995, 1997, 2002), rationalization, decentralization, strategic scope, and size should have positive effects on union effectiveness in organizing and ‘‘overall.’’ In addition, for the SUIT dataset a measure of organizing commitment is available and expected to have a positive relation to organizing effectiveness. Results for ‘‘Comparative Statics’’ Models Table 5 presents bivariate correlations pertaining to the hypotheses for variables in both the NUS and SUIT datasets. Bivariate correlations typically provide only a limited indication of support (or lack thereof) for multivariate hypotheses (e.g., with ‘‘all else equal’’). With this caution in mind, the correlation results provide support for most of the hypotheses. Innovation shows a weak to moderate positive relation with rationalization, decentralization, strategic scope, and size in one or both datasets. Also, one or both effectiveness measures correlates positively with these measures and innovation in both datasets. Overall, the similarity of the relations is far more striking than the differences, although there some slight differences do appear. Table 6 provides standardized OLS regression results for innovation in both datasets, and for IT use in the SUIT dataset. Consistent with prior findings and predictions, positive effects are consistently shown for rationalization and size across all three equations. The hypothesized positive effect for decentralization only finds support in the SUIT data, however, and that for strategic scope is only supported in NUS data.
Can Structural Reform Revitalize the American Labor Movement?
Table 5.
Correlations Matrices (Decimals Omitted) for Selected NUS and SUIT Variables.
NUS Data Variables 1. 2. 3. 4. 5. 6. 7.
Organizational effectiveness Organizing effectiveness Innovation Rationalization Decentralization Strategic scope Size (membership)
SUIT Data Variables 1. 2. 3. 4. 5. 6. 7.
379
Organizational effectiveness Organizing effectiveness Innovation Rationalization Decentralization Strategic scope Size (membership)
1
2
3
4
5
6
76a 42 31 22 11 16
36 24 23 27 22
45 15 35 31
10 10 7
6 29
26
1
2
3
4
5
6
77a 62 38 42 18 36
69 21 34 32 39
42 29 14 35
16 0 25
17 13
33
Note: Sample sizes for NUS-based correlation range from 93 to 109. Sample sizes for SUITbased correlations range from 63 to 73. a This is a part-whole correlation since the Organizing Effectiveness item is one of three included in the Organizational Effectiveness composite. Significant at the 90% confidence level or better in a two-tailed test.
Table 6.
Standardized Regression Results for Innovation and IT Use.
Variable Rationalization Decentralization Strategic scope Size (membership) R2 Adj. R2 N
NUS Innovation
SUIT Innovation
SUIT IT Use
0.42 0.00 0.18 0.26 0.34 0.31 90
0.26 0.22 0.16 0.22 0.28 0.23 63
0.24 0.39 0.06 0.39 0.34 0.27 45
po0.10 or better (one-tailed tests for slope coefficients).
380
JACK FIORITO ET AL.
Table 7 offers standardized OLS results for (‘‘overall’’) organizational effectiveness. As predicted, and consistent with previous research, innovation and decentralization evidence positive effects across both datasets. The predictions for rationalization and strategic scope, however, only find partial support in the SUIT dataset. No support is found for the predicted effects of size or IT use. Finally, Table 8 shows OLS results for organizing effectiveness. In accord with previous findings and predictions here, innovation and decentralization show consistent positive influences on organizing effectiveness. Also as predicted, strategic scope, IT use, and organizing commitment show positive Table 7. Standardized Regression Results for Organizational Effectiveness. Variable Innovation Rationalization Decentralization Strategic scope Size (membership) IT use R2 Adj. R2 N
NUS
SUIT
SUIT
0.42 0.11 0.18 0.00 0.07
0.42 0.14 0.33 0.19 0.06
0.31 0.19 0.41 0.15 0.01 0.03
0.26 0.21 90
0.53 0.49 61
0.48 0.40 45
po0.10 or better (one-tailed tests for slope coefficients).
Table 8.
Standardized Regression Results for Organizing Effectiveness.
Variable
NUS
SUIT
SUIT
SUIT
Innovation Rationalization Decentralization Strategic scope Size (membership) IT use Organizing commitment
0.22 0.10 0.15 0.19 0.05
0.60 0.09 0.23 0.26 0.08
0.51 0.01 0.29 0.20 0.04 0.22
0.43 0.02 0.24 0.33 0.03 0.22 0.21
R2 Adj. R2 N
0.20 0.16 90
0.61 0.57 61
0.62 0.56 45
0.64 0.57 45
po0.10 or better (one-tailed tests for slope coefficients).
Can Structural Reform Revitalize the American Labor Movement?
381
impacts. All of these results are consistent across the two datasets when comparisons are possible. No support is found for the hypothesized effects of rationalization and size in either dataset. The results shown in Tables 5–7 for innovation, organizational effectiveness, and organizing effectiveness are generally consistent with previous findings. It is notable that these relatively parsimonious versions of more complete models incorporating environment measures (e.g., Delaney et al., 1996; Fiorito et al., 1995) produce relatively similar results for common variables. This suggests that omitting environment measures has not produced important distortions in the results for the models herein. This is not to say that environment measures are not important, of course. Previous studies have shown an important role in some instances, particularly in predicting more objective indicators of union organizing success (Fiorito et al., 1995). Longitudinal and Change Model Results Up to this point all models have reflected a static approach. Outcome (e.g., organizing effectiveness) or intermediate outcome (i.e., innovation) measures based on data from a point in time were modeled as a function of contemporaneous predictor measures. That some results are consistent across datasets gathered at two points in time seven years apart is encouraging and perhaps hints of model stability in a ‘‘comparative statics’’ sense. More dynamic approaches are possible. One such approach is longitudinal. Our (partially) longitudinal models specify criterion measures from the 1997 SUIT or other 1997-era data as a function of contemporaneous predictor measures from the 1997 SUIT, with the important exception of the innovation measure, which is taken from the 1990 NUS. The reasoning behind this specification is that innovation’s effects are likely to unfold slowly. Further, since the innovation models we have estimated specify contemporary predictors, this specification avoids problems associated with endogeneity within the predictors of outcomes. Table 9 presents results for four models that are broadly comparable to models presented above except for the lag on innovation and the addition of Membership Growth as a criterion. As could be expected, results for the contemporaneous predictors in Table 9 are quite similar to those for comparable relations in Tables 7 and 8. The lagged innovation predictor is less successful, but still evidences a significant effect on organizing effectiveness. This is mildly encouraging, adding a bit more causally persuasive evidence to the mix, albeit just a small bit. One reason that the innovation results here
382
JACK FIORITO ET AL.
Table 9. Partially Longitudinal Models (NUS-Based Innovation; All Other Measures Involve SUIT). Union Union Organizing Organizing Effectiveness Effectiveness Effectiveness Effectiveness
Membership Growth
Membership Growth
Predictors Innovation Rationalization Decentralization Strategic scope Size
0.12 0.31 0.46 0.23
0.14 0.30 0.44 0.18 0.09
0.23 0.21 0.41 0.30
0.22 0.16 0.37 0.24 0.17
0.07 0.08 0.11 0.03
0.08 0.03 0.08 0.02 0.17
Summary statistics R2 Adj. R2 N
0.45 0.40 49
0.46 0.40 48
0.44 0.39 49
0.46 0.40 48
0.02 0.07 49
0.04 0.07 49
po0.10 or better (one-tailed tests for regression coefficients).
are less impressive than those shown earlier (Tables 7 and 8) might be that the seven-year lag is too long. We noted earlier that there had been substantial churning or shuffling of union ranks in some regards (see Table 4). The complete failure of the Membership Growth model may reflect an added complication. To wit, previously published results for more complete static models including criteria such as membership growth that are virtually independent of perceptual bias have underscored the importance of environmental as well as organizational factors (e.g., Fiorito et al., 1995). It may be possible to abstract from the environment in examining perception-based measures since survey respondents may net out much environmental ‘‘contamination’’ (Fiorito et al., 1995), yet foolhardy to take this approach for more objective measures that clearly are sensitive to environmental factors (e.g., membership is clearly sensitive to declining industry employment). Unfortunately, the data requirements for specifying environment measures at two points in time for environmental factors such as employer opposition and latent demand for union representation are substantial, and further investigation of this issue will remain an issue for future research.1 The change (i.e., ‘‘first differences’’) models in Table 10 fare slightly better than the longitudinal models, with the notable exception of the Membership Growth criterion (which is actually a growth rate and not a simple difference). For the remaining criterion measures, innovation, organizational effectiveness, and organizing effectiveness, all of the predictors except union size yield results fairly comparable to those in the static models (Tables 6–8). That is, the change model results suggest that changes in rationalization and
Change Models (All Variables Except Membership Growth are Simple Differences for 1997 vs. 1990).
Change IVs
Standardized Regression Coefficients for Predictors (Change IVs) Innovation
Innovation Rationalization Decentralization Strategic scope Size
0.16 0.22 0.00
Summary statistics R2 Adj. R2 N
0.23 0.15 41
0.33
Organizational Effectiveness
Organizational Effectiveness
Organizing Effectiveness
Organizing Effectiveness
Membership Growth
Membership Growth
0.29 0.28 0.43 0.05
0.29 0.28 0.43 0.03 0.09
0.43 0.00 0.24 0.05
0.43 0.01 0.23 0.04 0.11
0.23 0.01 0.04 0.07
0.23 0.01 0.04 0.08 0.03
0.50 0.44 40
0.51 0.43 40
0.28 0.20 40
0.29 0.19 40
0.05 0.05 41
0.05 0.08 41
po0.10 or better (one-tailed tests for regression coefficients).
Can Structural Reform Revitalize the American Labor Movement?
Table 10.
383
384
JACK FIORITO ET AL.
strategic scope predicted innovation changes, that changes in rationalization, strategic scope, and decentralization predicted organizational effectiveness changes, and that changes in innovation and decentralization predicted changes in organizing effectiveness. These results add to the circumstantial evidence supporting the causal inferences that we suggested earlier. No statistical procedure can prove causality, and one can craft noncausal or reverse-causality explanations for these results. The results for change models, however, add to a growing body of evidence that is consistent with positive rationalization and strategic scope effects on innovation, positive innovation, rationalization, and decentralization effects on organizational effectiveness, and positive innovation and decentralization effects on organizing effectiveness.
CONCLUSIONS AND FUTURE EXTENSIONS Three potentially major limitations affect this study, and should be noted before conclusions are offered. First, as noted in more detail above, there are methodology problems here that occur in most survey research: Nonresponse bias, unreliability, limited validity, etc. This paper has attempted to assess these issues and thereby alert readers that they might affect inferences. Second is the issue of heterogeneity, that is, the possibility that the national unions treated implicitly as internally uniform may be in fact heterogeneous. Clearly, this is a potentially greater problem in amalgamated or recently merged unions where distinct subunits may be relatively autonomous. Third, one must always be cautious in leaping from statistical associations to causal inferences, particularly in cross-sectional analyses where associations may quite plausibly reflect reverse causality or spurious associations. With appropriate caution noted, this research demonstrates that survey research can provide a useful tool for learning about unions. The various unions operating in the real world provide a ‘‘natural experiment’’ for researchers to study. Their varied policies and experiences should be informative in both failure and success. Of course, broad-based survey research such as this can most certainly be complemented by more in-depth studies of particular unions. Such research, e.g., case studies and intensive interviews, helps to better establish the meaning of findings. This is a good juncture to revisit the issue of comparability. Although the focus of the surveys shifted somewhat between 1990 and 1997, enough commonality was maintained to offer meaningful comparisons. In that several of the key findings from the earlier NUS were supported in the more
Can Structural Reform Revitalize the American Labor Movement?
385
recent SUIT, confidence in those findings is bolstered. Exploiting the longitudinal potential in the combined data and change models are potentially illuminating steps only possible by maintaining some consistency in surveys over time. In short, there is considerable virtue in consistency per se. As in research in some other areas, e.g., the General Social Survey, researchers need to think about the notion of ‘‘core items’’ which should be retained over time and should perhaps be the subject of discussion among researchers working along roughly parallel lines in different nations. The initial substantive findings constitute a constructive replication rather than an extension of previous research. Thus, confidence is now increased for earlier conclusions that rationalization and size are important determinants of innovation in national unions. In addition, there is stronger evidence now than in past work supporting a positive role for decentralization as an influence on innovation. Similarly, results for organizational and organizing effectiveness bolster previous findings of an important role for innovation and decentralization. At least for organizational effectiveness, there is stronger but still equivocal evidence that rationalization and strategic scope are positive influences. The results reviewed thus far, however, are based mainly on crosssectional analyses at two points in time. From a perspective of seeking to understand change, these results could be said to provide ingredients for a comparative statics analysis at best. The extensions of this work reflected in Tables 9 and 10, including more complete exploration of the longitudinal and dynamic potential in the data help clarify the nature of change per se in national unions. Although the longitudinal model results were generally disappointing, the change model results were generally quite encouraging. They show that most of the positive findings in previous cross-sectional analyses are also supported in terms of changes within unions. Thus for example, not only are high levels of innovation associated with (possibly causing) high levels of organizing effectiveness among cross-sectional samples at two points in time, but unions experiencing the largest increases in innovation are those experiencing the largest increases in organizing effectiveness across the 1990–1997 period. This too could reflect a mere association, but the odds seem strongly against it. Thus the change models’ results seem to move us substantially closer to strong causal inferences. Up to this point we have adopted a proper degree of academic caution. That approach is sometimes too conservative, and we would like to conclude with a more speculative and provocative extrapolation of the results and how they relate to the rebellion within the AFL-CIO. Based on the findings and the nature of the AFL-CIO vs. CWF schism as it has unfolded
386
JACK FIORITO ET AL.
to this point, the following observations, based largely on the results presented above, come to mind: Organizing commitment seems to matter. The CWF unions insist that organizing must come first. If they enact this priority, our results suggest there should be a noticeable impact in organizing success. Recognize, however, that pro-organizing rhetoric alone will not be sufficient for success! Rationalization might matter. In theory at least, mergers or consolidations of unions to better match contemporary industry and corporate structures should bring greater union efficiency, effectiveness, and perhaps innovation. However, union mergers and absorptions are often more opportunistic than rational, and politically constrained, so this requires observation and assessment. Innovation clearly matters. A persistent and strong finding across various union outcome measures is the positive role of innovation. Our innovation measures are rather modest in that they tend to tap incremental changes in organizing, bargaining, mobilizing, and servicing members. They do not reflect radical change, but rather what might be termed ‘‘best practices.’’ Yet, their impact is substantial. Aside from the efficacy of the various practices that make up our measures, it is worth noting that innovation tends to permeate all aspects of an organization and as such our measures are tapping a general proclivity to innovate. Such a proclivity requires adoption of a mindset that looks for new opportunities and ways to overcome obstacles rather than simply doing the same thing over and over again expecting different results. To the extent that the CWF unions’ secession allows and encourages more experimentation from unions in either federation, there should be a positive impact on union outcomes such as organizing success. Recall that one of the greatest periods of growth in American unions occurred during a time when there were two rival federations competing for members. How size matters. A key argument in the CWF plan is the need to restructure the union movement into a smaller number of larger unions that will have the resources to employ the state-of-the-art practices necessary to succeed. Our results suggest that size tends to have indirect effects on union outcomes through its impact on innovation and rationalization. The relationship between size and innovation is fairly robust and consistent with the CWF view. The relationship between size and rationalization is more mixed. Results from Jarley et al. (1997) suggest that size has a positive impact on structural differentiation in unions, but not on
Can Structural Reform Revitalize the American Labor Movement?
387
administrative practices. Furthermore, results from Fiorito et al. (1995) suggest that unions that specialized in organizing particular types of workers had greater organizing effectiveness than those that used a more ‘‘shotgun’’ approach. Taken together these results suggest that size may create slack resources that can be used to invest in innovation, but that focused growth rather than unrelated diversification might be necessary to yield efficiencies and organizing gains. Decentralization matters too. There are undoubtedly some things that are best done by or closely coordinated by central authorities and staff within unions. Perhaps establishing and guiding the Organizing Institute is a good example. How the graduates of the Institute are used is another matter, and this is where the virtues of decentralization may come into play. They also come into play more generally in terms of involving rankand-file members in conducting the union’s ‘‘business.’’ This result may be a ‘‘macro’’ (national union level) counterpart to conclusions about the organizing and servicing models or the value of rank-and-file intensive organizing campaigns. Consistent with claims from many impressionistic assessments and limited empirical evidence, this result suggests that unions that involve and empower members are more successful in both organizing and in overall union effectiveness. How the CWF and AFL-CIO unions contrast on decentralization is not yet clear. Certainly, some of the CWF affiliates have a relatively decentralized tradition in many regards that reflects a local product market orientation (e.g., SEIU, Teamsters, UFCW). Governance systems might also matter. Jarley, Fiorito, and Delaney (2000) used cluster analysis to analyze national union constitutions. They identified four distinct governance system types. We could not help but notice that all of the CWF unions (so far) for which Jarley et al. had data (six unions) fell into their ‘‘Group 3.’’ Group 3 unions had moderate codification, low member access, moderate federation, moderate division of authority, and low activity levels indicated in their constitutions. This is either an incredible coincidence, or a reflection of something systematic. It may be that governance systems partly reflect common environmental circumstances (e.g., local product markets and thus relatively strong locals vis-a`-vis their national), as Jarley et al. noted, and that these same environmental factors favor secession. In this sense, the relation may be spurious, due to a common cause, but nonetheless important in underscoring the role of the environment for various outcomes. Another possibility is that there is in fact a causal relation. We know that the secessions of 2005–2006 did not cause the union’s constitutional
388
JACK FIORITO ET AL.
characteristics, and thus if there is a causal relation it is in governance systems facilitating secession. Group 3 unions were not wildly extreme in any of the dimensions, but they did exhibit the lowest scores for member access and activity levels. These characteristics suggest unions where leaders are relatively insulated from member influence directly (e.g., recall elections, referenda), and indirectly via limited governance activity levels (e.g., infrequent executive board meetings or elections). Previously, the ‘‘coincidence’’ of persistent corrupt practices was noted for some Group 3 unions (Jarley et al., 2000). That too may be spurious, but again possibly underscoring the role of environment. More intriguing possibilities exist, of course, such as certain governance systems enabling the Lewis-like ‘‘strong leader’’ to enact significant change, possibly including ‘‘change to organize.’’ This deserves further study. In sum, there are some evidence-based reasons to be optimistic about the effects of the CWF unions’ secession. The national unions that seceded and makeup the new CWF coalition appear to have some of the characteristics associated with innovation, decentralization, and organizing success. Yet it is one thing to say that these unions have such characteristics and quite another to argue that the new CWF will be more effective than the AFL-CIO at influencing how their affiliates function and allocate resources. Decentralization may spur innovation in national unions, but some changes may require a top-down approach from a rather heavyhanded federation demanding change. Is Andy Stern today’s John L. Lewis? Will he provide the aggressive leadership that Lewis did in the 1930s? And, what will it take to tap today’s pent up demand for some form of collective representation that has been suggested in various polls (e.g., Freeman & Rogers, 1999)? These are questions that extend beyond the scope of this paper. But the answers to these questions will shape the labor movement’s future. If the CWF is successful in pushing for change and this change matters, it may spur similar reforms in the AFL-CIO just as the CIO split influenced the AFL back in the 1930s. We realize that some will see this as heresy or playing into the hands of the forces that would divide and conquer unions and workers. It may be heresy, and there is always the risk that division will mean weakness. But U.S. unions have been relatively united and shrinking in importance for 50 years. At a time when the issues historically championed by unions – e.g., health care, pensions, and more generally, equity – are increasingly important, maybe its time for something different.
Can Structural Reform Revitalize the American Labor Movement?
389
NOTES 1. We also considered fully longitudinal models comparable to those in Table 9, and for innovation, in all instances predicting 1997 SUIT-related measures (and membership growth for 1990–1997) solely from circa 1990 NUS-related measures. These models showed some results parallel with those in Tables 6–9, but were generally disappointing, seeming to underscore the lag period and omitted environmental predictors problems cited already.
ACKNOWLEDGMENTS The study reported here is possible due to the efforts of many individuals. First and foremost, the union officers and staff members who completed the National Union Survey (NUS) and the Survey of Union Information Technology (SUIT) made this study possible. In addition, there were many research assistants, colleagues, union officials, and family members who contributed to our efforts in various ways, and they are too numerous to mention individually. We apologize for failure to acknowledge each, but we are truly grateful for their assistance.
REFERENCES AFL-CIO Executive Council. (2005). Winning for working families. Posted on-line at www.aflcioorg Barney, J. B. (1997). Gaining and sustaining competitive advantage. Reading, MA: Addison-Wesley. Bronfenbrenner, K. (1997). The role of union strategy in NLRB certification elections. Industrial and Labor Relations Review, 50(2), 195–212. Bronfenbrenner, K. (2001). Changing to organize. The Nation, (September 3), posted on-line at www.thenation.com Cleeland, N. (2005). AFL-CIO President proposes shake-up. Los Angeles Times (April 29), posted on-line at www.latimes.com Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78(1), 98–104. Damanpour, F. (1991). Organizational innovation: A meta-analysis of effects of determinants and moderators. Academy of Management Journal, 34(3), 555–590. Delaney, J. T., Fiorito, J., & Jarley, P. (1991). Union innovation and effectiveness: Results from the national union survey. Iowa City, IA: Industrial Relations Institute, University of Iowa. Delaney, J. T., Jarley, P., & Fiorito, J. (1996). Planning for change: Determinants of innovation in U.S. national unions. Industrial and Labor Relations Review, 49(4), 597–614. Fiorito, J., & Hendricks, W. E. (1987). The characteristics of national unions. In: D. Lewin, D. B. Lipsky, & D. Sockell (Eds), Advances in industrial and labor relations (Vol. 4, pp. 1–42). Greenwich, CT: JAI Press.
390
JACK FIORITO ET AL.
Fiorito, J., & Jarley, P. (1992). Associate membership programs: Innovation and diversification in national unions. Academy of Management Journal, 35(5), 1070–1085. Fiorito, J., & Jarley, P. (2003). Union organizing commitment: Rhetoric and reality. Proceedings of the fifty-fifth annual meeting of the industrial relations research association (pp. 283–293). Champaign, IL: IRRA. Fiorito, J., Jarley, P., Delaney, J.T. (1993). National union effectiveness. In: S. B. Bacharach, R. Seeber & D. Walsh (Eds), Research in the sociology of organizations (Vol. 12, pp. 111–137) Greenwich, CT: JAI Press. Fiorito, J., Jarley, P., & Delaney, J. T. (1995). National union effectiveness in organizing: Measures and influences. Industrial and Labor Relations Review, 48(4), 613–635. Fiorito, J., Jarley, P., & Delaney, J. T. (1998). Luddites, IT, and cyberunions: Information technology and national unions. Final baseline report on the survey of union information technology. Tallahassee, FL: Florida State University. Fiorito, J., Jarley, P., & Delaney, J. T. (2000). The adoption of information technology by U.S. national unions. Relations Industrielles/Industrial Relations, 55(3), 451–476. Fiorito, J., Jarley, P., & Delaney, J. T. (2002). Information technology, union organizing, and union effectiveness. British Journal of Industrial Relations, 40(4), 627–658. Fiorito, J., Jarley, P., Delaney, J. T., & Kolodinsky, B. (2000). Unions and information technology: From luddites to cyberunions? Labor Studies Journal, 24(4), 3–34. Fiorito, J., Stepina, L. P., Jarley, P., Delaney, J. T., & Knudstrup, M. (1997). Visions of success: National leaders’ views on union effectiveness. Labor Studies Journal, 22(1), 3–19. Freeman, R. B., & Rogers, J. (1999). What workers want. Ithaca, NY: ILR Press. Gifford, C. D. (1997). Directory of U.S. labor organizations, 1997 Edition. Washington, DC: Bureau of National Affairs. Greenhouse, S. (2005). AFL-CIO Lays Off 105, but discord grows louder. New York Times (May 8), posted on-line at www.nytimes.com Heckscher, C. C. (1988). The new unionism. New York: Basic Books. Hurd, R. W. (2004). The failure of organizing, the new unity partnership, and the future of the labor movement. Working USA, 8(September), 3–25. Jarley, P., Fiorito, J., & Delaney, J. T. (1997). A structural contingency approach to bureaucracy and democracy in U.S. national unions. Academy of Management Journal, 40(4), 831–861. Jarley, P., Fiorito, J., & Delaney, J. T. (2000). Union governance: An empirically grounded systems approach. Journal of Labor Research, 21(2), 227–246. Jarley, P., Harley, B., & Hall, R. (2002). Innovation and change in Australian trade unions. Industrial Relations, 41(2), 228–248. Knoke, D. (1990). Organizing for collective action: The political economies of associations. New York: de Gruyter. Leibenstein, H. (1966). Allocative efficency v. x-efficiency. American Economic Review, 56(June), 392–415. Masters, M. F. (1997). Unions at the crossroads. Westport, CT: Quorum Books. Masters, M. F., & Atkin, R. S. (1997). The finances of major U.S. unions. Industrial Relations, 36(4), 489–506. Price, J. L., & Mueller, C. W. (1986). Handbook of organizational measurement. Marshfield, MA: Pitman. Shostak, A. B. (1991). Robust unionism: Innovation in the labor movement. Ithaca, NY: ILR Press.
Can Structural Reform Revitalize the American Labor Movement?
391
Shostak, A. B. (1999). Cyberunion: Empowering labor through computer technology. Armonk, NY: M.E. Sharpe. Stratton, K., & Brown, R. B. (1989). Strategic planning in U.S. labor unions. Proceedings of the forty-first annual meeting (pp. 523–531). Madison, WI: Industrial Relations Research Association. Tannenbaum, A. S., & Kahn, R. L. (1958). Participation in local unions. Evanston, IL: Row, Peterson. Templer, A., & Solomon, N. (1988). Unions and technology: A survey of union use of information technology. Relations Industrielles, 43(2), 378–393. U.S. Bureau of Labor Statistics. (1980). Directory of national unions and employee organizations, 1979. Washington, DC: U.S. Government Printing Office. U.S. Bureau of Labor Statistics. (2005). Union members in 2004. Posted on World Wide Web at ftp://ftp.bls.gov/pub/news.release/union2.txt
