Money, Financial Intermediation and Governance

Money, Financial Intermediation and Governance To our teachers Money, Financial Intermediation and Governance Dino ...

Author: Dino Falaschetti | Michael J. Orlando

79 downloads 1760 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Money, Financial Intermediation and Governance

To our teachers

Money, Financial Intermediation and Governance Dino Falaschetti Florida State University College of Law, Stanford University Hoover Institution and Economic Advisors, Inc., USA

Michael J. Orlando Economic Advisors, Inc., USA

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Dino Falaschetti and Michael J. Orlando 2008 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited Glensanda House Montpellier Parade Cheltenham Glos GL50 1UA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2008926566

ISBN 978 1 84542 870 9 Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents List of ﬁgures Foreword, Antoine Martin About the book About the authors

vii ix xi xii

Introduction PART I

1

THE ECONOMIC METHOD OF INQUIRY

1. Modeling 2. Foundation 3. A model economy PART II

9 15 25

MONEY IN A STATIC ECONOMY

4. What is money? 5. Money and the level of economic well-being PART III 6. 7. 8. 9.

MONEY IN A DYNAMIC ECONOMY

Money in a classical economy Money in a Keynesian economy Should monetary policy be active? Is monetary policy active?

PART IV

35 42

53 66 80 87

GOVERNING MONEY

10. Measuring monetary services 11. Organizing the production of monetary services 12. The case of the ‘Fed’

99 102 109

PART V INTERMEDIATION, GOVERNANCE AND ECONOMIC PERFORMANCE 13. Asymmetric information

115 v

vi

Contents

14. Financial intermediaries and their governance 15. Corporate governance 16. Financial development and economic performance

121 135 145

Questions Bibliography Index

151 159 175

Figures I.1 1.1

Organization of the book Reduced form relationship between red wine and heart disease 1.2 Omitted variables bias 1.3 Structural evidence 2.1 The moose’s energy maximization problem 3.1 The household’s utility maximization problem 3.2 An increase in relative prices shrinks the household’s consumption possibilities 3.3 The ﬁrm’s proﬁt-maximization problem 3.4 An increase in output prices increases output 4.1 Commodity money distorts relative prices 4.2 Commitment problems when moving from commodity to ﬁat money? 5.1 Inferiority of the (costly) barter equilibrium 5.2 Inferiority of barter in a ‘linear city’ 5.3 Barter further constrains economic performance in production economies 6.1 The ﬁrm’s proﬁt-maximization problem 6.2 The labor demand curve 6.3 Equilibrium in the labor market 6.4 Classical (long-run) aggregate supply 6.5 Classical aggregate demand 6.6 Goods market equilibrium 6.7 Goods market response to an increase in the money supply 6.8 A negative technology shock, and its inﬂuence on the quantity of labor demanded for any real wage 6.9 Labor market reaction to a negative technology shock 6.10 Goods market reaction to a negative technology shock 7.1 Keynesian (short-run) aggregate supply 7.2 Involuntary unemployment with eﬃciency wages 8.1 Lucas’s critique and the Phillips Curve’s breakdown 9.1 Optimal policy in the hangman game is inconsistent 9.2 The optimal monetary policy is inconsistent 13.1 Cash-constrained proﬁt maximization vii

2 12 12 13 18 26 27 28 29 38 40 43 45 48 54 55 57 58 60 61 62 63 64 64 67 73 81 89 92 116

viii

Figures

13.2 Moral hazard 14.1 Put option analysis of public insurance crises 14.2 A put option’s value increases with the underlying asset’s risk 16.1 Channels from ﬁnancial development to economic performance

119 128 130 147

Foreword Most of modern economics relies on the assumption that economic actors are ‘maximizers’ (i.e., households maximize utility and ﬁrms maximize proﬁt). This assumption has proved fruitful because it permits a systematic study of the interactions between economic agents through well-developed models. Despite the success of this approach, it is largely absent from existing intermediate teaching materials. Money, Financial Intermediation and Governance addresses this void by continually exploiting the assumption that economic actors are ‘maximizers’ to deepen our understanding of how ﬁnancial services, and the organizations that govern them, inﬂuence economic performance. Maximizing agents have an incentive to economize on resources to facilitate transactions. A well-developed monetary system, for example, expands consumption possibilities by minimizing the costs of bartering for goods and services (that is, opportunities that households forgo to ﬁnd consumption bundles that satisfy a ‘double coincidence of wants’). Likewise, a well-developed capital market expands ﬁrms’ production possibilities by facilitating trades between ‘idea-rich’ entrepreneurs and ‘resource-rich’ ﬁnanciers. But while maximizing-behavior provides the incentive to economize on transactions costs, it is also at the heart of an impediment to economic activity; the problem of credibly committing against narrowly egoistic actions. To see this problem, consider the diﬃculty that societies experience in establishing and maintaining eﬃcient systems of money. As maximizers, monetary authorities want to expand consumption possibilities beyond those that barter makes available. They can do that by issuing money. However, for money to be used, the monetary authority must be able to commit not to inﬂate away its value. Successful monetary authorities must adopt a type of organization or governance that constrains them enough to eliminate the temptation to inﬂate. The problem of credible commitment also says a lot about how systems of ﬁnancial intermediation should be (and are) organized. Idea-rich entrepreneurs and resource-rich ﬁnanciers enjoy considerable prospects for mutually beneﬁcial trade. But entrepreneurs have an incentive to be too optimistic about the future performance of their projects and misuse borrowed capital after the fact. Likewise, narrowly egoistic ﬁnanciers have an ix

x

Foreword

incentive to expropriate the product of entrepreneurs’ sunk investments. Well-developed systems of ﬁnancial intermediation and corporate governance help maximizers credibly commit against such actions, and thus expand general economic opportunities. Money, Financial Intermediation and Governance is both rigorous in its treatment of the material and accessible. By avoiding shortcuts it challenges the reader to think seriously about deep economic incentives. At the same time, it provides the guidance necessary to master these important ideas. This is a great book for readers who want to learn and, more importantly, to think. Antoine Martin

About the book Money, Financial Intermediation and Governance uniﬁes the treatment of diverse, but deeply related, topics in ‘money and banking.’ It does so by continually building on the assumption that economic actors are ‘optimizers’ to explain how monetary and ﬁnancial services, as well as related governance mechanisms, inﬂuence economic performance (that is, levels, ﬂuctuations, and growth rates of wealth). By developing this explanation from the ground up, Money, Financial Intermediation and Governance not only lets readers make sense of today’s monetary authorities and ﬁnancial markets, it lets them see through superﬁcial complexities to fundamental inﬂuences that will shape those organizations for years to come. Mastering this analytical process is important for business, legal, and policy professionals, as well as individuals who are interested in their own ﬁnancial security. Indeed, it equips readers with an enduring ability to productively anticipate, respond to, and even shape macroeconomic and related political developments. This book’s greatest contribution may thus be to help ‘students’ at various stages of their education and careers to enjoy the lasting advantages of becoming careful ‘thinkers.’ This contribution is especially valuable for intermediate-undergraduate and professional curricula where popular, but loosely grounded, approaches can overwhelm students with a long list of apparently unrelated topics. Money, Financial Intermediation and Governance, instead, shows how simple micro-level incentives work through monetary and ﬁnancial channels to inﬂuence broader economic performance. In addition, the book’s careful treatment of business cycles makes it an attractive supplement to conventional intermediate- or masters-level courses in macroeconomics, and even a primary text for the ﬁrst course in a sequence that examines economic ﬂuctuations and then economic growth. In each case, readers should beneﬁt from illustrative ﬁgures that guide them through formal analyses, careful arguments that draw and build on respected research, and an enduring methodological approach that lets their understanding evolve with the ﬁeld’s rapid developments.

xi

About the authors Dino Falaschetti (PhD, MBA, CPA) is a Campbell National Fellow at Stanford University’s Hoover Institution, Associate Professor of Law and Economics at the Florida State University College of Law, and President of Economic Advisors, Inc. Dr Falaschetti has also served the Executive Oﬃce of the (US) President as a Senior Economist for the Council of Economic Advisers (with responsibilities for regulation and ﬁnancial services), held academic appointments at Montana State University, University of California at Berkeley, University of Tennessee and Washington University in St Louis, and fulﬁlled managerial responsibilities in corporate ﬁnance and accounting for a Fortune 100 company. He earned a PhD in economics from Washington University in St Louis (with ﬁelds in political economy, economic theory, and industrial organization), an MBA with highest honors from the University of Chicago (with concentrations in economics and ﬁnance), and a BS with distinction from Indiana University (with a major in accounting). Michael Orlando (PhD, MBA) is Principal Consultant of Economic Advisors, Inc. in Denver, Colorado. Dr Orlando served in the Economic Research Department at the Federal Reserve Bank of Kansas City, and then as Vice President and Branch Executive of the Bank’s Denver Branch. His primary responsibilities included regional and energy markets research, policy advising and public communication. Dr Orlando also served as Vice President for Research and Product Development at evolve24, LLC, a business analytics firm specializing in development and validation of algorithms used to produce quantitative measures of media content. He is also a practiced teacher, petroleum engineer and environmental engineer. Dr Orlando’s research spans a range of topics in applied microeconomics. He has published work on the economics of payments networks, corporate governance, the geography and industrial demography of innovation, energy and environmental policy. Dr Orlando earned a PhD in economics from Washington University in St Louis (with ﬁelds in industrial organization and macroeconomics), an MBA from Tulane University, and a BS with high distinction from The Pennsylvania State University (with a major in petroleum and natural gas engineering).

xii

Introduction Our objective for the present book is to develop a uniﬁed analysis of topics relating to ‘money and banking.’ This approach is attractive, both pedagogically and scientiﬁcally, but is largely missing from popular instructional materials. We ﬁll this gap by continually grounding our insights on ﬁrst principles. In particular, we gradually build on the assumption that economic actors are ‘maximizers’ (that is, households maximize utility and ﬁrms maximize proﬁt) to understand how monetary and ﬁnancial services, and the politics that govern their production, inﬂuence economic performance. Successful students will enjoy a lasting method with which to address important questions like the following: ●

●

●

●

● ●

●

What is money and how does its supply relate to a) an economy’s consumption possibilities and b) ﬂuctuations in economic performance (that is, business cycles)? Why have only history’s most recent societies enjoyed the beneﬁts that ‘ﬁat’ monetary systems can produce (that is, currency systems that lack a commodity backing)? How can the governance of monetary authorities (for example, the United States’ Federal Reserve System) facilitate this enjoyment, and why do so many economies fall short on this dimension? How does the organization of intermediaries (that is, proximate suppliers of loanable funds) and governance of corporations (that is, proximate demanders of loanable funds) inﬂuence ﬁnancial market eﬃciency? Is market discipline suﬃcient for ﬁnancial market eﬃciency, or is public regulation necessary? Why is ﬁnancial market eﬃciency so important for economic opportunity in general – that is, why is ﬁnance important for ‘Main Street’ as well as ‘Wall Street’? Why, despite this importance, do societies so often discourage ﬁnancial development and thus forgo superior economic performance?

In lieu of oﬀering a superﬁcial treatment of these questions, we develop an analytical method for building ﬁrmly grounded and internally consistent explanations. This approach improves upon compartmentalized analyses 1

2

Money, financial intermediation and governance

Part I Analytical foundation Ch. 1–3 Banking Part V

Money Parts II–IV

Information is asymmetric

Barter must satisfy a double coincidence of wants

Ch. 13

Maximizers can do better by economizing on the cost of transacting across time

Maximizers can do better by economizing on the cost of spot market transactions

Ch. 15

Ch. 14 Financial intermediation

Money Ch. 4–5

Corporate governance

Ch. 6–9

Consumption possibilities

Business cycles

Ch. 10–12

Ch. 16 Economic performance and the governance of monetary and financial services

Figure I.1

Organization of the book

that can overwhelm students with long lists of apparently unrelated rationalizations, and leave them ill-prepared to address germane issues that regularly arise in personal and professional settings but fall outside the scope of what their courses can explicitly consider. Indeed, it encourages students to advance themselves into thinkers by focusing on the ability ‘to understand and not merely to remember’ (Hilton, 1997).1 Figure I.1 illustrates how our book works toward this objective. Part I begins by developing a set of analytical tools with which to ﬁrmly ground insights to money and banking in particular and (perhaps more importantly) human sociality in general. Our goal here is for readers to become comfortable with (a) the general process of reducing complex problems into simple ones so that those problems become tractable, but not irrelevant (that is, modeling) and (b) a particularly powerful method

Introduction

3

of reduction – that is, assuming that units of observation (for example, households and ﬁrms) are ‘egoistic’ (self-interested) and logically deducing refutable implications therein. Opening a study of money and banking in this manner creates at least two beneﬁts. First, by making our analytical method transparent, it helps readers see how much conﬁdence can be placed in what they ‘know.’ Second, it motivates why ‘money and banking’ topics are more than academic curiosities – they fundamentally inﬂuence economic performance. Part I (Chapters 1–3) shines a bright light on this inﬂuence by making transparent an assumption that a (prerequisite) principles-level training can leave opaque – that is, the cost of transacting is negligible. In doing so, it not only strengthens our ability to ‘think like a social scientist,’ it also builds a bridge from students’ incumbent training to a set of phenomena that frequently falls under the ‘money and banking’ heading. Indeed, once readers confront the truism that transacting is costly, they begin to appreciate the capacity for monetary and ﬁnancial governance to aﬀect a society’s economic opportunities. Our introductory model of the household’s problem in Part I implicitly assumes that spot-trades (those that occur at a point in time) take place in a well-developed monetary system. Absent such a system, however, households ﬁnd themselves maximizing utility via ‘barter’ (a system where individuals trade only when they share a ‘double coincidence of wants’). There, the relatively high resource cost of consummating trades precludes superior (unanimously preferred) outcomes. This departure from a ‘ﬁrst best’ outcome motivates our investigation in Part II (Chapters 4–5) of how money, by facilitating spot-trades when a double coincidence of wants is absent, can improve the economic wellbeing of all households. We’ll learn that ‘ﬁat’ money (exchange media that lack a commodity backing) can improve economic performance, but societies nevertheless face persistent obstacles to developing and maintaining eﬃcient monetary systems. In particular, ‘maximizers’ not only want to economize on transactions costs, they also want to opportunistically substitute ‘inﬂation taxes’ for ‘real taxes.’ Unless monetary authorities can credibly commit against such opportunism, individuals will lack the conﬁdence that is necessary for ﬁat systems to economize on transactions costs. In Part III (Chapters 6–9), we synthesize and extend contributions to an important ﬁeld of macroeconomics – that is, ‘money and economic ﬂuctuations.’ We do so by extending Part II’s investigation from a static to dynamic setting, and learn that whether money should be actively managed is also sensitive to the formation of expectations. To the extent that maximizing-behavior inﬂuences how individuals form expectations, activist

4

Money, financial intermediation and governance

monetary policy can encounter diﬃculty when attempting to dampen economic ﬂuctuations.2 But while an activist policy appears ineﬀective in this regard, political agents may nevertheless face a strong incentive against remaining passive. Indeed, after individuals form expectations about monetary policy, a political agent’s ‘best’ action becomes to expand the money supply – that is, as policy interactions unfold, individuals’ incentives become inconsistent with society’s optimal plan. Part III thus complements Part II by further motivating the importance of governing monetary institutions so that otherwise egoistic political agents advance aggregate economic performance. We’ll address this dimension of governance in Part IV (Chapters 10–12) by examining how the organization of central banking can inﬂuence economic performance. Taken together, Parts I through IV (Chapters 1–12) develop, from the ground up, an understanding of how the production of monetary services should be and is organized. But while this understanding is ‘deep,’ it is not ‘wide.’ To be sure, it only illuminates how societies should (and do) organize spot market transactions – that is, trades at a particular instant in time. It leaves open, however, the related question of how maximizers can ease transacting across time. We treat this issue in Part V (Chapters 13–16) by rationalizing the existence of and prescribing organizational strategies for ﬁnancial intermediation and corporate governance. Rather than being a separate investigation, this one also builds from Part I’s methodological foundation. In particular, we maintain consistency by again highlighting the capacity for transactions costs to discourage egoists from achieving their goals, and rationalizing observed features of ﬁnancial markets (e.g., the organization of intermediaries and corporate governance systems) as emerging from incentives to economize on these costs. Consider, for example, the goal of proﬁt maximization. In Part I’s introductory model, ﬁrms operate in a ‘frictionless’ economy, and thus freely choose inputs to equate each factor’s marginal product and opportunity cost. But what happens when ﬁrms operate in a world that more closely portrays our empirical reality? In particular, what happens when a ﬁrm’s opportunity to employ inputs conﬂicts with its access to requisite resources? In this richer setting, we ﬁnd that the domain of inputs shrinks from what is possible for Part I’s ﬁrms – that is, ﬁrms become ‘cash constrained.’ Consequently, just as households in Part II cannot achieve the level of utility that is available when bartering is costless, ﬁrms in Part V cannot achieve the proﬁts that are possible when accessing ﬁnancial capital is costless. But just as Part II’s households need not rest at an inferior barter outcome, Part V’s ﬁrms need not rest at an inferior cash-constrained outcome. To see why, we again exploit our assumption that economic actors are maximizers. For example, recognizing that increased proﬁts are

Introduction

5

available except for cash constraints, an egoistic supplier of loanable funds might lend ﬁnancial capital to expand a ﬁrm’s set of feasible inputs. In this manner, successful intermediation can make everyone better oﬀ – that is, ﬁrms can pursue productive projects and lenders can share in consequent proﬁts. The capacity to realize superior payoﬀs rests, however, on whether suppliers of loanable funds can discourage demanders from opportunistically employing those funds. To be sure, note that maximizers not only want to economize on transactions costs, they also want to misrepresent the quality of assets being traded and strategically exploit bargaining positions (the strength of which can vary across a transaction’s history). Left unchecked, the prospect of such unproductive actions will constrain a ﬁrm’s opportunity to leverage capital-access into increased proﬁts. We’ll thus learn that, for egoistic suppliers of capital to trade with egoistic demanders, economic actors must credibly commit against narrowly pursuing their own self-interests!3 This understanding will help us rationalize why, say, loanable funds tend to be traded via intermediaries such as banks rather than more directly. It will also help us see why ﬁrms adopt restrictive organizational structures (e.g., hand-tying capital structures) or employ external monitors (e.g., ﬁnancial statement auditors). Finally, it will oﬀer insight to why ﬁnancial governance is so important for economic well-being, and why productive governance nevertheless encounters considerable political resistance. In particular, we’ll see that educating individuals about the merits of ﬁnancial development may not be enough to encourage productive reforms (or discourage unproductive policies). Indeed, unless development policies address distributional constraints, even the most (technically) productive advances may go unrealized, and those that already exist will face considerable political risks.

NOTES 1. ‘Adaptability to change is itself a hallmark of successful education, and it is change, not any speciﬁc technology, that most aptly characterizes life today and in the foreseeable future. A genuine education enables one to acquire, for oneself, the skills one happens, at a given stage of one’s life, to need. A training, on its own, contributes almost nothing to education and produces distressingly ephemeral advantages’ (Hilton, 1997). 2. We leave until Chapter 14 the consideration of how credit market frictions can channel monetary policy into real economic activity. 3. Miller (1991), Falaschetti and Miller (2001) and Skaperdas (2003), among others, examine such dilemmas.

PART I

The economic method of inquiry

We develop in Part I a ﬁrmly grounded and internally consistent framework for investigating phenomena associated with human sociality in general, and money and banking in particular. This framework, frequently referred to as ‘the economic way of thinking,’ can appear so abstract as to hold little promise for addressing substantive interests. We’ll see, however, that the capacity to facilitate careful abstraction lets us employ the economic way of thinking to scientiﬁcally examine any aspect of human sociality. Indeed, after developing our method of ‘logical deduction from self-evident truths,’ we’ll consider its profound implications for what can otherwise appear to be trivial policy issues (for example, the extent to which airbags enhance automobile-safety). We’ll thus begin to appreciate how a ﬁrmly grounded, internally consistent, analytical method can oﬀer empirically interesting insights that may otherwise remain hidden. Our goal for the present set of chapters is to persuade readers that thinking like a social scientist can constructively address, in a simple and uniﬁed manner, important problems that often appear overwhelmingly complex and unrelated. David Friedman (1996) describes this capacity as follows: To most people, economics is a dull science full of statistics and jargon, mainly concerned with money and designed to answer a narrow (but important) set of questions. To economists, economics is a powerful tool for understanding why armies run away, voters are ignorant, and divorce rates rise, as well as solving practical problems such as how not to get mugged. Its theme is not money but reason – the implications, especially the non-obvious implications, of the fact that humans act rationally. (Italics added)

8

The economic method of inquiry

Mastering this method early in a course on money and banking is important. Rather than leaving students with an ephemeral catalog of disjoint ideas, it provides an enduring analytical tool with which to build, from the ground up, an understanding of how money, ﬁnancial intermediation and ultimately governance inﬂuence economic performance.

1.

Modeling

Models do not represent literal truth. Rather, they are tools for viewing the world, its problems, and contexts. (Poirier, 1995: 2)

WHAT ARE MODELS, AND WHY ARE THEY NECESSARY? To better understand phenomena associated with money and banking, we will make extensive use of economic ‘models.’ Models are simple characterizations of forces that inﬂuence our empirical reality, and are necessary to examine diﬃcult to track issues in a logically consistent manner. Every problem that individuals confront is a simpliﬁed version of ‘reality.’ In the extreme, inherent physical limitations preclude us from immediately sensing many aspects of what is ‘true,’ and thus from ever acting on a problem without ﬁrst simplifying it. We rely on models long before bumping into physical limitations, however. Consider, for example, the last time you used a road map.1 Was your map a ‘simpliﬁed representation’ of an empirical reality? Of course it was – indeed, a ‘real’ map would have never ﬁt in your car! But notice that, rather than clouding the issue under consideration (for example, directions from one town to another), this simpliﬁcation made it more transparent – that is, it let you see past relatively unimportant issues to carefully contemplate questions of particular interest (such as how far, and in what direction, is your destination?). The question before us is thus not whether to simplify, but rather how to simplify problems so that they become tractable (without becoming trivial). The simpliﬁcation on which we’ll progressively build is that economic actors engage in ‘maximizing-behavior’ (that is, households maximize utility and ﬁrms maximize proﬁts). In short, we’ll assume that actors are ‘hard wired’ to selﬁshly pursue these rather simple goals – they are egoists.2 Because this assumption can appear coarse, and because it will act as the foundation from which we build our understanding of money and banking, it is important for us to see why treating it as axiomatic is reasonable. We thus turn in the present chapter’s remainder to the issue of what makes for 9

10

The economic method of inquiry

a ‘good’ model. We’ll then evaluate in Chapter 2 whether a model that is grounded on the axiom of maximizing-behavior can be a good one. Our objective in developing this preliminary material is to be transparent about how our methodology generates insights to substantively interesting topics. In particular, we want to highlight what has to be true for our conclusions to be valid, and will argue that the satisfaction of this necessary condition is self-evident (that is, rationalizing our foundation does not require another model).

WHAT IS A ‘GOOD’ MODEL? Just because models are necessary for understanding money and banking (and human sociality more generally) does not mean that any model will do. To eﬀectively pursue our inquiry, we not only need a model, we need a ‘good’ one. Good models build from a solid foundation to generate insightful predictions. Models have to start from somewhere, and our starting point will be the set of statements that a model takes as given. We’ll refer to these statements as a model’s ‘assumptions,’ and claim that they must be ‘selfevident’ if a model is to be ﬁrmly grounded. Indeed, if reﬂection alone does not establish these statements’ truths, then we would need another model to establish our assumptions.3 A solid foundation precludes this type of regression. It also makes transparent the answer to an important question that more loosely grounded approaches tend to hide – that is, ‘what has to be true for a model’s predictions to be valid?’ In this light, a model whose assumptions are self-evident appears preferable to one whose assumptions are not. In addition, holding other considerations constant, a model that predicts well lets us rationalize our empirical reality, and thus appears preferable to one that predicts poorly. We’ll thus prefer models that generate non-trivial predictions from an agreeable set of assumptions to those that predict less well from superﬁcial assumptions. Depending on the objective at hand, however, we may encounter models that generate superior predictions while maintaining inferior assumptions (or vice-versa). Before moving forward, we should consider how to constructively evaluate models in this less obvious case. In short, we will prefer models that explain ‘a lot’ with ‘a little.’ The late Nobel Prize-winning economist, Milton Friedman (1953), carefully developed this notion to establish a criterion for whether a model should be accepted as a part of the ‘body of systematized knowledge concerning what is.’ An acceptable model, according to this standard, is one that yields valid

Modeling

11

and meaningful (that is, more than truistic) predictions about phenomena not yet observed.4 Notice that, while a model must rest on a solid foundation to achieve this status, it cannot maintain an exhaustive set of assumptions. In other words, models necessarily assume that certain forces are, and others are not, important for understanding a particular class of phenomena. Whether our model is suﬃciently complete thus becomes an interesting question, and can be addressed by evaluating the costs of expanding our assumptions (for example, a reduction in the conﬁdence that our assumptions are selfevident) against the beneﬁts of developing a ﬁner theory (for example, an increase in the robustness of predictions).

EVIDENCE AND THE QUALITY OF A MODEL’S PREDICTIONS [We need to test models] as useful imitations of reality by subjecting them to shocks for which we are fairly certain how actual economies or parts of economies would react. The more dimensions on which the model mimics the answers actual economies give to simple questions, the more we trust its answers to harder question. (Lucas, 1980)

A ‘good’ model is one that rests on a solid foundation and generates insightful predictions. We’ll argue in Chapter 2 that the assumption of maximizing-behavior is self-evident, and can thus ﬁrmly ground our models of human sociality in general and money and banking in particular. We’ll also present evidence that models that are grounded as such generate non-trivial, but accurate, predictions. Before turning in that direction, then, we should understand what counts as solid evidence. Our investigation will generate two types of evidence: reduced form and structural.5 In short, reduced form evidence looks at the ultimate relationship between variables of interest, and can thus provide insight to how large is a variable’s gross impact. Structural evidence, on the other hand, looks more carefully at the channels through which variables of interest relate, and can thus oﬀer information about a variable’s partial impact. Reduced Form Evidence Each type of evidence maintains strengths and weaknesses. To see these qualities, consider the hypothesis that drinking red wine decreases the incidence of heart disease. Under this hypothesis, red wine and heart disease share a negative relationship, the reduced form of which can be illustrated as in Figure 1.1.

12

The economic method of inquiry

Red wine ↑ → Heart disease ↓ Figure 1.1

Reduced form relationship between red wine and heart disease

To evaluate whether this relationship is an empirically important one, we can look at the correlation between drinking red wine and contracting heart disease. However, even if this correlation is strong and negative (that is, even if individuals who drink red wine experience considerably less heart disease), we should be careful about the conﬁdence that we place in our model. This concern arises from the fact that correlation does not imply causation.6 To understand this concern, suppose that the incidence of drinking red wine varies with income. Suppose also that income inﬂuences the incidence of heart disease through, say, its eﬀect on access to preventive health services. In this case, reduced form evidence for our hypothesis would be spurious – that is, red wine would be correlated with something that aﬀects heart disease, but wouldn’t aﬀect heart disease per se. Figure 1.2 illustrates how this type of inferential mistake can emerge when reduced form models omit variables (for example, income) that relate to both the hypothesized causal variables (red wine) and outcome measures (heart disease). Income ↑

Red wine ↑ Figure 1.2

Heart disease ↓

Omitted variables bias

In addition to suﬀering from ‘omitted variables bias,’ reduced form evidence may leave open the hypothesis that causation runs in the opposite direction. Suppose, for example, that heart disease diminishes human capital, and thus one’s earnings capacity. Suppose also that wine is a ‘normal’ good – that is, its demand increases with wealth. Then a lack of heart disease could cause relatively high levels of earnings, which in turn could increase wine consumption. In this case, heart disease and wine consumption share a negative reduced form relationship, but the incidence of heart disease inﬂuences wine consumption.

Modeling

13

Structural Evidence The above illustrations suggest that, to the extent that our hypothesized causal variables are not experimentally controlled, we’ll need to be careful about inference drawn from reduced form evidence. Given this suggestion, we may want to evaluate our hypotheses via ‘structural evidence’ instead. Rather than looking at the direct relationship between a hypothesized causal variable and outcome measure, structural evidence looks at the channels through which a causal variable might exert its inﬂuence. Continuing with our wine example, grapeskins (which give red wine its color) appear to be a rich source of the chemical ‘resveratrol.’ Resveratrol, in turn, appears to reduce the incidence of genetic mutations and thus curb age-related disorders such as heart disease.7 Finding a chain of evidence as illustrated in Figure 1.3 can thus increase our conﬁdence in the hypothesis that red wine inﬂuences heart disease – that is, omitted variables and reverse causation are not clouding our inference. Red wine ↑ → Resveratrol ↑ → Genetic mutations ↓ → Heart disease ↓ Figure 1.3

Structural evidence

Given this apparent superiority of structural evidence, why should reduced form evidence ever deserve our attention? It does, and for the same reason that it creates diﬃculty – that is, it does not rely upon a particular structure for its validity. Numerous channels exist through which red wine might aﬀect heart disease. Hence, even if red wine truly inﬂuences the potential for heart disease, investigators might conclude that it does not if they limit their analysis to an incorrect channel. In other words, unless we identify the correct process (or a process close to what is correct) through which our hypothesized variable aﬀects an outcome of interest, structural evidence could suggest that our variables are unrelated (even if they strongly relate through another channel!). Reduced form evidence, on the other hand, would disclose this relationship since it leaves open the channel of causation. As long as at least one such channel exists, reduced form evidence can support the hypothesis.

CONCLUSION To make any progress in understanding non-trivial social phenomena, we must reduce those phenomena into tractable models. But to produce valid insights, this ‘reduction’ must ultimately rest on solid ground – that is, a set of self-evident assumptions.

14

The economic method of inquiry

Logically deducing hypotheses from such assumptions maintains the attractive potential to ‘explain a lot with a little.’ To evaluate this explanatory power, we’ll compare our models’ observable implications to either reduced form or structural evidence. Relying on reduced form evidence will mitigate the risk of dismissing good models, but also increase the risk of accepting bad ones. Structural evidence exhibits the opposite merits and diﬃculties. Like many of the conclusions that we’ll draw, this one does not oﬀer a neat set of recipes with which to tackle the ‘real world.’8 Rather, it carefully sets out the trade-oﬀs that we must make to move forward with our understanding of money and banking, and the margins on which those trade-oﬀs can logically take place.9

NOTES 1. We thank Rob Fleck for sharing this illustration. 2. Note that, even if this assumption is a good one for modeling purposes, it does not condone or condemn associated behavior. We will address this issue more carefully when we look at how models facilitate ‘positive’ insights (that is, an understanding of ‘what is’) versus ‘normative’ insights (that is, an understanding of ‘what should be’). 3. Philosophers have long recognized this necessity, perhaps most famously René Descartes (1596–1650) in his ‘Cogito, ergo sum’ foundation to his Discourse on Method. 4. Professor Friedman’s 1967 prediction about the instability of the Phillips Curve exempliﬁes this feature of a ‘good’ theory (see for example, Woodford, 1999: 16). We’ll consider this example more carefully in Part III. 5. Mishkin (2004) also employs this dichotomy. 6. This concern led Rajan and Zingales (2003: 109) to argue that ‘correlation is the basis for superstition, while causality is the basis for science.’ 7. Evidence to this eﬀect appears in the journal, Nature (August 2003). 8. And for good reason – such a set is unlikely to exist! 9. In particular, it does not say that economic models can rationalize any imaginable outcome – that is, predictions from good economic models are not vacuous.

2.

Foundation

The axioms are like the foundations of a building. No matter how carefully the [social scientist] constructs the walls and the rest of the structure, if the foundations are unsound, the entire structure may collapse. One false axiom, and everything that follows may be wrong or meaningless. (Keith Devlin, 1998)

INTRODUCTION We argued in our ﬁrst chapter that, to understand complex social phenomena, we must logically reduce those phenomena into models, and those models should build on a parsimonious set of ‘self-evident’ truths. In doing so, however, we must not make our simpliﬁcations simplistic – that is, our assumptions must aid tractability without trivializing the question of interest. We push forward in the present chapter by arguing that building models on the assumption of ‘maximizing-behavior’ fulﬁlls this requirement. Although (and perhaps because) it is extremely simple, this assumption can facilitate deep insight to human sociality in general, and thus phenomena associated with money and banking in particular. Absent such an assumption, and a consistent logic for deducing insights from it, understanding what truly motivates these phenomena would be diﬃcult, if not impossible.

AN AXIOMATIC APPROACH The heart of the economic approach, according to Gary Becker (1976b), is its relentless exploitation of the assumptions that agents optimally choose actions subject to a set of stable preferences. So deﬁned, this approach should appear familiar to those with an intermediate training in mathematics – that is, social scientists employ the ‘economic approach’ in much the same way that mathematicians employ axiomatic approaches. In particular, social scientists frequently take Becker’s postulates (i.e., agents maximize according to stable preferences) as given. They then develop insight to how systems behave by logically deriving observable implications therein. Note well that, if our derivation of such implications is valid, then the only part of our approach open to criticism is our choice of axioms. Once 15

16

The economic method of inquiry

we accept this choice, any implications that logically derive from it must be true for social systems that our assumptions characterize. In other words, any system that obeys these rules necessarily exhibits all of the properties that we can deduce from them.1 This approach will thus advance our understanding of social phenomena to the extent that our axioms meaningfully simplify forces that act on our empirical reality. Before going forward, then, we should be clear about what we are assuming and how we plan to exploit these assumptions.

THE AXIOM OF MAXIMIZING-BEHAVIOR Never assume the animal you are studying is as stupid as the one studying it. (Trivers, 2002)

The axiom about which we’ll be particularly careful is one that frequently appears disagreeable to non-economists – that is, the axiom of ‘maximizingbehavior.’ Perhaps the most important argument to understand from the present chapter is that this assumption not only facilitates our construction of a simple analytical framework, it is also more obvious than superﬁcial assessments suggest. The ‘As If’ Principle, in Principle David Friedman (1996) oﬀers several rationalizations for our assumption that egoistic motives drive economic actors: One reason to assume rationality [that is, maximizing-behavior] is that it predicts behavior better than any alternative assumption. Another is that, when predicting a market or a mob, what matters is not the behavior of a single individual but the summed behavior of many. If irrational behavior is random, its eﬀects may average out . . . A third reason is that we are often dealing not with a random set of people but with people selected for the particular role they are playing. (Italics added)

It is this third rationalization that especially deserves our attention. Social scientists ground their investigations on the axiom of maximizing-behavior, not because they see such behavior as being purposeful, but rather because forces that inﬂuenced their subjects’ selection left those who act as if they are maximizers. Note that this axiom embodies the biological principle of natural selection. To be sure, it implicitly treats the social system as an adoptive mechanism that chooses exploratory actions that were most successful in maximizing utility or proﬁts. Our inquiry does not ground itself on the

Foundation

17

assumption that actors explicitly adapt to their environment (that is, explicitly maximize utility or proﬁts), but instead on the assumption that relevant systems adopted for our observation agents who recognized success in this regard (Alchian, 1950). The actors that we observe can thus be usefully modeled as rational, even if they do not explicitly maximize.2 An Illustration of the ‘As If’ Principle Alchian (1950) nicely illustrates the ‘as if’ principle by considering what would happen if thousands of travelers randomly selected roads out of Chicago, but gasoline stations existed on only one of those roads. Given these conditions, only individuals who happened upon the road where gasoline is available can successfully complete their trips. In other words, only those who chose a road as if they were maximizers would make themselves ‘observable’ to an external analyst (for example, a social scientist). Moreover, even if subjects randomly choose roads, the ‘as if’ principle accurately predicts who would succeed if the road on which gasoline was available changed. In this case, a diﬀerent direction of travel would become optimal. But rather than adapting to this change, newly successful individuals would have been adopted by the changed environment. To be sure, what is optimal in the new environment will diﬀer from what is optimal in the old, but only those that act as if they optimally responded are available for observation (Alchian, 1950). The axiom of maximizing-behavior sometimes receives criticism because households or ﬁrms do not appear explicitly rational. After all, when was the last time that you formally measured marginal costs and beneﬁts before making a decision? The trouble that we have in answering this question suggests that explicit rationality is not self-evident. Alchian’s illustration instead suggests that implicit rationality is self-evident. It is this latter characterization of rationality that can be treated as axiomatic, and acts as the foundation on which our models will build.

CAN OUR AXIOMATIC APPROACH PREDICT WELL? In Chapter 1, we argued that good models build insightful predictions from self-evident truths. In this light, and coupled with an understanding of the ‘as if’ principle, models that build on the axiom of maximizing-behavior appear capable of being ‘good.’ To complete our evaluation of whether such models can indeed help us understand human sociality, then, we must consider their potential to rationalize important features of our empirical reality.

18

The economic method of inquiry

Evidence from (Non-human) Animals The ‘as if’ principle’s power is, perhaps, most clearly evident in the context of (non-human) animals. Consider, for example, a problem that likely confronts moose – that is, ‘choose’ a combination of aquatic and terrestrial plants that maximizes energy. Absent constraints on its behavior, our moose could arbitrarily choose its diet. However, just as households face fundamental constraints when choosing consumption bundles (for example, bounded budgets), so does our moose.3 First, our moose must choose a combination of plants that produces a minimal level of sodium. Indeed, just as travelers who do not choose roads on which gasoline is available become ‘extinct,’ so do moose that ignore this constraint. Second, our moose cannot accumulate energy without bound. In particular, its rumen (the ﬁrst compartment of its stomach) limits the volume of plants that it can feasibly consume. Rather than facing an unconstrained maximization problem, our moose thus faces a constrained problem – that is, when attempting to maximize energy, it can choose combinations of aquatic and terrestrial plants only from the shaded area in Figure 2.1. If our moose addresses this problem as would a ‘maximizer,’ then it will choose the combination of terrestrial and aquatic plants represented here as (x, y). To see why a maximizing-moose would make such a choice, notice that the moose’s ‘iso-energy curve’ is analogous to the ﬁrm’s iso-proﬁt curve or consumer’s indiﬀerence curve – that is, any combination of aquatic and Aquatic plants Iso-energy curves

Sodium constraint

y

Rumen constraint x Figure 2.1

The moose’s energy mazimization problem

Terrestrial plants

Foundation

19

terrestrial plants that lies on a particular curve produces an equivalent level of energy. The combination of terrestrial and aquatic plants that lies on the highest such curve, subject to the sodium and rumen constraints, is (x, y) – that is, this combination produces the maximum feasible level of energy. But why would our moose choose this particular diet? After all, if we are skeptical about humans’ cognitive capacity to explicitly optimize, then shouldn’t we be more skeptical about the moose’s capacity? Not when the ‘as if’ principle is empirically relevant. In this case, competitive environments will adopt only those moose who choose diets as if they maximize energy (subject to sodium and rumen constraints). Clearly, the root of causation here is not the moose’s capacity to think formally. Rather, it is the hypothesis that energy inﬂuences a moose’s capacity to replicate, and behavioral traits are heritable. Under these conditions, moose who make themselves available for our observation will tend to behave according to the axiom of maximizing-behavior. We’ve certainly succeeded in developing a model of how moose choose their diets. But is this a ‘good’ model? Apparently, yes! Rather than being a cute but empirically irrelevant illustration, our formal analysis of the moose’s diet heavily draws on biological research (see Krebs and Davies, 1993). Evidence from this research is consistent with moose choosing consumption bundles in exactly the manner that we set out here.4 Evidence from Humans We now see that, to advance our understanding of interesting social phenomena (including those associated with money and banking), we must reduce those phenomena into a tractable analytical framework – that is, we must employ models. The manner in which we will make this reduction, and the manner in which social scientists tend to make this reduction more generally, is to assume that the subjects of investigation (for example, households, ﬁrms) engage in maximizing-behavior. And while making this assumption appears to ask a lot from our subjects, we should recognize the potential for selection-forces to have favored those who behave ‘as if’ they are maximizers. But to what extent have contemporary humans maintained this potential? It turns out that, time and again, even societies’ most unsuspecting members appear to act ‘as if’ they are maximizers. Consider, for example, the following questions: ● ● ●

Does raising the speed limit decrease safety? Does driving a light truck make you safer? Do airbags increase safety?

20

The economic method of inquiry

Each of these questions appears rather straightforward. However, raising the speed limit need not decrease safety if doing so encourages individuals to drive on safer roads (for example, divided highways).5 Driving light trucks might increase the risk of injury (or fatality) if it accommodates more dangerous driving habits.6 Airbags, too, can similarly induce ‘unintended consequences.’ A bit of introspection makes this diﬃculty clear – how would your driving habits change if your airbag deployed razor-sharp spikes? While not immediately obvious to those unfamiliar with the economic way of thinking, each of these consequences logically follows from the axiom of maximizing-behavior. Moreover, just as with our model of the moose’s behavior, received evidence is consistent with these implications being empirical regularities rather than theoretical curiosities. Our axiom is not as abstract as it might ﬁrst appear, and the insights that it generates can be life-saving! Evidence Against Maximizing-behavior? The logic and evidence developed so far suggests that models based on the axiom of maximizing-behavior can be ‘good.’ That said, we frequently observe phenomena that appear diﬃcult to rationalize within this context. Consider, for example, the case of radical religious militias – how can modeling individuals as ‘maximizers’ shed light on this seemingly irrational behavior? Eli Berman (2003) put this question as follows: The Taliban, Hamas, and other radical Islamic groups present a challenge to behavioral scientists who assume rational choice in individual decisions. How do we explain destructive acts from which individuals derive no direct beneﬁt such as subjugation and institutional abuse of women, minorities and homosexuals, closing of schools and hospitals, enforcement of general strikes, abuse and murder of prisoners, or desecration of holy sites? Why do radical religious sects so often turn to militia activity, and why are their militias so eﬀective?

Addressing these questions is important if we are to have conﬁdence in our analytical method. Indeed, if ‘maximizing models’ miss something fundamental about a particular social phenomenon, then we should question its capacity to rationalize other phenomena about which we are less certain. In this case, before applying the economic way of thinking to any particular question, we would have to know whether that question falls within our method’s scope. Another model would then be necessary to determine when our economic way of thinking can produce valid insights, and when it cannot.

Foundation

21

Evaluated in this light, Berman’s (2003) work appears important not only because it searches for deep social forces at which policies can aim to check radical militias, but also because it increases our conﬁdence that the economic way of thinking is not compartmentalized. Berman (2003) succeeds by realizing that maximizers encounter diﬃculty when attempting to make promises. To address this diﬃculty, maximizers may play actions that superﬁcially appear to decrease utility, but ultimately make feasible the expanded set of consumption possibilities that credible individuals can enjoy.7 To see this idea more clearly, consider a set of maximizers who are attempting to produce a ‘local public good’ – for example, a protective wall around a village. Notice that, should any one individual from this set curb his or her eﬀort, others are likely to be able to construct the wall themselves. In other words, no single individual is pivotal in building the wall, so an individual ‘shirker’ may enjoy the wall’s security without forgoing the utility that he or she could have enjoyed from doing something other than contributing to the wall’s construction. But if each individual is a maximizer, then everyone may act in this manner – that is, no one will contribute to the public good’s production. So how might maximizers credibly commit to such production processes? One way is to decrease the opportunity cost of contributing. In the case of radical religious groups, this decrease can come from attempts to ‘limit choices (prohibitions) and to destroy resources and options (sacriﬁces)’ (Berman, 2003). Here, the utility that members might enjoy from allocating eﬀort to market activities represents the opportunity cost of contributing to the public good. To the extent that prohibitions and sacriﬁces diminish the attractiveness of such alternatives, members will contribute more eﬀort to the public good – that is, when commitments are important, maximizers can do better by credibly constraining themselves.8 In this manner, Berman (2003) (and contributors to related literatures – for example, Laurence Iannaccone) oﬀer a ﬁrmly grounded rationale for why apparently cruel and meaningless destruction can persist. As such, they not only increase conﬁdence in the scope of our analytical framework, they show how that framework can expose the roots of pathologies that might otherwise remain hidden.

POSITIVE VERSUS NORMATIVE ANALYSIS We now see that our economic way of thinking can be a powerful method for carefully rationalizing any social phenomenon of interest. But does the

22

The economic method of inquiry

fact that we can rationalize a particular phenomenon mean that it is socially desirable? Absolutely not, and it is important that we understand why. Social scientists are tempted to address two types of questions: ‘what is’ and ‘what should be?9 These classes of questions motivate, respectively, ‘positive’ and ‘normative’ modes of analysis. In this context, economic models can be viewed as logical rationalizations of social phenomena or as bases for policy decisions.10 But individuals (even economists) frequently ignore this distinction. Robert Fogel’s (1993) Nobel Prize-winning research on slavery, for example, largely consists of positive analysis (based on the axiom of maximizing-behavior). Nevertheless, popular media widely (and inaccurately) characterized it as an apology for an institution that lacks normative merit. Likewise, pundits frequently dismiss the capacity for egoists to produce socially desirable outcomes. As Baye (2000) observes, however, this capacity is very real:11 [P]roﬁts signal the owners of resources where the resources are most highly valued by society. By moving scarce resources toward the production of goods most valued by society, the total welfare of society is improved. As Adam Smith ﬁrst noted, this phenomenon is due not to benevolence on the part of the ﬁrms’ managers but to the self-interested goal of maximizing the ﬁrms’ proﬁts.

Not only can desirable outcomes emerge from individually egoistic behavior, however, so can unattractive phenomena. This disparity emerges from selection-forces acting at the level of individuals and normative inferences being drawn at the level of groups. When individual-behavior creates external eﬀects, adopted individuals (that is, those that have been selected for observation) will tend to induce inferior equilibria. The selection-based arguments that we developed above thus do not condone associated behavior. Rather, and quite simply, they attempt to identify deep forces that promote it.

ADVANTAGES TO THINKING LIKE AN ECONOMIST The economic approach provides a uniﬁed framework for investigating all human behavior. The heart of Becker’s (1976b) argument, for example, is that human behavior is not compartmentalized, sometimes based on optimizing, sometimes not, sometimes motivated by stable preferences, sometimes by volatile ones, sometimes resulting in an optimal accumulation of information, sometimes not. Rather, all human (and even non-human) behavior can be constructively modeled as emerging from individuals that maximize utility from a stable set of preferences.12

Foundation

23

Consequently, the application of this approach does not restrict itself to the market sector. The economic approach predicts the same kind of response to opportunity costs, whether they are measured by money prices in economic markets or imputed prices in political markets. In addition to oﬀering a uniﬁed framework for understanding what might otherwise appear to be unrelated phenomena, the economic way of thinking also encourages us to carefully pursue that understanding. Notice, for example, that it will require us to be explicit about our assumptions, and the manner in which we deduce subsequent insights. By clearly linking our conclusions to the assumptions on which they ultimately rest, we will thus establish an ‘audit trail’ for distinguishing unsupported conjectures or claims from logically derived propositions (Besanko et al., 2000: 3). To sum up, once we understand that we must employ models to aid our understanding of human sociality, we need to settle on a foundation that is ‘self-evident.’ Indeed, if we took a less careful approach, nothing would stop us from trivializing important questions by essentially assuming their answers. By thinking in a scientiﬁc manner, on the other hand, we hold out the prospect of gaining valid insight to what can otherwise be intractable problems.

NOTES 1. 2.

Kurt Gödel’s celebrated ‘incompleteness theorem,’ nevertheless, shows that properties may truly exist that our consistent system is unable to illuminate. Friedman (1953) summarizes this argument as follows: Let the apparent immediate determinant of business behavior be anything at all – habitual reaction, random chance or whatnot. Whenever this determinant happens to lead to behavior consistent with rational and informed maximization of returns, the business will prosper and acquire resources with which to expand; whenever it does not, the business will tend to lose resources and can be kept in existence only by the addition of resources from outside. The process of ‘natural selection’ thus helps to validate the hypothesis [the ‘as if’ principle] or, rather, given natural selection, acceptance of the hypothesis can be based largely on the judgment that it summarizes appropriately the conditions for survival.

3. 4. 5. 6.

See, however, Koopmans’ (1957) and Blume and Easley’s (1993, 1998) arguments that the ‘as if’ principle is more forceful at the individual than ﬁrm level. Our investigation of capital markets (documented in subsequent chapters) appreciates this criticism. Becker (1962) oﬀers a forceful (partial equilibrium) argument for the role of constraints, rather than explicit rationality, in producing behavior that appears ‘as if’ it was generated by maximizers. Robson (2001a: 14–15) oﬀers a related discussion of ‘foraging theory.’ Dee and Sela (2003) review the relevant literature. See, for example, White (2004). Cohen and Dehejia (2004) ﬁnd related evidence for automobile insurance – that is, insurance coverage accommodates accident-prone drivers, and can thus lead to a signiﬁcant increase in the rate of traﬃc fatalities.

24 7.

8.

The economic method of inquiry Interestingly, monetary authorities and ﬁnancial market participants also encounter diﬃculty when attempting to make credible commitments, and similarly employ ‘handtying’ mechanisms to expand their opportunities. In the case of radical religious militias, individuals who appear more ‘rational’ may lack credibility when promising to act on their ﬁnanciers’ preferences. We’ll see in Part V that, to the extent that this credibility is missing, demanders of ﬁnancial capital (in this case, terrorists) must make do with a constrained set of opportunities (that is, they face a ‘cash constraint’). Competitive environments, in turn, tend to select against such individuals. Berman (2003) thus surmises that: Since club members engage in joint production of local public goods during their hours of non-market time, market work is a distraction with a negative externality for other members. So, eﬃcient clubs should tax market wages. Lacking tax authority they might turn to prohibitions on consumption as a crude but feasible way of lowering wages. Sacriﬁces can be explained as a costly signal of ‘commitment’ to the community . . . A sacriﬁce is then an initiational rite allowing membership and with it access to club goods.

9. 10.

It is very important to understand that, just because such constraints might make feasible a higher level of utility for members, Berman’s (2003) argument does not condone associated behaviors. For example, to the extent that the constraints export costs to others (for example, those whose resources have been destroyed), militias will not maximize social welfare. Led Zeppelin address the related question of ‘What Is and What Should Never Be’ (Led Zeppelin II album). Silberberg and Suen (2001) nicely make this distinction as follows: Positive economics is concerned with questions of fact, which are in principle either true or false. What ought to be, as opposed to what is, is a normative study, based on the observer’s value judgments.

11. 12.

The validity of Baye’s observation relies on markets being complete (that is, all attributes of goods and services being priced). Early social scientists also identiﬁed this merit: It is universally acknowledged that there is a great uniformity among the actions of [individuals], in all nations and ages, and that human nature remains still the same, in its principles and operations. The same motives always produce the same actions: The same events follow from the same causes. (Redman, 1997, quoting David Hume)

3.

A model economy

INTRODUCTION In addition to understanding the necessity of modeling and what constitutes a good model, we now understand why the axiom of maximizing behavior can yield ﬁrmly grounded insights to human sociality. We conclude Part I by employing this axiomatic approach to build a very simple model economy. Building this model will not only strengthen our ability to think like a social scientist, it will also let us precisely deﬁne the fundamental issues in ‘money and banking’ – that is, how societies economize on transacting in spot and future markets, and how this organization relates to economic performance.

MAXIMIZING-HOUSEHOLDS AND THE DEMAND CURVE Equipped with the axiom of maximizing-behavior, we can now build (from the ground up!) a model economy. We’ll start by considering how households’ egoistic behavior yields the familiar principles-level insight that demand curves slope downward. Suppose that households enter our model economy with preferences represented by the indiﬀerence curves shown in Figure 3.1.1 Also assume that, subject to a budget constraint, our households choose consumption bundles ‘as if’ they maximize utility. Characterized as such, our household’s problem is formally identical to that of our moose in Chapter 2 – that is, maximizing-households choose bundles like (x1, x2) from feasible opportunity sets. Why does our axiom yield this implication? In short, any other choice would forgo superior, and feasible, levels of utility. If our household consumed less of good 1 than is represented by x1, for example, then it would not be maximizing. Indeed, for all such inferior bundles, the rate at which the household is willing to forgo units of good 2 to enjoy good 1 (represented by the indiﬀerence curve’s slope) exceeds the rate at which the market requires units of good 2 in exchange for good 1 (represented by the budget constraint’s slope). A symmetric argument can be made against choosing 25

26

The economic method of inquiry

Good 2

Budget constraint

x2 Indifference curve

x1 Figure 3.1

Good 1

The household’s utility maximization problem

bundles where the quantity of good 1 exceeds x1. For a consumption bundle to be an ‘extremum’ (a candidate for maximizing utility), each good’s marginal utility must equal its marginal cost when evaluated at that bundle. We now see that the axiom of maximizing-behavior implies that households choose consumption bundles that equate marginal utility and cost. Pushing the axiom further yields the implication that demand curves slope downward – that is, quantities demanded increase when associated prices decrease. To see this implication, suppose that the relative price of good 1 in Figure 3.1 increases. Following this increase, the ‘budget line’ must rotate inward – that is, the price increase shrinks our household’s opportunities to consume good 1. Presented with this pivot in the boundary of its consumption set, our maximizing-household rationally reduces its consumption of good 1. Figure 3.2 illustrates this implication. Notice that our household’s budget set shrinks (i.e., the triangular set of feasible consumption alternatives becomes smaller). Consequently, if good 1 is ‘normal,’ then a maximizer will demand a lesser quantity of it (that is, a decrease in wealth decreases demand for normal goods). In addition, a maximizer will also substitute toward what has become a less costly good 2. In the end, maximizers demand a lesser quantity of a good when the price of that good increases – that is, maximizing-behavior implies that demand curves slope downward. This theoretical relationship ﬁnds such widespread empirical support that economists refer to it as the ‘law of demand.’2

27

A model economy

Good 2

Budget set

Indifference curve Good 1’s price increases x'1

x1

Good 1

Quantity demanded of Good 1 decreases

Figure 3.2 An increase in relative prices shrinks the household’s consumption possibilities

MAXIMIZING-FIRMS AND THE SUPPLY CURVE The ﬁrm’s problem, and insights that we can draw from it, essentially mirror those of the household. In short, we can formalize the ﬁrm’s problem as choosing input-output combinations to maximize proﬁts on technologically feasible production sets. We denote such a combination in Figure 3.3 as (x*, y*). Notice that, at this combination, the rate at which the ﬁrm can transform inputs into outputs (as represented by the technological frontier’s slope) equals that at which the market willingly trades inputs for outputs (as represented by the iso-proﬁt curves’ slope). Any other feasible combination would be inconsistent with proﬁt-maximization – for example, the market’s compensation would exceed the ﬁrm’s opportunity cost of inputs for inputlevels below x*, in which case the ﬁrm could increase proﬁts by expanding production. In a formally identical manner to that for the consumer’s problem, this condition is necessary for the ﬁrm’s input-choice to be a maximand. Maximizers, whether they are ﬁrms or households, engage in marginal analysis. We can now derive a ﬁrm’s supply curve in a manner that mimics our derivation of a household’s demand curve. Without loss of generality, suppose that

28

The economic method of inquiry

Outputs Iso-profit curves

y* Profit – 2

Production set

Profit – 1

Inputs Figure 3.3

x*

Profit – 0

The ﬁrm’s proﬁt-maximization problem

(real) output prices increase. This increase ﬂattens the ﬁrm’s iso-proﬁt curve (to maintain any level of proﬁt, a relatively small increase in output is necessary to oﬀset a given increase in input-cost). Figure 3.4 illustrates this increase and its implications for the maximizing-ﬁrm’s choice of factor inputs. Here, we’ve drawn the old iso-proﬁt curve as a dashed line and the new curves as solid lines. By causing our ﬁrm’s iso-proﬁt curve to ﬂatten, our hypothetical price change induces the ﬁrm to produce a higher level of output. Notice that, at the old proﬁt-level and new iso-proﬁt curve, marginal revenue exceeds marginal cost. Faced with an output-price increase, our ﬁrm will thus want to increase production from the level that maximized proﬁt under the old relative prices. In this manner, an upward-sloping supply curve follows logically from the assumption of maximizing-behavior – that is, an increase in output prices causes maximizing-ﬁrms to increase the quantity of output that they willingly supply.

GOVERNING MONEY AND CAPITAL MARKETS Having worked this far through our book, you should see more clearly what’s behind a principles-level model of an economy. You should be curious too about why this realization is important for understanding phenomena related to money and banking.

29

A model economy

Outputs

Iso-profit curves

Output – new price Production set

Output – old price

Inputs Figure 3.4

An increase in output prices increases output

This question has at least two very good answers. First, recall that substantive boundaries need not constrain the insights that thinking like an economist yields. Our improved understanding of this analytical method should thus facilitate a fruitful investigation of any social phenomena, including those that fall under the ‘money and banking’ heading. Second, this development sheds a bright light on how a principles-level training can discourage investigations in this direction, and why pursuing such an investigation is nevertheless important. As we’ve built it so far, our model oﬀers insight to how high-level economic properties can emerge from individual-level actions. To achieve this progress, however, we kept our model rather simple. In particular, we implicitly assumed that households and ﬁrms expend a negligible amount of resources when choosing, respectively, consumption bundles and production factors – that is, we assumed that economic agents costlessly transact. If we want to investigate phenomena related to money and banking, however, then models that rest on such an assumption cannot be ‘good.’ Indeed, models like these imply that monetary and intermediation services are superﬂuous, and should thus be absent from economies that consist of maximizing households and ﬁrms. The fact that we observe such features strongly implies that something is missing from our preliminary model’s set-up. To see the diﬃculty that maintaining this relatively simple framework creates, consider our implicit assumption that households maximize utility

30

The economic method of inquiry

subject only to a budget constraint – that is, households can move costlessly from their endowments to feasible consumption bundles that maximize utility. Here, maximizers cannot have an interest in developing a monetary system. Indeed, forgoing the resources that are necessary to build such a system is inconsistent with maximizing-behavior since households can already costlessly transact. Abstracting from transactions costs can facilitate insights to economies whose monetary systems are ‘well-developed’ – that is, our principles-level model may usefully characterize economies where households readily exchange currency for goods and services. But even here, going forward with such an assumption sweeps aside the important question of how to maintain a well-developed monetary system. Even more, it ignores the questions of what is money, how does money’s management inﬂuence economic performance, and why individual maximizers might pursue this management in something other than a socially optimal manner. It is with these and related questions that our principles-level model of the household leaves us, and to which we turn in Parts II–IV. A principles-level model of the household thus appears to be a reasonable point of departure for our examination of ‘money.’ Similarly, a principles-level model of the ﬁrm helps set the stage for investigating how ﬁnancial intermediation inﬂuences real activity and why, despite a potentially sound understanding of that inﬂuence, individual maximizers might regulate those services in a less than socially optimal manner. To anticipate this investigation, recall that our principles-level model rests on the assumption that ﬁrms maximize proﬁt subject only to a technology constraint (that is, choose a feasible input-output bundle that intersects the highest iso-proﬁt curve). Here, price taking ﬁrms costlessly access input bundles to maximize residual earnings. But suppose that a ﬁrm must purchase inputs before it recognizes payoﬀs. Suppose also that the ﬁrm under consideration is ‘cash constrained’ – that is, it is ‘project rich’ but capital poor in the sense that it controls insuﬃcient resources for ﬁnancing at least one fundamentally proﬁtable project. Here, the ﬁrm cannot maximize proﬁts because accessing necessary factors is not possible. In Figure 3.3, the ﬁrm would have to restrict its input choice to what can be funded with internal resources. Input-levels as high as x* may lie outside of this feasible set. So why would we assume that ﬁrms do not face a cash constraint? In a manner analogous to that of our household’s problem, making such an assumption might be acceptable when examining ﬁrms that operate in well-developed capital markets (markets where moving ﬁnancial capital consumes little in the way of real resources). But this assumption also begs a number of important questions. For example, what does it mean for a

A model economy

31

capital market to be well developed? And if assuming that capital markets are well developed is innocuous, then why do so many ﬁrms ﬁnd themselves operating in poorly developed markets? Even more, why would maximizers settle for anything less than a well developed capital market? It is with these and related questions that our principles-level model of the ﬁrm leaves us, and to which we turn in Part V of the book.

CONCLUSION We have now strengthened our understanding of the foundation on which a principles-level model of the economy rests, and why this foundation appears ‘weak’ if we want to understand phenomena related to money and banking. In particular, we should recognize that maximizing-households have no incentive to develop monetary systems if bartering is costless, and ﬁrms have no incentive to develop ﬁnancial markets if hiring inputs is costless. Absent this recognition, we cannot begin to rationalize why maximizers would expend resources to develop monetary and intermediation systems. But even casual observations suggest that maximizers do allocate resources in this manner, and that economies that maintain well-developed systems in this regard also tend to exhibit superior performance. We thus leave Part I of our book with a set of scientiﬁcally interesting and socially important questions, as well as a sound analytical framework for addressing them.

NOTES 1. Frank (2006, Chapter 7), among others, oﬀers a more complete textbook treatment of this topic. 2. ‘Both prestige and signaling eﬀects could result in demand curves that slope upward for some range of prices. Even so, personal experience and countless studies from economics and marketing conﬁrm that the law of demand applies to most products’ (Besanko et al., 2000).

PART II

Money in a static economy

What is money? Why do people accept currency in exchange for, say, a new car? Cars, after all, produce easily observed and valued services – for example, transportation. But what services does money produce? More importantly, how do these services aﬀect economic performance and, thus, how should their production be organized? Addressing these questions in Part II, we’ll learn that money economizes on the cost of transacting in spot markets (that is, it facilitates trades at a point in time). In well-developed monetary systems, transacting exhausts little of an individual’s endowment, and thus leaves large ‘residuals’ with which to facilitate consumption. On the other hand, as monetary services become less developed, individuals must increasingly forgo consumption possibilities to facilitate trade, and thus make do with inferior levels of utility. But even if money inﬂuences economic performance, societies can (and often do) fail to productively organize their monetary aﬀairs. This failure emerges, in part, from the incentive for monetary authorities to substitute ‘inﬂation taxes’ for ‘real taxes.’ Indeed, commitment problems that arise from such incentives may have discouraged all societies, except those from the last several decades, from enjoying the superior economic opportunities that well-developed monetary systems make available. On its face, this ‘failure’ appears inconsistent with our axiom of maximizing behavior. It is not. Rather, deductions from our axiom identify the root cause of such failures, and how organizational strategies can productively respond. In short, our axiom implies that maximizers not only have an incentive to reduce transactions costs, but to also strategically manipulate the mechanisms that produce such eﬃciencies (that is, favor themselves in distributing eﬃciency gains). Unless societies can deter such opportunism, their economies will rest at inferior outcomes.

34

Money in a static economy

We’ll thus see that constraining maximizers can be necessary to achieve superior outcomes and that, despite the prospect of expanded economic opportunities, societies can exert considerable resistance to developing sound monetary systems (Miller, 1991 investigates the nature of, and organizational responses to, such dilemmas more generally). We’ll build on this understanding in Part III by pushing Part II’s ‘static’ framework (that is, an investigation that abstracts from timing issues) into a ‘dynamic analysis’ of how monetary services relate to economic ﬂuctuations (business cycles). Taken together, Parts II and III uncover money’s relationship to two important features of economic performance – an economy’s consumption possibilities and ﬂuctuations therein. These insights, in turn, point to the governance of monetary authorities as fundamentally inﬂuencing economic well-being, and thus set the stage for our investigation of central banking in Part IV.

4.

What is money?

MONEY FACILITATES EXCHANGE What is ‘money?’ On its face, this question appears rather trivial. After all, when a friend asks whether we have money, we know exactly what he or she is asking. Or do we? Is, for example, currency ‘money?’ Here, popular terminology happens to serve us well – currency is money. But why is currency money? In other words, what fundamental feature gives currency its ‘moneyness?’ To address this question, let’s consider what else might be ‘money.’ Could we, for example, employ housing units as money? Your immediate answer to this question is probably ‘no.’ After all, we don’t buy groceries by transferring to store-owners a right to consume housing stocks (at least not directly). But before dismissing this question, let’s consider the following (hypothetical!) proposition. Suppose that we attempt to employ you as a research assistant. Suppose further that, in return for your services, we’ll give you either 20 US dollars or a really big, well-constructed house, located in an attractive neighborhood. Holding all else constant, which of the two forms of remuneration would you choose? The house, of course! So, at least in principle, there exist assets other than those that we commonly refer to as money that can act like money. We can think of numerous other examples, but the bottom line for each is the same. For an asset to act like ‘money,’ it must be acceptable as a ‘medium of exchange’ – that is, anything that an economy’s agents readily accept in return for supplying goods and services.1 We’ll refer to as ‘moneyness’ the extent to which assets produce this type of transaction service. It is this service that currency, and to some extent even housing units, produce for an economy.2 And while this service is relatively hard to see, it strongly inﬂuences economic performance.

BOX 4.1

THE ‘ISLAND OF STONE MONEY’

Yap is part of the Federated States of Micronesia, a group of 607 small islands in the Western Pacific Ocean, lying about 2500 miles 35

36

Money in a static economy

southwest of Hawaii, and just above the Equator. This island’s citizens (Yapese) draw considerable attention from both tourists and social scientists for employing crystalline stone as money.This medium of exchange, known as ‘fei,’ consists of thick stone wheels that range in diameter from 1–12 feet (30 cm to 3.5 m), and weighs as much as 4 tons. A hole in the center of larger wheels lets owners transport fei by rolling them with an inserted pole (Friedman, 1992). A noteworthy feature of fei is that its owner does not have to possess the stone money to transact with it. For example, after concluding a trade in which the fei-price is so high that a stone cannot feasibly be moved, the seller is content with accepting a simple acknowledgment of ownership in the buyer’s fei. This feature of Yap’s monetary system is strikingly illustrated by a family whose wealth was acknowledged by all of the island’s citizens, even though no one (not even the family itself!) ever physically observed this wealth. You have probably anticipated that not all stones pass as fei. Rather, the Yapese used to quarry fei from the ‘rock islands’ of Palau. And quarrying was not cheap (imagine the inflation if it was!). Indeed, the Yapese had to canoe over 600 miles (one way) to reach the quarries (Land of Stone Money)! During a return trip from Palau, a Yapese family’s canoe sank, and with it, an enormous fei. But even though the family’s ‘money’ fell to the Pacific Ocean’s bottom, other islanders recognized its purchasing power just as if the stone was readily visible – ‘The purchasing power of that stone remains . . . as valid as if it were leaning visibly against the side of the owner’s house’ (Friedman, 1992). Another interesting illustration emerges from Germany’s colonial occupation of Yap from 1899–1919. During this period, Germans tried to employ Yapese to construct and repair the island’s infrastructure (for example, roads). But the Yapese weren’t very willing agents, at least until the Germans painted large black crosses on the islanders’ fei, many of which were leaning against their owners’ huts. Marked as such, the fei became worthless to the Yapese – no one would accept a defaced fei as an exchange medium. The Yapese were thus persuaded to work for the Germans. And once the islanders’ labor produced sufficient value to pay the Germans’ ‘fine,’ crosses were erased and the fei regained its purchasing power (Friedman, 1992). This story is important for our purposes because it illustrates how money’s value comes from its capacity to facilitate trade, not from anything intrinsic in the money per se. The parallel between

What is money?

37

fei and money in developed economies is striking.3 Consider the family whose fei rests at the ocean’s bottom. Other islanders knew about the stone, and that members of a particular family maintained the right to direct the stone’s purchasing power. They thus readily accepted (unobservable) portions of that fei as an exchange medium. But don’t the same characteristics describe today’s checking accounts? Suppose, for example, that our checking account physically resides in St Louis.Will this remote location discourage a merchant in, say, San Francisco from accepting some of our ‘St Louis stone’s’ purchasing power in return for goods or services? No! Checkable deposits act like money for exactly the same reason as do the Yapese fei – because parties to a trade readily accept it as a medium of exchange. Our deposits have no more intrinsic value than do fei – indeed, neither has any capacity to directly increase our consumption. Rather, both draw their value from being an agreed upon exchange medium, and thus facilitating transactions.

EVOLUTION OF EXCHANGE MEDIA From Commodity to Fiat Money Understanding how money facilitates transactions will help us see how money can inﬂuence economic performance. Examining money’s evolution is instructive in this regard. People appear to have ﬁrst used ‘money’ in the form of livestock over 11 000 years ago. Subsequently, seashells, leather, metals and other commodities found such employment (see the Public Broadcasting Service tutorial The History of Money). Indeed, prior to 1971, every major system of money (directly or indirectly) based itself on a commodity (Friedman, 1992). Remarkably, we are among the ﬁrst people in recorded history to use ‘ﬁat currency’ as an exchange medium! Fiat currency has no value outside of its capacity to facilitate transactions. Rather, it creates value by helping traders coordinate on a pure exchange medium. The US dollar, for example, simply declares that ‘This note is legal tender for all debts, public and private.’

38

Money in a static economy

Beneﬁts of Fiat Money The advantages of employing a ﬁat currency are readily observable. For example, consider the resource cost of transporting paper (or even electronic) money, versus that of transporting a commodity such as gold. By economizing on such costs, ﬁat systems can leave economies with relatively large residual endowments – that is, resources that are left untouched by transacting, and are thus available to facilitate consumption. In addition, commodity money can distort decisions by raising the relative price of commodities that produce both consumption and exchange services (see Blanchard and Fischer, 1989). To see the consequent constraint on economic well-being, consider Figure 4.1. Suppose that a society employs a commodity as money (for example, silver), and that this commodity also enjoys industrial applications. Also note that, if the services that our commodity produces are not easy to separate, then maximizers will develop and employ monetary services until the marginal beneﬁt of doing so equals the marginal cost of the joint services. In our figure, this equilibrium obtains at one unit of the commodity. But because the marginal cost of joint services exceeds that of either constituent service, this equilibrium cannot be eﬃcient. Indeed, the marginal beneﬁt of monetary services must exceed those services’ marginal cost.

Units

Marginal benefit of currency services

Marginal cost of currency services

Marginal cost of industrial services

Marginal cost of joint services

0

5

1

1

2

1

4

2

2

4

2

3

3

3

6

3

2

4

4

8

4

1

5

5

10

Figure 4.1

Commodity money distorts relative prices

What is money?

39

Here, commodity money precludes mutually beneﬁcial trades of monetary services. Fiat currency, on the other hand, allows for a closer alignment of each service’s marginal beneﬁts and costs, and thus facilitates the recognition of superior outcomes. By separating monetary services from others, a ﬁat currency makes feasible the types of mutually beneﬁcial trades that are impossible in our example’s ﬁrst case. In Figure 4.1, such trades induce an equilibrium at two currency units, where the marginal beneﬁt and cost of monetary services are equal.4 By allowing for the separate production of monetary and industrial services, ﬁat money lets those services migrate more freely to their highest valued uses. Commitment with Commodity Money If ﬁat currency is so attractive, then why has almost everyone in recorded history used commodity money (or barter) to facilitate trade?5 To answer this question, notice that the generation of any beneﬁt creates a cost. We should thus expect the beneﬁts of ﬁat money to come at a cost, and this cost to discourage ﬁat systems from developing. We’ll see in subsequent chapters that monetary authorities face a strong incentive to opportunistically manipulate money’s supply. The cost of acting on this incentive, however, is greater for commodity money than it is for ﬁat money. Increasing the supply of gold, for example, probably costs more than does increasing the supply of a paper currency.6 Committing against such increases may thus be easier within a commodity system. Data from Figure 4.2 support this hypothesis. Under President Richard Nixon, the US left the gold standard in the early 1970s. Moving to a ﬁat system, the economy enjoyed beneﬁts described above (for example, see our discussion surrounding Figure 4.1). However, it may have also sacriﬁced, the capacity for commodity systems to constrain monetary authorities from manipulating the money supply. Consistent with this constraint being relaxed, Figure 4.2 evidences a trend-break in consumer prices (an increase in inﬂation) in the early 1970s.

CONCLUSION Money is anything that acts as a medium of exchange. Until very recently, history’s societies employed various commodities as exchange media, many of which (for example, stones) strike us as being fundamentally worthless. Remember, however, that money’s value comes from facilitating exchange and thus indirectly expanding consumption possibilities. In

40

Money in a static economy

200.0 150.0 100.0

Source:

January-05

January-99

January-93

January-87

January-81

January-75

January-69

0.0

January-63

50.0

January-57

CPI (less food & energy)

250.0

Federal Reserve Bank of St Louis 2007.

Figure 4.2 Commitment problems when moving from commodity to ﬁat money?

this sense, ﬁat money exhibits greater ‘moneyness’ than do its commoditybased counterparts. To enjoy this beneﬁt, however, societies must develop organizational substitutes for producing the commitment beneﬁts of commodity money.

NOTES 1. Note that money must also act as a ‘store of value.’ If money did not store value, it could not facilitate anything but instantaneous exchange (but then money would be superﬂuous). Note also, however, that money is inferior to other assets as a store of value (consider, for example, interest bearing bonds). ‘What distinguishes money from other stores of value is its liquidity, and what underlies the liquidity of money is the fact that it is the common medium through which other commodities are exchanged’ (Ostroy and Starr, 1990: 4). This distinction has important implications for how the supply of money should be measured, an issue to which we turn in Chapter 10. 2. Note, however, that currency only produces transactions services, and thus mitigates price-distortions for monetary assets that produce other than exchange services. We address this and related issues in the following chapters.

What is money?

41

3. Friedman (1992) highlights this parallel by asking ‘how many of us have literal personal direct assurance of the existence of most of the items we regard as constituting our wealth? What we more likely have are entries in a bank account, property certiﬁed by pieces of paper called shares of stock, and so on and on.’ 4. Notice that ﬁat currency also expands the opportunities for mutually beneﬁcial trades in industrial services. 5. Selgin (2003) addresses this question by examining how the monetary system of ‘adaptive learners’ would evolve. 6. Even systems of commodity money, however, are exposed to the consequences of monetary authorities’ opportunistic incentives – consider, for example, the practice of seignorage.

5.

Money and the level of economic well-being

INTRODUCTION We now understand that money is anything that acts as an exchange medium, and informally developed the idea that money ﬁnds value from its capacity to facilitate trade. Our objective for the present chapter is to further develop this idea. We’ll do so by examining how money aﬀects an economy’s consumption possibilities (that is, consumers’ budget sets from Part I), and the level of real economic activity more generally (that is, the level of Q or ‘quantity’ from equilibrating Part I’s demand and supply sides). Leaving this chapter, we should better appreciate how money, despite having no direct eﬀect on consumption possibilities, strongly inﬂuences the level of welfare that societies enjoy.1

THE INFERIOR BARTER EQUILIBRIUM An Illustration Suppose, for a moment, that we ﬁnd ourselves in a money-less world. Suppose also that, while we enjoy a substantial resource endowment (that is, we are ‘rich’), there also exist consumption bundles that would make us happier than would consuming our endowment. How might we trade for these more attractive bundles? For example, if our endowment does not include a car, and we maintain a strong preference for a car, how would we act on this preference? Notice that we cannot simply go to an auto dealer and exchange ‘money’ for a car. Rather, we must happen upon an individual who owns a car we want and maintains a suﬃciently strong preference for elements of our endowment. In other words, to enjoy consumption bundles that are superior to our endowment, we have to ﬁnd individuals with whom there exists a ‘double coincidence of wants.’ Having to satisfy a double coincidence of wants, however, diminishes our opportunities. To see this implication, let’s continue with the assumption that 42

Money and the level of economic well-being

43

Boundary of set of feasible consumption bundles when bartering is costless Good 2

Boundary of set of feasible consumption bundles when bartering is costly

u'

u

e2

e1 Figure 5.1

Good 1

Inferiority of the (costly) barter equilibrium

we are ‘rich’ (that is, our resource endowments are ‘large’). Given our endowment’s size, we might think that obtaining a new automobile is easy. But such is not the case in a money-less economy. Here, unless the party from whom we demand a car has a suﬃciently strong preference for elements of our endowments, it will not trade. We may, as a consequence, be stuck with our original resource endowment and the inferior level of utility that it produces. Furthermore, even if we successfully trade for a car, we will likely exhaust a considerable part of our endowment in searching for a suitable trading partner. Figure 5.1 illustrates this phenomenon. Consuming our endowment is feasible in both monetary and barter economies (consuming the bundle of goods (e1, e2) is feasible, with or without money). In a monetary economy, however, we can easily trade away from our endowment. Here, the act of transacting negligibly exhausts our endowment – that is, our trading partner receives almost everything that we give up in a transaction. We thus represent the ease of transacting in a monetary system by assuming

44

Money in a static economy

that our budget constraint maintains the same slope on either side of its endowment. Bartering, by comparison, exhausts a considerable amount of resources. Absent a well-developed monetary system, we must search longer for a suitable trading partner (that is, we have to satisfy a double coincidence of wants) and may ﬁnd it necessary to make a greater number of trades before reaching an optimal outcome. In either case, transacting exhausts more of our endowment than does trading in a monetary economy, leaving our trading partner with considerably less than what we give up in a transaction. We represent the phenomenon shown in Figure 5.1 by drawing the budget constraint in a barter economy with a kink at the endowment. The diﬀerent slopes of our constraint represent the increased cost of transacting as we move further away from our endowment (that is, the increased diﬃculty of satisfying a double coincidence of wants or the increased number of trades that are necessary to reach a superior outcome). Notice that our opportunity set in this economy must be smaller than that in a well-developed monetary system. Demand for each good must thus be (weakly) less in a barter economy than it would be in a monetary economy. Likewise, the car’s supplier must recognize an inferior level of utility. Goods and services (almost) certainly exist for which our potential supplier would willingly trade his or her car. But while our resource endowment is ‘large,’ the supplier’s preferences may be such that he or she does not place much value on its elements. Hence, unless another auto-demander presents him- or herself to our supplier, and that demander’s endowment includes elements that the supplier ﬁnds more preferable, the supplier will also be stuck with his or her endowment (and the consequent inferior level of utility). Moreover, even if the supplier ultimately ﬁnds a suitable trading partner, having to forgo preferable consumption bundles while searching (as well as relatively high search costs) will constrain consequent increases in his or her utility. A Formal Model Our illustration suggests that the relatively high cost of bartering constrains economic opportunity set. A more formal treatment of this phenomenon conﬁrms our suggestion. Consider, for example, a society of three individuals (to whom we’ll refer as ‘players 1, 2 and 3’) and assume that they reside in a ‘linear city’ as shown in Figure 5.2 (see (i) Players’ locations). In addition, suppose that each individual arrives with an endowment of good X, Y or Z (see (ii) Endowments), but maintains preferences

45

Money and the level of economic well-being

1

2

3

(i) Players’ locations

Figure 5.2

1

2

3

(ii) Endowments

Z

X

Y

(iii) Preferences

X Y Z

Y Z X

Z X Y

(iv) Equilibrium allocation

Z

Y

X

(v) Efficient allocation

X

Y

Z

Inferiority of barter in a ‘linear city’

over those goods as in the figure’s third row (see (iii) Preferences, where players’ most preferred goods are listed ﬁrst). Let’s examine, then, how well our players can do by trading their endowments. Assuming that trade is feasible only with immediate neighbors (for example, the cost of encountering a distant neighbor may be prohibitive), our society’s members will ultimately rest at the Equilibrium allocation (iv).2 But while this allocation is stable, it is not optimal. Indeed, comparing it to the Eﬃcient allocation in (v), we see that transaction costs induce an ‘inferior equilibrium’ – that is, trading goods Z and X would make both players 1 and 3 better oﬀ without making anyone worse oﬀ. Nevertheless, our economy cannot reach the optimal outcome because (a) trading takes place only when a double coincidence of wants exits and (b) while player 1’s and 3’s wants ‘coincide,’ the physical distance between them discourages their meeting. We thus see how a barter system, by necessitating a double coincidence of wants, can stop individuals from enjoying the relatively high levels of utility that lower cost transaction-technologies make feasible.

46

Money in a static economy

THE SUPERIOR MONETARY EQUILIBRIUM An Illustration Radford (1945) oﬀers an interesting illustration of how individuals can be better oﬀ when money facilitates trade than when trade relies on satisfying a double coincidence of wants. While a prisoner of World War II, Radford observed that his ‘standard of material comfort’ depended on the ease with which he could exchange goods from his ration’s ‘endowment.’ He also observed that, by circumventing the necessity to satisfy a double coincidence of wants, the emergence of cigarettes as the prison camp’s exchange medium expanded his opportunities to consume goods and services that were preferable to his endowment (that is, the camp’s ‘monetary economy’ produced better outcomes than did its ‘barter economy’). Indeed, when the camp’s supply of cigarettes encountered interruptions, prisoners increasingly used barter as an exchange mechanism. During such periods, the camp’s overall level of economic activity also appeared to decrease, as did the prisoners’ material well-being.3 A Formal Model Figure 5.2 shows how the lack of transaction services can stop individuals from realizing attractive equilibria. More importantly, it shows that the associated welfare-loss does not emerge from a resource constraint, but rather from the relatively high cost of trading in barter systems (for example, opportunities that one must forgo while attempting to satisfy a double coincidence of wants). Radford’s account of the POW economy oﬀers empirical support, then, for the hypothesis that monetary systems can facilitate superior outcomes. We can develop this implication more carefully by extending our previous model of an exchange economy. Let’s do so by carrying over each player’s endowment (that is, each player maintains the endowment speciﬁed in Figure 5.2 (ii) Endowments), and introducing a ‘clearinghouse’ (CH) that stands willing and able to trade a single ‘piece of paper’ (POP) for a unit of any good X, Y or Z. Why might an equilibrium in this monetary economy (notice that our POPs act as an exchange medium) dominate that which obtains in its barter analog – that is, an economy where a simple medium like POPs is missing? To address this question, suppose that each of our players sells his or her endowment to the CH. Note that the CH now ‘owns’ one unit of each good X, Y and Z, and each of our players owns one POP. But what will our players do with their POPs? After all,

Money and the level of economic well-being

47

POPs are simply pieces of paper and thus cannot increase utility by themselves. Given that our players are maximizers, it follows that they will trade inherently worthless POPs for utility-enhancing commodities. In particular, each maximizer will ‘buy’ from the CH his or her most preferred good (that is, player 1 will exchange a POP for one unit of good X, and so on), and these transactions leave each player with his or her most preferred good. Remarkably, by introducing fundamentally worthless pieces of papers as an exchange medium, we moved from the unattractive barter equilibrium to an attractive monetary equilibrium – an outcome from which no possibility exists to make any player better oﬀ without making another worse oﬀ. So how did this transformation take place? Why, for example, could our players under the barter system not have simply dumped their endowments into a bucket and taken turns withdrawing each individual’s most preferred good? Notice that such a process would have left our players with the same optimal allocation as does trading under the monetary system. But also notice that bartering players would have confronted an important diﬃculty. In particular, after dumping their endowments into a bucket, they would have to know whose right it is to grab goods back out. We can easily imagine the resources that our players would have exhausted in coming to and enforcing such an agreement.4 Moreover, the very prospect of exhausting resources in this manner could have dissuaded our maximizers from contributing goods to the bucket in the ﬁrst place! In either case, our players would again ﬁnd themselves at an inferior equilibrium. By simply introducing a system of POPs, on the other hand, our monetary system appears to have circumvented such potential diﬃculties. Indeed, after throwing their endowments into the bucket (the CH), each player’s ‘right’ to grab goods back out is evidenced by how many POPs he or she holds.5 Consequently, as long as the players have conﬁdence that POPs can continually be traded for goods, they will move toward the ‘ﬁrstbest’ equilibrium that our principles-level consumers enjoy, and away from the inferior allocation at which maximizers in a barter system must rest.6

BARTER VERSUS MONEY IN A PRODUCTION ECONOMY We now see that money can increase welfare by letting individuals trade, even when their ‘wants’ do not coincide. To motivate this argument, however, we restricted our attention to ‘endowment economies’ – that is, economies (like the POW camp) where individuals do not produce

48

Money in a static economy

anything, but rather attempt to maximize utility by trading elements from their endowments. By considering a production economy (that is, one in which individuals transform raw endowments (such as labor) into consumables), we can recognize even deeper beneﬁts. Notice that, by permitting an economy to operate at a relatively low level of transactions costs, money encourages individuals to specialize in production. Our barter economy, on the other hand, discourages maximizers from specializing by increasing the cost of acquiring goods and services that individuals do not produce themselves. In barter economies, maximizers can do better by making a variety of goods, and thus rationally forgo becoming eﬃcient producers of any. This compromised eﬃciency ultimately diminishes an economy’s productive capacity. Here, a lack of specialization hinders learning, and thus leaves ﬁrms with inferior technologies. At any price level, ﬁrms must supply a smaller quantity of goods or services. This supply-side eﬀect exacerbates the adverse demand-side consequences of bartering in an exchange economy (see Figure 5.1) to forgo an even larger part of an economy’s potential. Figure 5.3 illustrates this phenomenon. Households’ demand with costly bartering

P

Households’ demand with costless bartering AS'

AS Firms’ supply with superior technology Firms’ supply with inferior technology

PE P'E

AD' Q'E

QE

AD Q

Figure 5.3 Barter further constrains economic performance in production economies

Money and the level of economic well-being

49

CONCLUSION If monetary systems can greatly improve welfare, then why don’t all economies employ eﬃcient ones?7 This question is important, and the reason that it remains open for now stems from our condition that ‘players have conﬁdence that POPs can continually be traded for goods.’ To carefully address this question, we must understand why players (who are themselves maximizers) should maintain such conﬁdence. We’ll build this understanding in Part IV by investigating how monetary authorities govern themselves to build credibility. Before turning to this investigation, however, let’s be clear about how the present chapter’s problem motivates it. Notice that, if ‘clearinghouses’ are maximizers, they want to exchange new POPs for our players’ endowments, and thus generate real revenues through ‘seignorage’ (a strategic increase in the money supply).8 In cases like this one, money’s ability to ‘store value’ weakens, and thus so does its acceptability as an exchange medium. By assuming that ‘players have conﬁdence that POPs can continually be traded for goods,’ we have set this issue to the side for now. We might also be curious about whether the incentive to pursue seignorage is the only source of skepticism for a monetary authority’s constituents. Indeed, while considerably richer than our principles-level model from Part I, the models that we develop in Part II remain simple in that they abstract from timing issues. Before examining the organization of monetary authorities, then, we should also look at our clearinghouse in a ‘dynamic economy’ (that is, one in which trades take place across diﬀerent periods). We’ll turn to this issue in Part III, and see that monetary authorities not only have an incentive to raise real revenue by opportunistically increasing the quantity of POPs, they also have an incentive to time monetary changes to inﬂuence ﬂuctuations in a production economy’s performance. In addition to promising against inﬂation taxes, then, a ‘sound’ monetary authority must also promise against opportunistically manipulating the money supply to inﬂuence an economy’s output level.

NOTES 1. Economists have heretofore encountered diﬃculty in attempting to achieve this understanding (see Blanchard and Fischer, 1989). Making progress in this direction is important, however, for seeing how the management of an economy’s money supply inﬂuences economic performance. We’ll employ these insights in Part IV to consider how societies should organize their production of monetary services. 2. This particular allocation becomes an equilibrium by exhausting all feasible and mutually beneﬁcial trades.

50

Money in a static economy

3. ‘True, a prisoner is not dependent on his exertions for the provision of the necessaries, or even the luxuries of life, but through his economic activity, the exchange of goods and services, his standard of material comfort is considerably enhanced . . . Everyone receives a roughly equal share of essentials; it is by trade that individual preferences are given expression and comfort increased . . . Cigarettes became the normal currency . . . While the Red Cross issue of 50 or 25 cigarettes per man per week came in regularly, and while there were fair stocks held, the cigarette currency suited its purpose admirably. But when the issue was interrupted, stocks soon ran out, prices fell, trading declined in volume and increasingly became a matter of barter . . . Most of our economic troubles could be attributed to this fundamental instability’ (Radford, 1945). 4. For example, having each player show up at the same time might facilitate agreement. It would also consume more resources, however, than would impersonal trade. 5. Ostroy and Starr (1990: 8–9) oﬀer a related illustration of money’s ability to facilitate ‘record-keeping’ in this regard. 6. The examination of ‘coordination games’ can oﬀer insight to how conﬁdence in a particular exchange-medium emerges. Centralized solutions to such games often dominate decentralized solutions, for example, arguing against the eﬃciency of competitive markets in producing monetary services. Fitness landscapes that characterize such games can also favor the evolution of norms, such as those that govern individuals’ willingness to accept a particular exchange-media. Richerson and Boyd (2001) oﬀer an introduction to this idea. 7. We might even ask why 11 000 years had to pass for monetary systems to move from those that are commodity-based to those that rely on ﬁat currency! 8. To the extent that the quantity of goods is stable, increasing the money supply increases nominal prices between the time the CH exchanges money for a good and the time another individual arrives to demand that good. During this period, real goods held by the CH will increase in value (as measured by the number of POPs that can be acquired for goods) and nominal money held by households will decrease in value (as measured by the number of goods that can be acquired with POPs).

PART III

Money in a dynamic economy

In our previous set of chapters, we argued that money ‘matters’ because it can improve an economy’s performance relative to what barter makes available. We also learned that, to enjoy this improvement, economic actors must share conﬁdence that their exchange medium is readily acceptable for trade. (After all, why did maximizers trade real goods and services for the ‘pieces of paper’ (or POPs) that circulated in our model monetaryeconomy?) Realizing this conﬁdence’s importance, we want to address how societies build or maintain it. Before moving in this direction, however, we’ll further motivate in Part III the importance of how societies organize monetary services by considering how money relates to ﬂuctuations in aggregate economic activity (that is, business cycles, or ‘high-frequency’ components of relevant economic time series). One channel through which money might aﬀect ﬂuctuations is that of ‘nominal rigidities.’ ‘Rigidities’ refer to the phenomenon that an economy’s nominal prices (those measured in units of exchange media) need not move together. For example, the ‘price’ of labor (that is, wages) might move more slowly than does that of ﬁnal goods and services. In cases like this one, expanding money’s supply can cause labor prices to increase more slowly than do output prices. This heterogeneous response, in turn, encourages proﬁt-maximizing ﬁrms to expand output until slow-moving increases in marginal costs catch up with more ﬂuid increases in marginal revenue. Nominal rigidities can thus open the door to money exerting a real inﬂuence (that is, aﬀecting economic opportunities). While nominal rigidities enjoy pedagogical prominence (and frequently make their way into popular media and policy debates), however, they encounter several problems. For example, we’ll ﬁnd the above explanation diﬃcult to maintain if we don’t ﬁrst relax the axiom of maximizing-behavior.

52

Money in a dynamic economy

Our axiom implies that individuals form expectations ‘rationally,’ and thus cannot be systematically fooled by a rate of increase in nominal wages that lags that of output prices. Moreover, important features of the data do not support nominal rigidity theories. For example, contrary to the implication of standard rigidity models, real wages appear to be pro-cyclical, and the rate at which nominal prices actually change appears to be greater than what is consistent with rigid prices. Despite these diﬃculties, money continues to exhibit a rather robust reduced form relationship with output. Hence, while we might not be comfortable with the capacity of nominal rigidities to rationalize this relationship’s structure, we should still be curious about other channels through which money might exert a real inﬂuence. We’ll thus consider alternative channels through which money can inﬂuence aggregate ﬂuctuations, including those associated with ‘unexpected inﬂation,’ ‘eﬃciency wages’ and ‘credit.’ This consideration will further our understanding of whether and how money aﬀects economic stability. It will also leave open two very important questions. The ﬁrst is normative – should political agents actively manage the supply of monetary services? In addressing this question, we’ll see that monetary authorities encounter considerable diﬃculty in ‘ﬁne tuning’ economic performance. Moreover, even if authorities can overcome such diﬃculties, their capacity to improve economic well-being by dampening economic ﬂuctuations may be small. These arguments suggest that monetary authorities should passively manage money’s supply. Despite such prescriptions, however, our empirical reality is one where monetary authorities tend to pursue activist policies. A second question, then, is why political agents actively manage an economy’s money supply, even if doing so might shrink economic opportunities. To address this question, we’ll consider an interesting paradox that emerges from the problem of ‘time inconsistency’ – that is, even benevolent monetary authorities want to pursue inﬂationary policies, and this incentive is robust to constituents perfectly anticipating the authorities’ actions. Indeed, bargaining positions can change over time (even if agency and informational problems are absent) and thus ultimately render individual incentives ‘inconsistent’ with socially optimal policies. Economic performance can thus improve when the organization of monetary authorities mitigates this inconsistency or increases the cost of acting on it. Notice that this normative inference is remarkably similar to that in Part II – that is, maximizers can enjoy an expanded set of economic opportunities by constraining themselves from acting in an opportunistic manner. We’ll build on this robust inference in Part IV to consider how societies do and should organize their central banking services.

6.

Money in a classical economy

THE CLASSICAL ECONOMY’S SUPPLY SIDE The Firm’s Problem and Labor Demand Our investigation of how money relates to economic ﬂuctuations starts with a ‘classical’ model of the macroeconomy – that is, a model that ﬁrmly grounds itself on the axiom of maximizing-behavior while setting aside potentially important economic frictions. This approach will let us see how far a simple model goes in rationalizing business cycles, and highlight empirical regularities that deserve further attention. Our classical model develops macro-insights from micro-foundations by deﬁning a ﬁrm’s (real) proﬁt as follows: f(L) (wP) L where f transforms labor (L) into output, w denotes a nominal wage for which individuals can trade leisure in the labor market, and P denotes the price level of ﬁnal goods and services. Before moving forward, we should understand some of this representation’s details. First, the ﬁrm’s technology can reasonably be characterized as transforming labor into output at a positive but decreasing rate. Stated more formally, fL 0 and 2fL2 0. This latter condition appreciates the diminishing rate at which ﬁxed capital stocks can augment the productivity of marginal labor units (as well as those units’ contribution to congestion costs). In addition, we’ll deﬁne the economy’s price level P as an ‘average’ nominal price for each good and service that an economy produces – for example, the ‘consumer price index,’ or CPI, that popular and business media frequently report. Finally, we’ll assume (and ultimately derive the hypothesis) that input and output prices move together (on average). This assumption implies that changes to the economy’s price level P leave the relative price of labor (w/P) unchanged.1 Our axiom of maximizing-behavior thus suggests that ﬁrms can be modeled as choosing units of labor to maximize proﬁts as follows:

53

54

Money in a dynamic economy

Goods Y*

f(L); f' > 0,f" < 0 Maximum distance between f(L) and (w/P)L

(w/P)L

L

Profit

f(L) – (w/P)L Profit/L > 0

L*

Profit/L < 0

L

Profit/L = 0 Figure 6.1

The ﬁrm’s proﬁt-maximization problem

max f(L) w P L L

Before solving this problem analytically, let’s develop our economic intuition by considering its graphical representation in Figure 6.1. Notice that, given the production function f’s curvature (that is, the rate at which ﬁrms can transform inputs to outputs, and changes therein), we can illustrate the ﬁrm’s problem as follows. The ﬁrm’s optimal choice of labor, L*, equates labor’s marginal product and cost. Evaluated at any other choice of labor, the ‘gap’ between output f(L) and labor’s share of output (w/P)L is less than the maximum feasible diﬀerence. Consider, for example, any LL*. Evaluated at these levels of labor, the rate at which f(L) increases exceeds (w/P). Given our axiom of maximizing-behavior, ﬁrms will not rest at such choices since labor’s marginal cost falls short of its marginal product.2 Analytically, we can deﬁne this optimal choice as having to satisfy the following ‘ﬁrst order condition.’ f w L P

55

Money in a classical economy

Goods

(w/P)' w/P

Ld L' Figure 6.2

L*

L

The labor demand curve

This expression comes from diﬀerentiating the ﬁrm’s objective and recognizing that, evaluated at a ‘maximand’ (a choice that fulﬁlls the objective), the objective function’s value cannot change. We’ve thus conﬁrmed our graphical intuition that, presented with any real wage (w/P), proﬁt maximizing ﬁrms will employ labor to satisfy the above necessary ﬁrst order condition. We can now develop the labor demand relationship – that is, the relationship between labor’s price and quantity demanded. This relationship plays an important role in determining the level of output at which our model economy will rest, and we’ll refer to this relationship’s graph as the ‘labor demand curve’ – that is, the quantity of labor that a ﬁrm willingly employs at any given real wage (w/P). Figure 6.2 illustrates this relationship. To see why the demand curve for labor slopes downward, note that the ﬁrm’s optimal choice of labor L* must satisfy the above ﬁrst order condition for any real wage (w/P). We can thus examine how the ﬁrm’s optimal labor choice changes with the real wage as follows: f(L* ( wP )) wP 0

56

Money in a dynamic economy

This expression simply says that an optimal choice of labor must satisfy our necessary ﬁrst order condition for any real wage. To see how the optimal labor choice relates to the real wage, then, we can diﬀerentiate the above expression as follows: 2f L* 10 L2 ( wP ) L* 1 0 ( wP ) (2fL2 ) Since 2fL2 0, the (representative) ﬁrm’s labor demand curve must slope downward. Equilibrium in the Labor Market Households constitute the labor market’s supply side, and we can formalize their objective as being symmetric to that of ﬁrms – that is, for any real wage (w/P), supply a quantity of labor that maximizes utility. Given this symmetry, we should expect the labor supply curve to slope upward. Indeed, assuming that the incentive to ‘substitute’ labor for leisure as real wages increase overwhelms the incentive to consume more leisure as ‘wealth’ increases, satisfying this objective requires price-taking households to increase the quantity of labor supplied when real wages increase.3 Collecting both the labor supply and demand curves in one ﬁgure, equilibrium in the labor market occurs where the labor demand and supply schedules intersect – that is, at (LE, (w/P)E) in Figure 6.3. To see why an economy of maximizers rests at this intersection, notice that any other candidate is either infeasible or leaves open mutually beneﬁcial trades. Consider, for example, a real wage that exceeds the equilibrium level depicted above (that is, (w/P) (w/P)E). Evaluated at these levels, households exist that are willing to supply labor at a lower real wage, and ﬁrms exist that are willing to employ them. But if households are maximizing utility, and ﬁrms are maximizing proﬁts, then absent bargaining frictions, the economy’s dynamics cannot stop until they exhaust all such mutually beneﬁcial trades. The Classical Aggregate Supply Curve Equilibrium in the labor market thus determines two important macroeconomic variables – that is, the levels of real wages (w/P)E and employment LE. This equilibrium, in turn, ﬁxes the quantity of goods at which our

57

Money in a classical economy

w/P Ls

(w/P)E

Ld LE Figure 6.3

L

Equilibrium in the labor market

model economy must rest. To see this last implication, notice that when each ﬁrm employs its share of equilibrium labor, LE, ﬁrms aggregately supply m . f(LE/m)F(LE) ‘goods’ to our model economy (where our economy consists of m representative ﬁrms). That aggregate supply (AS) cannot change with the price level P follows from our assumption that input and output prices move together. Notice that changes in the price level P do not change the real wage, and thus do not inﬂuence the quantity demanded or supplied of labor. Moreover, changing the price level does not aﬀect the ﬁrm’s technology or the household’s preferences, and thus leaves the labor demand and supply schedules unaltered. Because changes in P do not aﬀect the labor market’s equilibrium, and output is a function of labor, the classical economy’s aggregate supply curve must thus be vertical at YE as shown in Figure 6.4.

THE CLASSICAL ECONOMY’S DEMAND SIDE We now see how equilibrium in the labor market, coupled with the assumption that nominal rigidities are negligible, determines several important economic variables – that is, equilibrium employment LE, real

58

Money in a dynamic economy

P AS

YE Figure 6.4

Y

Classical (long-run) aggregate supply

wages (w/P)E and output YE. But while we can draw inference about the relative price of labor (w/P), we haven’t said anything about the economy’s price level P. Moreover, our model so far ignores money. To address these issues, we need to develop the demand side of the classical economy’s goods market. Velocity and the Quantity Theory of Money The ‘quantity theory of money’ will aid our thinking about the economy’s demand side. An important ingredient to this theory is the ‘velocity’ of money – that is, the average number of times that a monetary unit (for example, a dollar) turns over to purchase an economy’s ﬁnal goods and services. To make this idea concrete, let M denote the (nominal) stock of money, Y the real aggregate output in one year, and P a price index. Then money’s velocity can be expressed as the ratio of nominal aggregate income during a period to that period’s money stock: V PY M The ‘equation of exchange’ follows immediately from this deﬁnition. It says that aggregate nominal income equals the money stock times money’s average turnover rate:

Money in a classical economy

59

MVPY Notice that this equation is really an accounting identity – that is, by deﬁnition, an economy’s money stock must turn over V times to purchase all ﬁnal goods and services. To construct a theory of aggregate demand, we need to say something about the forces that independently shape these variables. The ‘quantity theory’ lets us push forward in this direction. The classical quantity theory posits that an economy’s institutional features (for example, ATM machines, credit cards, check-writing facilities, and so on) determine money’s velocity.4 Moreover, because the theory assumes that these institutions evolve gradually, it implies that velocity slowly changes and can thus be treated as more or less constant in the short run. Equipped with this ingredient (that is, a process by which one of our variables is pre-determined), we can move forward in developing a ‘theory’ of money demand. Notice that, if institutional features determine velocity, and these features evolve slowly, then money demand can be expressed as a constant multiple of nominal income: Md k·PY where k 1/V. The Classical Aggregate Demand Curve The aggregate demand (AD) curve maps the levels of aggregate output Y that are demanded at each price level P. We can derive this relationship from the monetarists’ classical quantity theory by mapping the points that satisfy the ‘equation of exchange’ for a given velocity and quantity of money (and assuming that money markets equilibrate – that is, Ms Md). The resulting AD curve has the property that as the price level falls, ceteris paribus, the quantity demanded of real output increases. We illustrate this relationship in Figure 6.5. To see why this relationship holds (given our assumptions), consider the case of a price drop (P ↓) and notice what has to happen to the quantities demanded for the equation of exchange to hold. In this case, real money holdings (Md P) increase (at least initially) above the level that households held (in aggregate) before the price drop. But before the price drop, maximizing households must have been holding an optimal level of real cash balances – that is, the level at which cash balances’ marginal cost (for example, forgone interest) equaled the associated marginal beneﬁt (for example, transaction services). Immediately after the price drop,

60

Money in a dynamic economy

P

AD Y Figure 6.5

Classical aggregate demand

households must thus be holding too high a level of real balances (if marginal costs increase and marginal beneﬁts decrease in real balances). The price drop thus encourages households to reduce their real balances by trading them for goods – that is, increasing the quantity of goods demanded above what the original price level warranted.5

EQUILIBRIUM IN THE CLASSICAL ECONOMY Recall that the labor market determines equilibrium employment LE and the real wage (w/P)E (that is, the rate at which households can trade leisure for commodities). This employment level, coupled with the economy’s production technology, determines output in the goods market. And since nominal rigidities are negligible in the classical theory (that is, wages and output prices move together on average), this output level obtains for any price level at which the economy might rest. As a result, aggregate demand (as represented by the AD curve) determines the classical economy’s price level P for a given level of equilibrium output YE, but cannot inﬂuence the level of equilibrium output. Figure 6.6 illustrates these relationships.

61

Money in a classical economy

P AS

PE AD

YE Figure 6.6

Y

Goods market equilibrium

MONEY, PRICES AND REAL ACTIVITY Money and Inﬂation We should now understand how the classical economy comes to rest at particular levels of real wages, employment, output and prices. Ultimately, however, we are interested in how monetary policy aﬀects these variables. Let’s move in that direction by examining how changing the supply of money inﬂuences the price level. Recall that, in the monetarist approach, the AD curve relates Y and P for given levels of M and V. Since this model treats V as a constant, it points to the economy’s stock of money M as the primary demand shifter. Figure 6.7 illustrates this inﬂuence. The price level and quantity of aggregate demand share a negative relationship for any stock of money. For a given P, however, expanding the money stock increases the quantity of aggregate demand – that is, holding P and V constant, Y must increase with M. Moreover, because increasing M increases the quantity of aggregate demand for any P, the AD curve must shift outward – for example, from AD(M0) to AD(M1) in Figure 6.7, where M1 M0. Changing the money supply thus inﬂuences the evolution of the classical economy’s price level.6

62

Money in a dynamic economy

P

AS

P(1) AD(M1) P(0) AD(M0)

YE Figure 6.7

Y

Goods market response to an increase in the money supply

Money and Real Activity While money aﬀects the classical economy’s price level, it leaves output unchanged. In other words, money exerts a nominal eﬀect, but is ‘neutral’ with respect to economic opportunities. This neutrality implies that ﬂuctuations in aggregate output come from ‘technology shocks’ – that is, changes to the economy’s production possibilities. Figures 6.8, 6.9, 6.10 illustrate how such shocks create ‘real business cycles’ (RBC), and why the classical model implies that monetary policy is an ineﬀective tool for dampening ﬂuctuations. In Figure 6.8, a (negative) technology shock shrinks the ﬁrm’s production possibilities from those that were available under f(L) to those that remain under g(L). Such a shock can emerge, for example, from a natural disaster (for example, a hurricane) that diminishes an economy’s (physical) capital stock or constrains the supply of complementary inputs (for example, oil). In cases like these, the loss of capital or complementary inputs reduces the productivity of each labor unit. Confronted with a decrease in marginal product, maximizing ﬁrms reduce their willingness to employ labor. To see this implication, recall that L* maximizes proﬁts for a given real wage (w/P) under the technology f(L). But since labor’s marginal product is everywhere less under g(L) than under f(L), L* cannot maximize proﬁts after the shock. Indeed, marginal productivity under g(L) must be less than (w/P) when evaluated at L*. Firms thus

63

Money in a classical economy

Goods

f(L) g(L)

(w/P)L

L

Profit

L'

L*

f(L) – (w/P)L g(L) – (w/P)L L

Figure 6.8 A negative technology shock, and its inﬂuence on the quantity of labor demanded for any real wage optimally respond to negative shocks by scaling back employment until the marginal productivity of labor again equals the marginal real wage – for example, at L'L* in Figure 6.8. Since this response occurs for an arbitrary real wage, it must occur for any real wage. A negative technology shock thus causes the labor demand curve to shift inward, and decreases the equilibrium level of labor employed. This decrease, coupled with the inferiority of the post-shock production technology, then causes a leftward shift in the economy’s aggregate supply curve. Figures 6.9 and 6.10 illustrate how a negative technology shock ultimately inﬂuences the markets for labor and goods/services in this manner. A Case Against Activist Policy Figures 6.9 and 6.10 also highlight an important normative implication of the classical model – that is, because the economy perpetually equilibrates, monetary policy cannot improve upon the level of welfare at any point in the business cycle. Indeed, as Figure 6.10 shows, increasing aggregate demand in the face of a negative supply shock exacerbates the shock’s inﬂuence on the price level but does nothing to mitigate the shock’s real eﬀects.7

64

Money in a dynamic economy

w/P Ls

(w/P)E (w/P)'E

L'd L'E Figure 6.9

Ld L

LE

Labor market reaction to a negative technology shock

P AS'

AS

AD(M1)

AD(M0) Y'E Figure 6.10

YE

Y

Goods market reaction to a negative technology shock

Money in a classical economy

65

To the extent that the classical model accurately portrays the ‘true’ economy, it suggests that a ‘passive’ monetary authority can do better than can an ‘activist’ authority (where ‘better’ is measured relative to real economic opportunities). It thus argues for organizing monetary authorities in a manner that weakens pressures that might encourage political agents to ‘ﬁne tune’ the economy.8 Before extending our treatment of how monetary authorities are and should be organized, however, we should learn how Keynesian models rationalize ﬂuctuations in economic aggregates. These models hold out a greater prospect for activist policy to improve welfare. Understanding their merits can thus facilitate greater conﬁdence in the normative prescriptions that we’ll soon develop for governing an economy’s money supply.

NOTES 1. We’ll soon see that this last assumption is pivotal to the classical model’s normative implications, and thus pay considerable attention to how an economy might behave when nominal prices are not so ﬂuid. 2. A symmetric argument shows that ﬁrms will not rest at a choice of labor that exceeds L*. 3. Given the previous sub-section’s detail, we leave this requirement’s formal derivation to our readers. 4. Notice, for example, that as the supply of ATMs increases, any quantity of money (M) can sustain the purchases of an increasing level of goods and services (PY). 5. Notice that this argument diﬀers from its analog in a model of the microeconomy. There, as the price of a particular good increases, so does the opportunity cost of consuming that good. In addition, a price increase shrinks an individual’s budget set. For normal goods, these substitution and income eﬀects work together to decrease the quantity demanded. In our model of the macroeconomy, however, an increase in P represents an increase in the price level, not an increase in relative prices. Consequently, the aggregate demand curve slopes downward not from substitution and income eﬀects, but because changes in P change the level of nominal balances that households are willing to hold. 6. Even more, if we deﬁne ‘inﬂation’ as a continual increase in the price level, then ‘inﬂation is always and everywhere a monetary phenomenon’ (Friedman, 1992: 193). While supply shocks and changes in government expenditure may have transitory eﬀects on the price level, they cannot, by themselves, continually increase prices. Both the monetarist theory developed here and the Keynesian counterpart developed below thus agree that inﬂation comes from continually growing the money supply. 7. Nobel Prize-winning economists Finn Kydland and Edward Prescott are prominent contributors to the RBC literature in this regard, forcefully arguing that, because technology shocks drive economic systems, policies that aim to dampen these shocks tend to be counterproductive (see Rolnick, 1996). 8. This organizational implication is similar to what we developed in Part II. There, we argued that a passive monetary authority can facilitate conﬁdence in ﬁat exchange media, and thus expand economic opportunities by economizing of bartering costs.

7.

Money in a Keynesian economy

It’s not a moving equilibrium. When we have 11 percent unemployment in 1982 or 25 percent in 1932, I don’t regard that as being a labor market equilibrium with supply and demand equal. And I don’t believe that productivity, technology, go up and down in anything like waves which would be consistent with the business cycles we observe. (James Tobin, quoted by Fettig, 1996)

INTRODUCTION The classical model oﬀers a ﬁrmly grounded rationalization of important macroeconomic phenomena such as inﬂation (especially hyper-inﬂation), and even ﬂuctuations in real output. Indeed, prominent contributors to the RBC (real business cycles) literature argue that technology shocks explain 70 percent of the post-war (WWII) economy’s ﬂuctuations (see Rolnick, 1996 and Romer, 2001, Table 4.4). But this model encounters diﬃculty when attempting to rationalize phenomena such as involuntary unemployment, and ﬂuctuations therein. In addition, persistent reduced form evidence suggests that, at least in the short run, money is not neutral.1 The Great Depression shed a bright light on these diﬃculties, encouraging that era’s most prominent economist, John Maynard Keynes, to consider how aggregate demand can aﬀect employment and output. A distinguishing feature of these early models is an allowance for nominal rigidities. Recall the classical assumption that constituent prices move together (on average), and its implication that aggregate supply does not vary with the price level but remains ﬁxed at a level consistent with the labor market’s competitive equilibrium. By instead assuming that prices are ‘sticky,’ conventional Keynesian models imply that the (shortrun) aggregate supply curve slopes upward, and thus open the door for activist monetary (and ﬁscal) policy to productively address economic ﬂuctuations.

66

67

Money in a Keynesian economy

THE KEYNESIAN ECONOMY’S SUPPLY SIDE Nominal Rigidities, the Phillips Curve and Policy Eﬀectiveness Suppose that, instead of tracking the price level (P), wages (w) do not move in proportion to output prices. This ‘rigidity’ or ‘stickiness’ might emerge from laborers (formally or informally) entering inﬂexible wage contracts. For any ‘contracted’ nominal wage w, increasing the price level P (say by expanding the money supply M) would decrease the real price of labor w/P. Faced with a lower real price of labor, ﬁrms increase the quantity of labor demanded (that is, ﬁrms move down their labor demand curves).2 To the extent that households are willing to supply labor at lower real wages, the level of labor that an economy employs will increase with P, and the aggregate supply curve will slope upward per Figure 7.1.3 Notice that, if the aggregate supply curve slopes upward (for example, because wages are sticky), then policy makers might trade oﬀ higher inﬂation for lower unemployment. As in the classical model, expanding the money supply continues to increase the price level. But instead of being neutral in real terms, this increase increases the quantity of labor that ﬁrms demand and thus increases employment. Early post-war economists found considerable empirical support for this implication. Indeed, Alban W. Phillips famously graphed the relatively stable negative relationship that, prior to 1970, appeared between P AS

Y Figure 7.1

Keynesian (short-run) aggregate supply

68

Money in a dynamic economy

unemployment and inﬂation. Coupled with rationalizations like that described above, the ‘Phillips Curve’ encouraged monetary authorities to actively pursue full employment. Rational Expectations and the Phillips Curve’s Breakdown While the Phillips Curve enjoyed considerable popularity (and does so even today), it also encounters important diﬃculties. For example, notice that a demand-driven ‘economic boom’ should increase the price level, and thus reduce real wages. But during economic ﬂuctuations in general, and the Great Depression in particular, real wages are pro-cyclical – they decrease with economic slowdowns. Moreover, even Keynes realized that assuming that workers (mistakenly, but persistently) focus on nominal wages could not logically produce an upward sloping supply curve (Blanchard, 1990). Before we consider more ﬁrmly grounded arguments of why the aggregate supply curve might slope upward, let’s turn our attention to how maximizing-individuals form expectations, and how this formation creates diﬃculty for the sticky wage channel outlined above. To the extent that inferential mistakes are costly, maximizers avoid systematic errors in judgment. In other words, maximizers optimally collect and evaluate information so that expectations are ‘rational,’ or correct on average.4 Rational laborers, for example, might incorrectly forecast changes in the price level, but those errors would not exhibit bias in a particular direction (unless the costs of mistakes are asymmetric). In this light, the ‘nominal rigidity’ channel for a positively sloped aggregate supply curve appears to rest on a weak ground.

BOX 7.1

THE EFFICIENT MARKET HYPOTHESIS

Models that characterize expectations as being ‘rational’ appear to be ‘good’ by the standards we developed in Part I. Indeed, the rational expectations hypothesis builds from the axiom of maximizingbehavior and finds considerable empirical support, especially from financial market data. Applied to the pricing of financial assets, rational expectations theory is known as the ‘Efficient Markets Hypothesis’ (EMH). Its testable implications include the hypothesis that (risk adjusted) forecast errors equal zero on average – that is, they are unpredictable for a given set of information.

Money in a Keynesian economy

69

To see this implication, suppose that available information tells investors that security prices will increase, but prices have not yet incorporated this knowledge. Investors would thus profit from buying a security today and selling it after realizing the expected price increase. But for such opportunities to be available in the first place, investors must have been ignoring a ‘sure thing.’ In other words, for security prices to reflect something less than the set of information on which individuals can feasibly transact, investors cannot be maximizers! We thus see that the rational expectations hypothesis is another logical implication of our axiom of maximizing-behavior. The extent to which the EMH exhibits empirical merit thus depends on how reasonable and complete is our axiom. Recall that this reasonableness does not rely on individuals consciously maximizing, but rather on how strong are the forces that adopt individuals who act ‘as if’ they are maximizers. In light of how competitive are financial markets, we may thus be relatively comfortable with building asset pricing models from the axiom of maximizing-behavior. (Andrei Shleifer, 2000 critically reviews this evidence, but does not firmly ground the preferences that might give rise to behavioral anomalies. Robert Frank, 2006 offers a textbook treatment of this latter issue.)

By pointing to this weakness, the rational expectations hypothesis increased skepticism among economists such as Milton Friedman and Edmund Phelps about the policy trade-oﬀs that the Phillips Curve implies are feasible.5 Indeed, Friedman used his presidential address to the American Economic Association to predict the Phillips Curve’s collapse and oﬀer the ‘expectations-augmented’ Phillips Curve in its place. The traditional Phillips Curve follows Keynes’ early conjecture that workers focus on nominal wages rather than real purchasing power. But if individuals are maximizers, then they rationally form expectations. In this case, suppliers of labor correctly anticipate (on average) changes in the price level so that, when these changes are realized, real wages tend to remain steady. Friedman thus argued that only unanticipated inﬂation can cause unemployment to deviate from its natural rate (that is, the rate at which competitive labor markets equilibrate). The ‘expectations-augmented’ Phillips Curve formally characterizes this relationship as follows:6 (Y Yn ) ( e ) s

70

Money in a dynamic economy

This more ﬁrmly grounded model implies that attempts to inﬂate one’s way to full employment ultimately result in high inﬂation and no employment beneﬁts (that is, ‘stagﬂation’). In particular, it implies that Y increases only when individuals experience surprise-inﬂation – that is, when ( e ) 0. Moreover, to the extent that expectations improve as individuals learn from new information, any such surprises are unlikely to create long-lasting eﬀects. As workers pressure ﬁrms to maintain real wages, for example, the AS curve would shift leftward and oﬀset any employment eﬀect from surprise increases in the money supply. Summary The Phillips curve was absorbed quickly in econometric models, in policy discussions, and in policy decisions. The notion of a stable and reliable ‘trade-oﬀ’ between inﬂation and unemployment made the journey from professional literature to the counsels of policy and even the Councils of Economic Advisers more rapidly than most new ideas. Sociologists or historians of science may one day explain why economists and oﬃcials accepted the Phillips curve so readily. Economists are more often attracted by theories without evidence than by evidence lacking a clear relation to theory. (Brunner and Meltzer, 1976: 2, emphasis added)

We should now understand that, if nominal rigidities are considerable, then the aggregate supply curve can slope upward and money can be nonneutral (at least in the short run). We should also recall from our introduction to methodology (see Part I) that when models ground themselves on a weak foundation, they are unlikely to produce valid insights. By apparently wrapping the wrong structure around persuasive reduced form evidence of how M relates to Y, many post-war economists may have realized this problem. The Phillips Curve exhibited considerable power to rationalize the 1960s data, but appears to rest on a weak foundation – that is, one that ignores evolving expectations and the strategic responses of maximizers. Consistent with this criticism, the Phillips Curve relationship broke down when policy makers attempted to exploit it. The 1970s instead saw ‘stagﬂation’ – that is, high inﬂation and unemployment.

EFFICIENCY WAGE THEORIES To maintain a consistent analytical framework, we cannot simply give up on the axiom of maximizing-behavior whenever it creates diﬃculty for rationalizing observed phenomena. For example, if individuals are indeed maximizers, then we cannot readily rationalize a supply curve’s upward slope via the simplistic nominal rigidity story that we criticized above.

Money in a Keynesian economy

71

We may nevertheless be curious about whether nominal rigidities can arise without violating axioms of rationality. ‘Eﬃciency wage’ theories show that they can. Blanchard and Fischer characterize this class of theories as follows: Eﬃciency wage theories all start from the assumption that productivity may be aﬀected by the wage the ﬁrm pays, though the precise mechanism through which the wage aﬀects productivity varies from theory to theory. When workers’ eﬃciency is aﬀected by the wage, a reduction in the wage may in the end increase rather than decrease cost. (Blanchard and Fischer, 1989: 455)

Recall that the ‘classical ﬁrm’ chooses factors to maximize proﬁts, given real factor prices – that is, factor prices do not aﬀect factor productivity. Extending that model in the present section, we’ll relax the implicit assumption that real wages do not aﬀect the eﬀort-choice of laborers, and thus expand the ﬁrm’s problem to include the choice of wages – that is, ﬁrms in this section choose L and w. Working through this more general model, we’ll ﬁnd that ﬁrms optimally choose a wage above that which clears the classical labor market. At this relatively high wage, the quantity of labor that households supply exceeds that which ﬁrms demand – that is, strictly positive unemployment occurs in equilibrium. This rationalization of strictly positive equilibrium unemployment also oﬀers a more ﬁrmly grounded rationalization of why nominal variables (for example, w) might be rigid. A Richer Proﬁt Maximization Problem Our formal investigation begins by specifying the ﬁrm’s problem as choosing labor and nominal wages to maximize proﬁts – that is, it characterizes ﬁrms as choosing w and L to maximize the following objective:

max f(e( wP )L) w P L w, L

This objective diﬀers from what we examined in the classical model. Importantly, f now transforms ‘eﬃciency units’ of labor into output at a positive but decreasing rate (that is, f '0, f "0). Here, e transforms (at a positive but decreasing rate) each real wage that the ﬁrm indirectly chooses into ‘eﬀort’ that each unit of labor optimally exerts. Output thus becomes a function not only of the quantity of labor units employed (L), but also of how much eﬀort (e) each unit exerts, and that eﬀort depends on the ﬁrm’s wage choice. Optimal choices of w and L must simultaneously satisfy the following ﬁrst order conditions:

72

Money in a dynamic economy

f '( · ) e'( wP ) (LP) (LP) 0 and f '( · ) e( wP ) (wP) 0 respectively. We can re-write the ﬁrst condition as f '( · ) e'( wP ) (LP) (LP) f '( · ) e'( wP ) 1 f '( · ) 1 e'( wP ) and the second as f '( · ) (wP) e( wP ) Recognizing that the ﬁrm’s optimal choice of L and w must simultaneously satisfy these conditions, we can substitute the necessary condition for L into that for w and arrive at the following (Solow) condition: 1 e'( wP ) (wP) e( wP ) e( · ) (wP) 1 (wP) e( wP ) In words, ﬁrms choose a nominal wage such that the elasticity of eﬀort with respect to the real wage equals 1 (that is, eﬀort changes proportionately with the real wage). Given this wage, the necessary condition for choosing L determines the ﬁrm’s choice of (raw) labor units. And since the ﬁrm’s optimal choice of w need not coincide with what would have cleared the classical labor market, the ﬁrm’s choice of L need not clear the labor market.7 Eﬃciency Wages and Involuntary Unemployment This ﬁrmly grounded, but richer, model of the ﬁrm implies that strictly positive and involuntary unemployment can persist in an economy. To see this implication more clearly, let’s examine Figure 7.2.8 Suppose that individuals face a ‘dichotomous’ choice of eﬀort (denoted here as e) where the choice set includes ‘no shirking’ (e 1) and ‘shirking’ (e 0). Recall that utility-maximizing households forgo leisure for labor market opportunities until the marginal cost of doing so (for example, disutility of eﬀort) equals the associated marginal beneﬁt (for example,

73

Money in a Keynesian economy

(w/P)

No shirking constraint Ls

(w/P)E

f'(L) = Ld

1

LE Figure 7.2

N

Individuals

Involuntary unemployment with eﬃciency wages

utility from expanded consumption possibilities). Given our characterization of eﬀort as taking one of two levels, and further assuming that the cost of exerting such eﬀort is uniform across households, we can represent the quantity of labor that households are willing to supply at each level of the real wage (w/P) by the vertical line labeled Ls.9 This curve says that, for any real wage that exceeds the marginal cost of eﬀort (that is, (w/P) 1), households willingly choose e 1. As a consequence, for any (w/P) 1, all N of our economy’s homogeneous households oﬀer their labor services to ﬁrms. To see why ﬁrms must nevertheless oﬀer a nominal wage above that which compensates households for choosing e1, consider what happens if they do not. When monitoring is costly, a positive probability exists that ﬁrms will not detect their employees’ shirking. And employees who escape this detection will be really well-oﬀ – that is, they receive a strictly positive (real) wage payment, but do not incur any eﬀort cost. More formally, employees who shirk but are not caught receive a net payoﬀ of (w/P)0, rather than the zero payoﬀ ((w/P) – e0) that they would have realized by not shirking. When the ﬁrm catches a shirker, however, it does not remunerate the employee – since the shirker didn’t exert any eﬀort, its payoﬀ nets to zero. So, if employees always choose e 1, then they certainly receive a net payoﬀ of zero. But, if employees always shirk, then they enjoy an expected payoﬀ that strictly exceeds zero (to the extent that verifying eﬀort is costly).

74

Money in a dynamic economy

Absent an eﬃciency wage, employees always do at least as well shirking as they do not shirking. An interesting question, then, is how much greater than (w/P)1 must the ﬁrm raise its wage? Figure 7.2 illustrates the locus of all points that induce no shirking by the ‘No shirking constraint’ curve. Notice that, given our previous paragraph’s argument, the no shirking constraint must lie everywhere above the level of (w/P) 1. Moreover, as the economy’s employment level reaches its potential (that is, as L → N), the wage that ﬁrms must pay to guarantee no shirking increases. To see why, notice that the unemployment rate decreases as L → N, and thus so does the expected duration of unemployment for those who are caught shirking. In this manner, increases in the unemployment rate decrease the net beneﬁts of shirking. The eﬃciency wage that is necessary to discourage shirking thus decreases as (N – L) grows. Finally, let’s consider the ﬁrm’s labor demand schedule. Recall that the necessary condition for proﬁt maximization requires the ﬁrm to choose L so that labor’s marginal product ( f '(L) ) equals its marginal (real) cost (w/P). We graph all such points in Figure 7.2 via the curve Ld. Given this set-up, we see that the labor market equilibrates at less than ‘full employment.’ Indeed, at the equilibrium real wage (w/P)E, N individuals oﬀer their labor services to ﬁrms, but only LE N ﬁnd employment (that is, the unemployment rate equals (N LE )N .)10 Note, however, that since the real wage is no longer exogenous, we cannot employ our illustration of the classical labor market to see how unemployment might obtain in equilibrium (it is nonsensical to evaluate how the ﬁrm’s optimal choice of labor responds to its optimal choice of eﬃciency wages – these two choices are determined simultaneously). Rather, we have to assume that ﬁrms set wages subject to the ‘no shirking constraint’and then ask at what employment level does the marginal product of ‘non-shirkers’ equal the associated marginal cost (that is, the ‘eﬃciency wage’). And since the eﬃciency wage exceeds that which fully remunerates a laborer’s eﬀort, the equilibrium level of employment must be less than that obtained in the classical model (where laborers did not enjoy an informational advantage and thus optimally oﬀered labor to the market at a price equal to the marginal cost of eﬀort).11 Micro-based Theories of Nominal Rigidities This eﬃciency wage model advances our simple Keynesian model by oﬀering a ﬁrmly grounded rationalization of involuntary unemployment. It does not, however, yield a positively sloped aggregate supply curve – that is, eﬃciency wage theory does not (by itself) say that nominal wages are sticky, but rather that eﬃciency wages do not clear labor markets.12

Money in a Keynesian economy

75

However, it is possible to move from our rationality-based explanation of ‘involuntary unemployment’ to a nearly rational explanation of cyclical unemployment. Perhaps the most authoritative research in this regard is that of Akerlof and Yellen (1985). This work develops from an application of the ‘envelope theorem’ – that is, evaluated in the neighborhood of an optimally chosen eﬃciency wage, the proﬁt function is ﬂat. In other words, ﬁrms that do not optimally respond to aggregate demand shocks (by adjusting their choice of eﬃciency wage) face little in the way of negative selection pressures – that is, they will still act as if they are ‘nearly’ rational.13 Akerlof and Yellen (1985) formalize this notion by showing that the loss to ‘inertial’ ﬁrms from not optimally responding to an -small shock to the money supply is proportional to 2, but the economy experiences employment losses (and thus output losses) proportional to . To see why this must be true, consider the case where all ﬁrms are inertial – that is, no one reoptimizes in the face of a change in M. Here, while each ﬁrm remains near the maximum level of its objective, the response in output must be proportional to the size of the shock (that is, changes in M must pass through to Y in our equation of exchange when P is rigid). In this setting, small departures from maximizing-behavior can let demand shocks move the economy from its classical equilibrium. This implication follows from proﬁts in eﬃciency wage models being rather robust to nominal rigidities, whereas employment is not. Hall (2003) explores further in this direction. His objective, like that of Akerlof and Yellen (1985), is to rationalize how small nominal rigidities might induce large (real) economic ﬂuctuations. Rather than relying on departures from rationality to motivate rigidities, however, Hall exploits the condition that market clearing wages are elements of a (non-singleton) ‘bargaining set’ – that is, a non-degenerate range of wages that satisfy both households and ﬁrms. While ﬁrms would proﬁtably hire units of labor at any element of the bargaining set, the surplus that they recognize therein will decrease as realizations of that bargaining process move toward the set’s upper bound. Moreover, the intensity with which ﬁrms rationally pursue recruiting activities depends positively on this surplus’s size. Consequently, because small productivity decreases or input cost increases essentially move bargaining realizations toward the set’s upper bound, ﬁrms optimally respond by curbing their recruiting activities. As a result, labor market frictions increase (that is, costs for qualiﬁed laborers and ﬁrms to ﬁnd each other grow), causing involuntary unemployment to grow as well. From a purely ‘rational’ model, then, Hall (2003) oﬀers an explanation for why we observe cyclical involuntary unemployment.14

76

Money in a dynamic economy

EVIDENCE AGAINST NOMINAL RIGIDITIES For money to aﬀect real activity in a model with nominal rigidities, the period over which prices remain sticky must be longer than what is necessary for monetary authorities to recognize that a policy change would be beneﬁcial, craft and implement that change and have that implementation work its way through the economy. An important question thus becomes for how long can we expect prices to be rigid? Peter Klenow and various co-authors (mainly, Mark Bils) address this question by analyzing data on how frequently diﬀerent prices in an economy change.15 The resulting evidence creates potentially serious diﬃculties for models of money and real activity where the causal structure is nominal rigidities. First, these authors ﬁnd that prices change about three times per year on average – that is, more frequently than a theory of nominal rigidities might predict. Second, the frequency with which individual prices change appears to vary systematically across sectors. For example, while goods prices appear to change every 3.2 months, those for services change much less frequently (every 7.8 months). More generally, goods and services that embody little in the way of value added (for example, gasoline, fresh fruit) exhibit considerably more price ﬂexibility than do their more highly processed counterparts. Indeed, prices for ‘raw goods’ change every 1.6 months and those for ‘processed goods’ change every 5.7 months. But, for monetary policy to aﬀect real activity via the nominal rigidities channel, input prices (that is, prices for ‘raw goods’) need to be rigid! Stock and Watson (1999) review what can be considered corroborating evidence. They argue, for example, that ‘real wages have no predictive content for output growth at the one- or four-quarter horizons’ (Stock and Watson, 1999: 42). In this light, while the microeconomic theories of sticky (nominal) wages can oﬀer ﬁrmly grounded rationalization of why aggregate demand might inﬂuence economic performance, they nevertheless appear to lack empirical merit. Indeed, to the extent that nominal rigidities (micro-based or not) cause cyclical unemployment, we should observe a rather strong and counter-cyclical relationship between real wages and output.

OTHER MONETARY TRANSMISSION MECHANISMS In the previous subsection, we carefully developed causal channels through which money might inﬂuence economic activity. But while these models appear sound in the sense that they build on the axiom of maximizing-

Money in a Keynesian economy

77

behavior, we learned that their implications leave unexplained (or are incongruent with) important features of the relevant data. Let’s now recall our evaluation of structural versus reduced form evidence. Notice that our micro-based models of nominal rigidities oﬀer structural explanations of how money might inﬂuence an economy’s real aggregates. That the evidence does not support such explanations, however, does not imply that money is neutral. Indeed, money might exert a real eﬀect and we’ve simply looked at evidence through the wrong structure. We thus push forward by looking at reduced form evidence of money’s potential to inﬂuence an economy’s real aggregates. Friedman and Schwartz’s (1963) historical ‘experiments’ are frequently cited as being persuasive in this regard. In short, they carefully identify episodes in US monetary history where changes in the money supply can conﬁdently be interpreted as random. They then look at what happens to output and ﬁnd that it tends to move in the same direction. Contemporary statistical studies largely corroborate Friedman and Schwartz’s evidence that money exerts a gross eﬀect on output. However, we’ll see again that nominal rigidities are an unlikely channel through which money exerts this inﬂuence. We’ll thus conclude with a brief review of how money might inﬂuence real activity not through labor market channels but rather through those associated with ﬁnancial markets. Christiano et al. (1999), among others who followed Friedman and Schwartz (1963), evaluate how economies respond to monetary policy shocks.16 They do so by ﬁrst estimating parameters of the Federal Reserve’s reaction function (that is, the rule by which the Fed transforms economic information into monetary policy) and then treating the corresponding residuals (that is, unexplained variation in the bank’s policy choices) as evidence of exogenous policy shocks. By evaluating how economic aggregates respond to various speciﬁcations of these shocks, they ﬁnd robust evidence that (unexpected) monetary policy exerts real inﬂuences as follows: After a contractionary monetary policy shock, short term interest rates rise, aggregate output, employment, proﬁts, and various monetary aggregates fall, the aggregate price level responds very slowly, and various measures of wages fall, albeit by very modest amounts. (Christiano et al., 1999: 69)

However, like that which we review above from Klenow (see Clement, 2003) and Stock and Watson (1999: 42), this evidence also oﬀers little support for nominal rigidities as a causal channel. For example, Christiano et al. (1997) report that real wages decline slightly when money experiences a negative shock, and this decline appears especially prominent in the manufacturing sector. But recall that a nominal rigidities story says that real wages increase in the face of contractionary shocks.

78

Money in a dynamic economy

Moreover, recall that a popular rationalization of nominal rigidities as a causal channel is that laborers (for example, those employed in manufacturing) enter long-term nominal wage agreements. Christiano et al.’s (1997) evidence, however, not only suggests that real wages react in the wrong direction, it suggests that this inconsistency is especially pronounced in a sector that is, in principle, most susceptible to nominal rigidities! The persistence with which scholars ﬁnd reduced form evidence for money’s real eﬀects, but evidence against nominal rigidities as a plausible transmission mechanism, has motivated searches for alternative channels through which real aggregates respond to money. One such channel emerges from monetary policy aﬀecting the capacity for ﬁnancial intermediaries to supply credit services. A negative shock, for example, can decrease banks’ reserves, and thus the level of assets from which they can generate loanable funds. To the extent that prospective borrowers cannot readily substitute for this source of ﬁnancial capital, real economic activity can decline in the face of a negative monetary policy shock.17

CONCLUSION How should a monetary authority respond to demand or supply shocks? Our analysis of aggregate supply in this chapter shows that, in principle, monetary policy can oﬀset shocks and thus enhance economic stability. Note, however, that any such opportunity depends on the rate at which market forces can ‘correct’ deviations from full employment. Given the considerable delay in economic data becoming available, and lags between policy considerations, implementation, and eﬀects, economists are often skeptical about the capacity for ‘ﬁne tuning’ to outpace self-correcting mechanisms. Moreover, as we’ll see in following chapters, ﬁnancial developments may have contributed to a marked reduction in the cycles that advanced economies experience, and possibly left little room for even the most eﬀective policy to improve economic performance.

NOTES 1. 2. 3. 4.

See, for example, the literature following Friedman and Schwartz (1963). This change does not inﬂuence the quantity of labor that ﬁrms are willing to employ for any real wage – that is, it does not shift the labor demand curve. In some models, the willingness of households to supply labor constrains increases in the quantity of labor demanded. There, the aggregate supply curve slopes upward only until full employment is reached. Stated more formally, the expectation operator E is ‘rational’ if E[x E[x] ] 0 where the variable x represents an outcome of interest (for example, the price level P).

Money in a Keynesian economy

5. 6. 7. 8. 9. 10.

79

According to this deﬁnition, expectations can be rational (that is, an optimal forecast) even if they prove inaccurate after resolving the uncertainty over which they were formed. Both Friedman and Phelps won Nobel Prizes in economics, in part for their prescient criticisms of the Phillips Curve. In this expression, the n-subscript denotes ‘natural rates’ (for example, natural rates of output or unemployment), the e-subscript denotes expected values, and s-denotes a supply shock. Below, we’ll see that this choice is ‘optimally’ too high. This analysis draws heavily from Blanchard and Fischer’s textbook treatment (1989: 456–61, Figure 9.7) which, in turn, draws from Shapiro and Stiglitz’s (1984) seminal contribution to the literature. Implications that we are about to draw from this model appear robust to these apparent ‘over’-simpliﬁcations. Yellen (1984: 202–3) explains this implication as follows: If all ﬁrms pay an identical wage, and if there is full employment, there would be no cost to shirking and it would pay all workers, assumed to get pleasure from loaﬁng on the job, to shirk. In these circumstances, it pays each ﬁrm to raise its wage to eliminate shirking. When all ﬁrms do this, average wages rise and employment falls. In equilibrium, all ﬁrms pay the same wage above the [classical] market clearing, and unemployment, which makes job loss costly, serves as a worker-discipline device. Unemployed workers cannot bid for jobs by oﬀering to work at lower wages. If the ﬁrm were to hire a worker at a lower wage, it would be in the worker’s interest to shirk on the job. The ﬁrm knows this and the worker has no credible way of promising to work if he [or she] is hired.

11. 12.

13. 14. 15. 16. 17.

In this sense, it might be more accurate to characterize eﬃciency wage theories as measuring diﬀerences in employment levels due to diﬀerences in monitoring technologies, rather than rationalizing involuntary unemployment per se. Woodford (1994: 1), for example, observes that ‘Much of the eﬃciency wage literature has been partial equilibrium in character, and content with a static analysis of the determinants of a steady-state level of unemployment.’ Yellen (1984: 204) oﬀers a related observation. And while beginning a textbook treatment of this diﬃculty, Blanchard and Fischer (1989: 461) conclude that ‘Eﬃciency wages . . . do not necessarily imply that small costs of changing wages will lead to large nominal rigidities.’ ‘Failure to choose wages in precisely optimal fashion results in little loss to the ﬁrm, so that sticky wages in response to shocks result in only small losses to a ﬁrm that does not exactly maximize’ (Akerlof and Yellen, 1986: 17, italics in original). Note, however, that this model also implies that real wages are counter-cyclical. Clement (2003) summarizes this research. See also, for example, Romer and Romer (1989). Christiano et al. (1999) review evidence on how shocks that work through this mechanism may create diﬀerent eﬀects for diﬀerent agents in an economy. For example, large businesses may not realize large direct consequences of a shock since their capital market substitutes for intermediaries subject to reserve requirements are relatively rich.

8.

Should monetary policy be active?

THE LUCAS CRITIQUE AND POLICY EVALUATION Robert Lucas famously extended the rational expectations hypothesis to consider whether historical data can inform considerations about proposed policies. His conclusion is damning – that is, if individuals form expectations rationally, then reduced form models ‘provide no useful information as to the actual consequences of alternative economic policies’ (Lucas, 1976: 20, emphasis in original).1 Lucas’s ‘critique’ builds on Muth’s (1961) seminal insight that expectations depend on the structure in which they develop. This structure shapes how today’s economic conditions evolve into tomorrow’s. Hence, if policies change an economy’s structure, then the manner in which maximizers react to policy variables can also change. In this case, policy makers must predict individuals’ reactions under the new regime from data on the old regime – that is, information about how individuals reacted to policy variables before the economy’s structure changed! Insights to the Phillips Curve’s Breakdown To better understand Lucas’s critique, let’s consider how it rationalizes the Phillips Curve’s breakdown. Recall that the Phillips Curve, on its face, evidences the potential for policy makers to trade higher inﬂation for lower unemployment – that is, it suggests that monetary authorities can dampen real ﬂuctuations if they are willing to incur nominal ﬂuctuations. But evidence from which this trade-oﬀ became apparent takes a reduced form – that is, it largely rests on the negative correlation between price and unemployment levels during the 1960s. Given that correlation does not imply causation (see Part I), we begin to appreciate Friedman’s, Phelps’s, and others’ early concern about the Phillips Curve.2 This concern grows when we note that, before the Phillips Curve, monetary authorities were less likely to pursue an inﬂation–unemployment trade-oﬀ. But evidence for the Curve comes from this relatively ‘passive’ period. There, changes in the price level would have (by deﬁnition) largely been unexpected. Consequently, if the ‘Lucas supply curve’ (the supply curve associated with the ‘expectations-augmented’ Phillips Curve) 80

Should monetary policy be active?

Figure 8.1

Passive policy

Active policy

Random inflation

Systematic inflation

Unexpected inflation

Unexpected inflation

Output

Output

81

Lucas’s critique and the Phillips Curve’s breakdown

reasonably summarizes forces that act on aggregate supply, then this period’s output would have indeed shared a negative relationship with measured inﬂation – that is, random changes in inﬂation would have created unexpected inﬂation, and thus shared a negative association with unemployment in reduced form evidence. Figure 8.1’s left-hand column illustrates how such a structure could have generated this reduced form relationship. Supported by this relationship, however, policy makers actively pursued an inﬂation–unemployment trade-oﬀ. Once they did, ﬂuctuations in the price level became more systematic, rather than random as they tended to be during the data generating period. Consequently, the reduced form relationship between inﬂation and output broke down. Realizing a change to more active policy, maximizers rationally updated the manner in which they interpreted price changes, and thus left (again) only unexpected price changes to inﬂuence real activity. So while unexpected

82

Money in a dynamic economy

price changes may have still inﬂuenced real activity during this latter period, the Phillips Curve relationship would have been overwhelmed by real output’s insensitivity to what became a relatively large expected component of price changes.3 As the right panel of Figure 8.1 illustrates, policy can change how individuals form expectations, and this change can obviate relationships that appear in historical data. Lucas’s critique is much deeper than the simplistic observation that estimates of policy eﬀects are inherently noisy, or even that policies have ‘unintended consequences.’ Rather, it shows how policies not only aﬀect variables of interest, but also the manner in which individuals anticipate those variables’ evolution. Because this anticipation inﬂuences expectations, and expectations inﬂuence behavior, relationships that appear in historical data may contain little information about how individuals would react to policy changes. Identifying an economy’s structural parameters (that is, parameters that are insensitive to policy changes), and how policies inﬂuence social forces that emanate from them, thus becomes necessary to logically evaluate proposed policies. The Lucas Critique: A More Formal Development Suppose that the quantity of output that a ﬁrm supplies during any period consists of both permanent and cyclical components as follows:4 yit ypit ycit Here, yit denotes ﬁrm i’s (real) output in period t, and superscripts p and c denote ‘permanent’ and ‘cyclical,’ respectively. Let’s also assume that a ﬁrm’s cyclical supply varies with the relative price of its output as follows: ycit (pit peit ) Here, pit denotes the price of ﬁrm i’s output in period t, while peit denotes ﬁrm i’s expectation of the economy’s general price level. This characterization implies that the cyclical quantity supplied by ﬁrms depends on how they interpret departures of their own price from the level that obtains in the economy more generally. For example, to the extent that ﬁrm i interprets departures as coming from ﬂuctuations in relative (as opposed to aggregate) demand, it will heavily weight that departure in deciding what quantity to supply. In low-inﬂation/stable price economies, ﬁrms will thus tend to interpret the diﬀerence (pit peit ) as evidence of a change in relative demand, and a high value of the parameter will govern ﬁrms’ quantity choices.

Should monetary policy be active?

83

Now let’s consider how a policy maker might exploit this model’s insight to evaluate whether a proposed policy will achieve its objective. Recall from our examination of the classical and Keynesian models that changing the money supply inﬂuences an economy’s price level. One might thus invoke our model of cyclical output to evaluate whether changing the money supply dampens e short-term economic ﬂuctuations – that is, M → (pit pit ) → ycit. To carry out such an evaluation, however, we need to estimate – that is, uncover information about the ‘structure’ through which changes in policy variables inﬂuence real activity. But cannot be identiﬁed from reduced form evidence. To see this diﬃculty more clearly, suppose that ﬁrms rationally expect price changes, and that available information includes the price at which their own output sells (that is, pit) and the mean of the economy’s past general price levels (which we’ll denote by pt). The ﬁrm’s expectation of the economy’s general price level (and thus its relative price) may thus be represented as follows: peit (1 ) pit pt Here, (1) denotes the relative weight that a ﬁrm places on information about its own price when forming expectations about the economy’s general price level, given information about the economy’s past price level. Substituting this expression into that for a ﬁrm’s cyclical output-component (that is, ycit) and re-arranging terms, we arrive at the following relationship: ycit (pit pt ) Finally, averaging across all ﬁrms i (i1. . . N), we ﬁnd the following expression for the economy’s cyclical component of aggregate supply: yct (pt pt ) In words, this expression says that (aggregate) deviations from trend supply ( yct) vary with deviations from past price levels. This relationship is qualitatively identical to that suggested by the Phillips Curve. However, even if a monetary authority could cause (pt pt ) to diﬀer from zero, it will be unable to evaluate its policy from estimating the above relationship. Indeed, the resulting reduced form evidence would confound the eﬀects of and . In other words, the reduced form evidence would identify the gross eﬀect of a change in (pt pt ) , but fail to distinguish the manner in which information enters a ﬁrm’s price level expectation () from that in which a perceived relative price change inﬂuences the quantity of output that ﬁrms choose to supply ( ).

84

Money in a dynamic economy

Finally, even if policy makers solve this identiﬁcation problem, the resulting evidence would still have to ground itself on historical data. But if policy inﬂuences an economy’s structural parameters, then the estimated values of and will diﬀer (in a perhaps unknowable way) from those that govern the post-implementation economy. Lucas (1976: 37–8) highlights this diﬃculty by suggesting that, while one can reasonably expect the parameter to be stable, evolution of is sensitive to variability in the economy’s price level and thus inherently unpredictable as a function of demand policy. He thus concludes that ‘the long-run output-inﬂation relationship as calculated or simulated in the conventional way has no bearing on the actual consequences of pursuing a policy of inﬂation’ (Lucas, 1976: 39, emphasis in original).

POTENTIAL BENEFITS FROM STABILIZATION POLICY Taking US performance over the past 50 years as a benchmark, the potential for welfare gains from better long-run, supply-side policies exceeds by far the potential from further improvements in short-run demand management. (Lucas, 2003: 1, italics in original).

Lucas’s critique highlights an important diﬃculty for activist monetary policy – even if a Phillips Curve-type trade-oﬀ is available, the eﬀects of policies that might exploit that trade-oﬀ cannot be easily estimated. But even setting this serious diﬃculty aside, stabilization policy may now have little room to improve economic performance. As the above quote from Lucas argues, for example, policies that aim to increase growth rates may be superior to those that attempt to dampen economic ﬂuctuations in terms of potential to improve economic performance. Lucas draws this strong conclusion from a relatively straightforward thought experiment. He asks, quite simply, how much better oﬀ consumers would be if consumption’s variability about trend was eliminated. Measuring welfare in this manner (that is, as the average level of consumption that a consumer would forgo to avoid high frequency changes in consumption), Lucas (2003:1) calculates that even the most successful policy would improve welfare by only one-half of one-tenth of 1 percent! Lucas develops this conclusion from the following indiﬀerence condition between a stochastic and deterministic stream of consumption (patterns that might emerge from a volatile and stable economy, respectively). In particular, he evaluates how much more consumption (measured by the parameter ) a risk-averse consumer must receive on average when faced with a stochastic consumption stream (measured by the variable ct at each period

85

Should monetary policy be active?

t) than with a deterministic stream Aet (which is constant for each t and equals the average level of the stochastic stream of ct):

E

t0

t

((1 ) ct ) 1 E 1

t0

t

(Aet ) 1 1

The value of that satisﬁes this condition tells us how much compensation (in terms of average consumption) is necessary to make a consumer indiﬀerent between receiving a risky and deterministic stream of consumption. Doing so, Lucas ﬁnds that 0.5 · 2. He then looks to the data for estimates of how risk averse are consumers (an estimate of ) and how volatile is consumption (an estimate of 2). Equipped with these estimates, he calculates that 0.0005. In words, this calculation says that, evaluated at empirically plausible levels of risk aversion and consumption volatility, consumers would willingly forgo only one-twentieth of 1 percent of their average consumption to remove all uncertainty about future consumption prospects! Moreover, this calculation appears robust to particular assumptions about the coeﬃcient of risk aversion (see Lucas, 2003: 8). Lucas thus concludes that stabilization policies (beyond those that steadied monetary aggregates and nominal spending in the post-war United States) promise (at most) negligible increases in welfare. Given this small potential, and the diﬃculties that we encountered with both the classical and Keynesian models of high-frequency movements, one might be skeptical about the capacity for pursuing such an improvement to create net beneﬁts. Others are investigating, however, the potential for stabilization policy to inﬂuence welfare not only via its eﬀect on short-term ﬂuctuations but also from its spillover to long-term trends.5 For example, stability could inﬂuence an economy’s equilibrium level of ‘growth-promoting’ activities, such as investment. Barlevy (2004) characterizes this potential as follows: Intuitively, eliminating ﬂuctuations reallocates investment from periods of high investment to periods of low investment. Since the marginal return to investment is higher when there is less investment, this allows agents to achieve more growth from the same resources.

And even if this eﬀect is small, its welfare consequences can be considerable in light of compounding growth rates. Indeed, Barlevy’s (2004) model implies that the ‘cost’ of ﬂuctuations equals the value of annually forgoing 8–10 percent of trend consumption (versus 201 percent for Lucas).

86

Money in a dynamic economy

NOTES 1. Notice that this conclusion is especially relevant for monetary policy since, as we just saw, structural evidence on how money relates to real activity appears relatively diﬃcult to develop. 2. Robert Lucas was a graduate student under Milton Friedman’s instruction. 3. This implication is consistent with observations that, rather than breaking down, the Phillips Curve relationship has simply changed its intercept. 4. This model abstracts from one that Lucas develops in his original ‘critique’ (Lucas, 1976, Section 5.3). 5. Note that Lucas (2003) implicitly assumes that dampening ﬂuctuations would not aﬀect the trend rate of growth to which the economy would return.

9.

Is monetary policy active?

To breed an animal with the right to make promises – is not this the paradoxical task that nature has set itself in the case of man? (Friedrich Nietzsche, 1887)

INTRODUCTION The last chapter’s analysis largely argues against activist monetary policy. We saw, for example, that reduced form evidence oﬀers little guidance to activist policy. Moreover, even if policy could fully eliminate business cycles, consumers may place a negligible value on marginal increases in economic stability. The present chapter turns this analysis in a more positive direction. Here, we’ll address the important question of why, despite discouraging arguments against policy activism, monetary authorities can maintain a strong incentive to actively inﬂate. This paradox emerges from the problem of ‘time inconsistency’: The problem in a nutshell is as follows: The best policy plan has the property that after following the plan for a while, everybody agrees that there is a better alternative than to continuing the original plan. But, if a group of people cannot commit to its original plan, the ex ante best plan [that which initially appeared optimal] is not feasible. (Edward C. Prescott, quoted by Rolnick, 1996)

We’ll argue that the ‘ex ante best plan’ is a policy that produces price stability. Notice, however, that once individuals anticipate such a policy, expectations in the Lucas supply curve are ﬁxed. At this point, the monetary authority has an opportunity to trade inﬂation for employment (this opportunity does not exist before expectations are formed). Despite the optimality of low-inﬂation policies, monetary authorities can thus prefer high-inﬂation policies. It is important to highlight that the problem here is not one of agency. Rather, dynamic inconsistency emerges from the general manner in which bargaining positions change over time. Actions that comprise an initially optimal plan can induce inferior outcomes after other parts of that plan have been implemented. Even for optimal plans, ‘staying the course’ is a fundamentally diﬃcult objective for maximizers since they (by deﬁnition) want to re-optimize as the plan rolls out. 87

88

Money in a dynamic economy

To be productive, then, mechanisms that govern money must either dampen the incentive for opportunistically making such changes or increase the cost of acting on that incentive. Our next set of chapters shows that agents who focus more on price levels than on employment or output gaps face a considerably lesser incentive in this regard. This insight implies that societies can better commit to policies that minimize economic ﬂuctuations by delegating monetary authority to an ‘imperfect agent.’ In other words, an imperfect policy agent can do a better job of maintaining price stability than can a more accountable authority. Moreover, received evidence lends support to this implication – that is, economic performance exhibits a strong and positive relationship with central bank ‘independence.’

THE TIME INCONSISTENCY PROBLEM The folk song (as arranged by Led Zeppelin), ‘Gallows Pole,’ nicely illustrates how optimal policies can be inconsistent in the sense that parts of those policies become sub-optimal when the time comes to implement them. Consider, for example, the following excerpt: Accused to the hangman: Hangman, hangman, hold it a little while, I think I see my brother coming, riding a many mile. Accused to the brother: Brother, did you get me some silver? Did you get a little gold? What did you bring me, my brother, to keep me from the Gallows Pole? Brother to the accused: Brother, I brought you some silver, I brought a little gold, I brought a little of everything to keep you from the Gallows Pole. Hangman to the accused: Your brother brought me silver . . . But now I laugh and pull so hard . . . see you swinging on the Gallows Pole!

At least initially, an ‘optimal’ policy for the ‘hangman’ and the ‘accused’ may be for the accused to bribe the hangman and the hangman to free the accused. As the song illustrates, however, the hangman’s most attractive action after receiving the bribe is to hang the accused anyway – that is, the optimal policy is time inconsistent. The game tree shown in Figure 9.1 formally illustrates this inconsistency. Viewed from the game’s initial node, the actions (Bribe, Don’t Hang) induce an ‘optimal’ outcome – that is, an outcome from which no player can be made better oﬀ without making the other worse oﬀ. However, once the Accused chooses an action (any action!), the Hangman’s best response is to Hang. Even though both players strictly prefer the payoﬀs that result from

89

Is monetary policy active?

Accused

Don’t bribe

Bribe

Hangman Don’t hang

5 0

Figure 9.1

Hangman Hang

1 1 Inferior equilibrium

Don’t hang

Hang

3 3 Optimal outcome

0 5

Optimal policy in the hangman game is inconsistent

(Bribe, Don’t Hang), pursuing their (unconstrained) interests induces the inferior outcome (Don’t Bribe, Hang). More generally, time inconsistency emerges from the inability of (narrow) maximizers to credibly commit against optimizing at each point of a plan’s implementation (even if continually re-optimizing induces an inferior outcome!). This inability stems from actions that comprise an optimal plan being inconsistent with a player’s objective when the time comes to play them. For example, the action Don’t Hang is part of the hangman game’s optimal plan, but when the time comes for the Hangman to choose an action, Don’t Hang produces an inferior payoﬀ. If the hangman is a maximizer, then promises to implement the optimal policy will lack credibility.

THE PROBLEM FOR MONETARY POLICY The problem of time inconsistency implies that, absent opposing institutional forces, maximizers encounter diﬃculty in committing to optimal policies. Constituents may thus rationally expect sub-optimal actions from their political agents (even perfect ones), and this expectation will be

90

Money in a dynamic economy

fulﬁlled in equilibrium. Even if a low-inﬂation policy is optimal, for example, monetary authorities may be unable to credibly commit to such policies. Constituents, in turn will play actions in expectation of a suboptimal monetary policy. In a manner that is qualitatively identical to the hangman game, an economy of maximizers can become stuck at an outcome that everyone agrees is inferior. Seeing this problem’s root will let us carefully consider the eﬃcacy of how monetary authorities organize themselves. Before we turn to that consideration, then, let’s further develop our insight to why optimal monetary policies can be inconsistent. Our investigation will build from an unremarkable game theoretic model where:1 ● ● ●

constituents begin by forming expectations over monetary policy; given these expectations, a ‘benevolent dictator’ chooses policy;2 and this policy choice inﬂuences the economy’s level of social welfare.

We can solve this game by starting at its end and moving backward (that is, by employing a method known as ‘backward induction’). Let’s begin, then, by evaluating the dictator’s ‘best response’ to any expectation that its constituents might form. In particular, suppose that our dictator’s objective is to maximize the following measure of social welfare: Z 2 (Y Yn ) 2 where Y denotes (real) aggregate output, Yn denotes the level of output associated with ‘full employment’, and denotes the inﬂation rate. Qualitatively, the important feature of our measure is that welfare (Z) increases with the economy’s realized output Y and decreases with departures from stable prices (that is, with 0). If monetary policy only aﬀects nominal prices (and thus inﬂation), then our authority’s optimal policy would be to induce no inﬂation, and developing the optimal policy would be trivial, at least from a purely economic perspective (that is, 0 would maximize Z without aﬀecting Y or Yn). But previous chapters suggest that monetary policy can inﬂuence real activity if it departs from expected policy (that is, to the extent that ( e ) 0 where e denotes ‘expected’ inﬂation). Let’s consider a richer speciﬁcation of social welfare, then, that permits just such a departure. In particular, let’s assume that the following speciﬁcation of the expectations-augmented Phillips Curve adequately captures the relationship between unexpected inﬂation and ﬂuctuations from output’s natural rate:

Is monetary policy active?

91

Y Yn ( e ) (Y Yn ) ( e ) Substituting this last expression into our original objective gives us the following measure of social welfare: Z 2 ( e ) 2 Our policy maker thus confronts the problem of choosing a rate of price change to maximize welfare (that is, choose to maximize Z). To solve this problem, note that an ‘optimal’ choice must (by deﬁnition) be one from which our dictator has no incentive to change. In other words, the welfare-maximizing choice of inﬂation, say *, must exhaust all opportunities to increase Z by changing . Satisfying this condition implies that the authority’s optimal choice of inﬂation must be such that the ﬁrst derivative of Z with respect to must equal zero (that is, evaluated at an optimal rate of inﬂation, social welfare must not vary with a small change in inﬂation). To see this implication, note that choosing from an interval where inﬂation increases welfare (that is, where Z 0) leaves superior levels of welfare unrealized. Indeed, any where the ﬁrst derivative of Z with respect to exceeds zero cannot be an optimal choice since choosing a higher rate of inﬂation would be feasible and increase welfare. Likewise, any from which decreasing inﬂation would increase welfare (that is, such that Z 0) also forgoes superior levels of welfare. Policy choices that maximize welfare must thus satisfy the condition that changing the choice negligibly aﬀects welfare (that is, choices of where Z 0). The following set of equations thus tells us that, acting ‘as if’ it is a maximizer, our benevolent dictator will choose 1 for any constituentexpectations e. Notice that, even a perfect policy agent’s best choice (that is, *) can induce strictly positive inﬂation in equilibrium: Z 2 2 0

* 1 Given the dictator’s best response, we can now work back to how constituents form expectations about monetary policy, and what the process of expectation-formation implies for the level welfare that is feasible for societies (that is, the level at which society rests when each of its members rationally pursues its objective).

92

Money in a dynamic economy

Recall that rational expectations do not systematically induce mistakes. Given this deﬁnition, and our deduction that even benevolent dictators will choose 1 when the game’s play reaches them (that is, after constituents form their expectations), rational constituents will expect 1 (re * 1).3 We can thus calculate our economy’s equilibrium welfare level (denoted by ZE) as follows: ZE 2 (* re ) (* ) 2 2 (1 1) (1) 2 1 Our analysis thus far can be summarized as follows. First, note that our model assumes away any prospect of political agents imposing ‘inferior’ policies upon constituents as a function of agent-preferences diverging from constituent-preferences. Indeed, we endowed our dictator with the same objective as that of society – that is, maximize welfare. Given this setup, then, and the order of play (that is, constituents set expectations before dictators choose policy), dictators maximize social welfare by choosing a strictly positive inﬂation rate. Correctly anticipating (on average) that dictators will choose this policy, constituents rationally expect an inﬂationary policy. Notice the paradox with which this game leaves us – that is, society settles on an inferior policy, even though the dictator’s objective is to maximize social welfare. To see this implication, consider what would have happened if our dictator could credibly commit to a policy of zero inﬂation (that is, to setting 0). In this case, our constituents would have rationally expected zero inﬂation (e 0). Consequently, as Figure 9.2 illustrates, society would have enjoyed a level of welfare in excess of what obtains under a non-commital, but benevolent, policy maker (that is, Z|0 2 (00) – (0)2 01ZE). But because payoﬀs to inﬂating exceed those to not inﬂating for any expectation, our benevolent dictator cannot promise

Figure 9.2

␲e = 0

␲e = 1

␲=0

Z=0

Z = –2

␲=1

Z=1

Z = –1

The optimal monetary policy is inconsistent

Is monetary policy active?

93

against maximizing social welfare, and society must rest at an inferior equilibrium!4 In addition, note that society cannot ‘rationally’ fool itself by setting e 0. To be sure, each member of society would be better oﬀ if everyone else set expectations in this manner. Here, policy makers would continue to choose 1 (inﬂating is, after all, a dominant strategy) and society would rest at a superior level of welfare (that is, Z|0 2 (1 0) (1)2 1). But each individual could enjoy a higher level of utility by expecting inﬂation, while letting others irrationally form expectations – that is, individuals maintain homogeneous tastes over social welfare, but heterogeneous tastes over the distributional consequences of policies that can aﬀect welfare. Each member thus has an incentive to cheat on ‘agreements’ to irrationally form expectations. Absent the enforceability of such agreements, members cannot escape the inferior equilibrium that results from time inconsistency.5

NOTES 1. See Barro and Gordon (1983b) and Blanchard and Fischer (1989, Chapter 11) for slightly more rigorous, though qualitatively identical, illustrations of how the time inconsistency problem relates to monetary policy. 2. While this assumption appears to conﬂict with our maximizing-behavior axiom, we adopt it here to emphasize that the time inconsistency problem does not emerge from inconsistencies between preferences of policy makers and their constituents. Rather, even when preferences are perfectly aligned, actions that appear optimal when considered today can appear sub-optimal when considered at a future date. Moreover, this inconsistency emerges simply from the passing of time, not because the relevant policy environment might change over time. We’ll thus see that, even when our dictator’s constituents would enjoy maximal feasible utility when the price level remains constant (that is, when the dictator chooses a policy of zero inﬂation), the benevolent dictator will inﬂate! 3. The subscript ‘re’ denotes ‘rational expectation.’ 4. This type of problem (that is, one in which expectations inﬂuence outcomes which, in turn, feed back to inﬂuence policy decisions) is a very common one. It arises in a variety of contexts including the economics of patents and copyrights, capital taxation, nominal government debt and even child rearing! 5. Chari et al. (1989) formally highlight this feature of the time inconsistency problem.

PART IV

Governing money

In Part II, we argued that money can enhance welfare by economizing on bartering costs. To realize this potential, however, an economy’s participants must be conﬁdent that prospective trading partners will accept nominal exchange media for real goods and services. Indeed, since money does not directly generate utility, maximizers who lack this conﬁdence will not trade consumption possibilities for ‘money.’ Moreover, absent this willingness to trade via nominal exchange media, individuals must revert to barter, and realize an inferior equilibrium. We thus learned that governing an economy’s monetary authority in a manner that promotes conﬁdence in an exchange medium is important for achieving high levels of social welfare. We then saw in Part III that, in addition to inﬂuencing levels of welfare, money can also inﬂuence ﬂuctuations in an economy’s aggregate output (and even growth in a model like Barlevy’s (2004)). (In Part V, we’ll see that how an economy governs its ﬁnancial markets can also inﬂuence longerterm trends in output.) To be sure, channels through which money inﬂuences ﬂuctuations may not have been satisfactorily identiﬁed, but persuasive reduced form evidence supports the hypothesis that this relationship is causal. If this relationship largely emerges from changes in unexpected inﬂation, however, actively managing monetary policy is unlikely to improve economic performance (and, indeed, may diminish it!). Nevertheless, even ‘perfect’ political agents face strong incentives to pursue such policies (recall the problem of time inconsistency). Parts II and III thus help us understand (a) why monetary systems are desirable and (b) why commitment problems can work against their development. In Part IV, we’ll consider how the organization of monetary authorities can insulate policy makers from these discouraging incentives, or increase the cost of acting on them.

96

Governing money

We’ll start in Chapter 10 by quantifying the level of services that monetary authorities produce. Our measure will build on Part II’s insight that exchange media vary in their ‘moneyness’ – that is, the extent to which these media facilitate transacting. In particular, we’ll see that weighting monetary assets by their degrees of moneyness can more accurately characterize an economy’s production of monetary services than can more traditional simple sum measures. Equipped with an understanding of relevant measurement issues, we’ll think about how an economy should govern its monetary authority. To aid this consideration, recall that Part II’s monetary authority is a passive clearinghouse. But had this authority been a maximizer, it would have opportunistically printed ‘pieces of paper’ (POPs) to acquire real goods and services. Indeed, governments that do not have access to well-developed tax systems frequently resort to the ‘printing press’ in just this manner. And because doing so dampens conﬁdence in the capacity for currencies to act as exchange media, individuals must forgo the superior equilibria that are available in well-developed monetary systems. An important part of our argument, then, will be how societies can discourage monetary authorities from substituting increases in nominal money for increases in real taxes. Moreover, recall from Part III that monetary authorities also face a time inconsistency problem in the realm of economic stability policy. To the extent that authorities enjoy discretion when forming stability policy, they can strategically act on temporal changes in bargaining positions. Constituents that anticipate such behavior (those that rationally form expectations) will, in turn, act in a manner that induces less than optimal outcomes (for example, inﬂation costs without output beneﬁts). In addition to understanding how the governance of monetary authorities should relate to that of ﬁscal authorities, we’ll thus want to understand how governing monetary authorities can mitigate opportunistic policy actions as they relate to economic ﬂuctuations. We’ll argue that monetary authorities should organize themselves in a manner that diminishes the incentive to pursue discretionary policies or increases the cost of acting on that incentive. Note well that this prescription does not dissuade policy makers from being maximizers. Rather, it improves welfare within the constraints that emerge from maximizingbehavior – for example, maximizers have an incentive to strategically act on changes in bargaining positions. We’ll learn that, by delegating monetary authority to agents whose objectives encourage time-consistent policies, and discouraging the circumvention of that delegation, ‘central bank independence’ (CBI) maintains an attractive capacity to achieve such improvements. Equipped with an understanding of how monetary authorities should be organized, we’ll conclude this set of chapters on a positive note. Our

Governing money

97

objective is to evaluate the extent to which monetary authorities have adopted our prescriptions and the success of CBI where it has been implemented. We’ll learn that CBI shares a strong and positive relationship with economic performance across countries, and consider the eﬃcacy of a particularly inﬂuential central bank (the US’s Federal Reserve System) in mitigating the problems of inﬂation taxes and time inconsistency.

10.

Measuring monetary services

INTRODUCTION Before we can contemplate how monetary authorities are and should be organized, we should understand how to measure (one of) their output(s) – that is, transactions services. We’ll address this issue by arguing that exchange media should be counted as ‘money’ only to the extent that they contribute to an economy’s stock of transactions services. This method of counting diﬀers from more common ‘simple sum’ methods that implicitly treat diﬀerent forms of money as equally contributing to an economy’s production of exchange services.

SIMPLE SUM MEASURES To measure an economy’s stock of monetary services, authorities tend to use ‘simple sum’ indexes. As their name suggests, these measures add, one for one, the units of exchange media circulating in an economy during a particular point in time. For example, the United States’ Federal Reserve System measures the US money supply at three levels – that is, M1, M2 and M3. Each level, moving from M1 to M3, progressively includes less liquid forms of money. In particular:1 ●

● ●

M1 includes ‘currency in the hands of the public, travelers’ checks, demand deposits, and other deposits against which checks can be written’; M2 includes M1 plus ‘savings accounts, time deposits of under $100 000, and balances in retail money market mutual funds’; and M3 includes M2 plus ‘large-denomination ($100 000 or more) time deposits, balances in institutional money funds, repurchase liabilities issued by depository institutions, and Eurodollars held by US residents at foreign branches of US banks and at all banks in the United Kingdom and Canada.’2

Notice that each measure treats its components as perfect substitutes. For example, M3 treats each monetary unit held in the form of a 99

100

Governing money

‘large-denomination time deposit’ and ‘currency in the hands of the public’ as producing the same level of monetary services. Clearly, however, we can more readily employ currency as an exchange medium than we can time deposits. Indeed, transferring rights over currency’s transactions services can be as easy as physically handing currency to a trading partner, whereas transferring rights over a time deposit may exhaust a far greater level of resources (for example, those associated with having to wait for the deposit’s maturity). In this light, we see that measuring an economy’s supply of monetary services by (at least implicitly) treating each factor as equally productive can upwardly bias our measure and complicate that measure’s comparison over time.

WEIGHTED AGGREGATES By explicitly accounting for monetary assets’ variation in liquidity, weighted sum aggregates can oﬀer a relatively accurate measure of an economy’s stock of monetary services. Barnett (1987) carefully derives the ‘optimal’ weights for a monetary aggregate by examining a model of optimal cash holdings. In short, Barnett builds from microeconomic foundations (that is, the household’s and ﬁrm’s maximization problems) the insight that a monetary asset’s ‘user cost’ should determine its weight in a monetary aggregate. While economically intuitive, this insight leaves open the question of how, exactly, to calculate an asset’s weight. Barnett shows that we can measure an asset’s user cost by evaluating the diﬀerence between its nominal rate of return and that which commensurate resources would have earned from buying a ‘benchmark asset.’ In other words, to measure an asset’s monetary services, Barnett simply compares the asset’s rate of return to that of an asset that only transfers consumption capacity across time (that is, a pure savings vehicle). Notice that, for maximizers to rationally hold a monetary asset whose rate of return falls short of a benchmark, the monetary asset must produce alternative beneﬁts. And since these beneﬁts are not (by deﬁnition) transferring consumption capacity across time, they must come from the asset’s ‘moneyness.’ To understand this method more clearly, let’s consider the weight with which currency would enter Barnett’s aggregate. Note that currency’s nominal rate of return is zero – for example, a dollar stored in your pocket today will still be a (nominal) dollar tomorrow. Rather than physically possessing a unit of currency, however, you could have traded that unit for the ‘benchmark asset’ – for example, purchased a long-term obligation from a stable and productive government. So why would you rationally forgo the return on the benchmark asset?

Measuring monetary services

101

A maximizer does so to the extent that currency produces transaction services. And since the opportunity cost of holding currency exceeds that of holding any other monetary asset, we see that currency logically receives the greatest weight in a monetary aggregate. Other assets earn a strictly positive nominal return (that is, a nominal return in excess of that associated with currency), but not as much as does the benchmark asset. These assets thus receive a weight that is less than that of currency but greater than that of the benchmark asset – their value comes from producing monetary services and transferring consumption capacity over time. And since the benchmark asset only transfers consumption capacity over time, it receives zero weight in a monetary aggregate – none of its value comes from acting as an exchange medium. Weighted aggregates can thus oﬀer a relatively accurate measure an economy’s monetary services because they explicitly account for the ‘user cost’ of monetary services. User costs measure the forgone interest (opportunity cost) of holding a monetary asset when a higher yielding benchmark could have been held. Maximizers rationally incur these costs to the extent that monetary assets produce transaction services. Using these costs to weight monetary assets can thus facilitate a more precise measure of an economy’s stock of transaction services.

NOTES 1. Source: Federal Reserve Bank of New York. Accessed 2 April 2008 at http:// www.newyorkfed.org/aboutthefed/fedpoint/fed49.html. 2. Note that M3’s components are unlikely to ﬁnd direct employment as an exchange medium.

11.

Organizing the production of monetary services

And people wondered, if we can explain the laws of the universe, can’t we explain the principles that should control the government? (Anthony Kennedy, US Supreme Court Justice, 2006 address to the ABA assembly, Honolulu, HI)

INSTITUTIONS AND COMMITMENT We saw in Part III that ‘time inconsistency’ is problematic in that a policy’s optimality is sensitive to when one evaluates it. The policy that appears best before economic agents play actions can eventually appear sub-optimal as the policy unfolds. This change does not emerge from new information revealing itself over time, or from the preferences of policy makers and constituents diverging, but rather from the manner in which past actions change bargaining positions. As such, time inconsistency is a resistant problem that can leave maximizers able to commit to only suboptimal policies. In this section, we’ll investigate how monetary authorities can govern themselves in a manner that expands this set of commitments. Our model of the monetary authority’s problem (see Chapter 9) highlights a potentially fruitful direction – that is, implement governance features that either increase the cost of acting on inﬂationary incentives (to choose = 1) or discourage the maintenance of such an incentive in the ﬁrst place. To the extent that authorities successfully implement such features, constituents will rationally expect the socially optimal policy (for example, = 0 in Chapter 9’s model), and monetary authorities will fulﬁll that expectation. An interesting insight here is that maximizers can do better by increasing the cost of pursuing their incentives, or even fundamentally changing those incentives! Rules and Optimal Contracts Laws (and institutions more generally) can enhance credibility by raising the cost and lowering the beneﬁt from deviating from a given policy. (Drazen, 2000: 134, italics in original) 102

Organizing the production of monetary services

103

As the title of Kydland and Prescott’s (1977) seminal paper suggests, a rule-bound monetary authority can rationally choose policies that dominate those of a discretionary authority (even a benevolent one!). Consider, for example, a rule that would proscribe Chapter 9’s monetary authority from ‘inﬂating’ (that is, choosing the action = 1). Confronted with this rule, our authority might have been constrained to choose policy in a manner that induces a superior equilibrium – for example, one where authorities choose to not inﬂate and constituents rationally expect this choice. Flexibility and enforceability We emphasize ‘might’ because rules encounter at least two diﬃculties in mitigating time inconsistency. The ﬁrst stems from ‘ﬂexible rules’ being able to dominate rigid ones. To be sure, rigid rules can erase a monetary authority’s inﬂationary bias. But they also preclude an authority from optimally responding to economic shocks. In principle, then, ﬂexibility can enhance welfare. But allowing for ﬂexibility also exacerbates the enforcement problem. A rule-breaking authority might, for example, masquerade as one that optimally draws on ﬂexibility to dampen economic ﬂuctuations. More generally, as ﬂexibility under the rules grows, so do opportunities for rule breakers to disguise their actions. While ﬂexibility enjoys beneﬁts, then, it also creates costs in terms of forgone detection capacity. Who rules the ruler? The second problem stems from the dilemma of ‘who rules the ruler?’ If authorities are maximizers, then we should be skeptical about their incentive to obey rules for which they also act as enforcers. Indeed, breaking the rules can be attractive (for example, doing so might improve one’s election prospects) while punishing oneself is not (at least not immediately). In this manner, the ‘rules solution’ to time inconsistency pushes back the question of how monetary authorities can enhance credibility to the question of how political agents can credibly enforce rules. Repetition and Reputation Our investigation of the time inconsistency problem, and the potential for formal rules to oﬀer relief from it, implicitly rests so far on the assumption that monetary authorities and constituents meet once – they do not engage in ‘repeated interaction.’ Recall, for example, the games we considered in Chapter 9. There, even allowing for dynamics, players enjoyed only ‘one shot’ at each game.

104

Governing money

We may thus be interested in whether time inconsistency remains problematic when constituents repeatedly interact with monetary authorities. Let’s investigate, then, whether time inconsistency is an artifact of our simplifying assumption that players only meet over bounded lengths of time. Oﬀering players the prospect of inﬁnitely repeated interaction makes feasible a much richer set of equilibria than what emerged from our one-shot games. Notice that, if players suﬃciently value the future, they can threaten to punish their opponent for deviations from the optimal path, and these threats can be credible. In this case, the present cost that players incur to punish others is less than the consequent future beneﬁts (since those beneﬁts do not receive an overly heavy discount). A monetary authority that gives suﬃcient weight to future punishments may thus optimally forgo any short-run beneﬁts of inﬂating. In addition, repetition facilitates the building of reputations. Notice that ‘reputation’ is a vacuous notion for our one-shot games. But in repeated games, players’ actions can reveal information about their otherwise hidden ‘types.’ To develop this insight, suppose that monetary authorities are endowed with one of two immutable types – that is, one that maintains a strong taste for price stability and the other that does not. Suppose also that this information is private – only the monetary authority knows its true type. Then, by repeatedly sending ‘costly signals’ (for example, not inﬂating, or playing the action = 0 in our Chapter 9 model), a monetary authority can build a reputation as an inﬂation hawk. Indeed, an authority whose type is truly ‘inﬂation-averse’ will see the action of ‘not inﬂating’ as being dominant. To the extent that constituents continue to expect inﬂation, they will experience the inferior payoﬀ associated with the actions ( = 0, e = 1). Maintaining an expectation for inﬂation (e = 1) would be irrational in this case (e re). The problem of multiplicity At least on its face, the prospect of repeatedly interacting with a monetary authority oﬀers an escape from the dismal implications of one-shot interaction. But simply allowing for repetition need not induce a superior equilibrium. First, rather than pushing an economy toward superior outcomes, repeating a one-shot game can push an economy toward any outcome. To be sure, as players’ valuations of the future increase, they become increasingly patient, and the set of credible punishments thus grows. But as this set grows, so does the set of outcomes that can be sustained in equilibrium. In the limit, ‘anything can go.’

Organizing the production of monetary services

105

The problem of mimicking A second potential problem for our repetition argument is that maximizers can maintain incentives to masquerade as types other than their own – that is, they can strategically manipulate their reputations (see Backus and Driﬃll, 1985). Realizing, for example, that a reputation for being an inﬂation hawk can induce constituents to expect price stability, even authorities that do not maintain a strong aversion to inﬂation want to appear like they do. Indeed, to the extent that a ‘non-hawk’ authority successfully mimics inﬂation hawks in this regard, it can enjoy the beneﬁts of inﬂating (since constituents will have expected no inﬂation). Delegation and Insulation The ﬁnal, and perhaps most persuasive, governance ‘solution’ that we’ll consider is delegating monetary authority to an insulated bureaucracy. The idea here is a general (though paradoxical) one – a maximizer can make itself better oﬀ by delegating authority to an imperfect agent and credibly promising not to circumvent that delegation. Schelling (1980 [1960]: 143) oﬀers a seminal insight to this possibility:1 Just as it would be rational for a rational player to destroy his own rationality in certain game situations, either to deter a threat that might be made against him and that would be premised on his rationality or to make credible a threat that he would not otherwise commit himself to, it may also be rational for a player to select irrational partners or agents.

Vickers (1985: 138) went on to make the following observation: If control of my decision is in the hands of an agent whose preferences are diﬀerent from my own, I may nevertheless prefer the results to those that would come about if I took my own decisions.

To see this implication, suppose that a ﬁrm enjoys market power, but faces the prospect of entry from a competitor. To protect its proﬁts, the incumbent ﬁrm’s management might threaten to sharply drop prices upon the competitor’s entry. But such a threat would lack credibility if incumbent management takes proﬁt maximization as its objective – that is, if the ﬁrm’s owners hired managerial agents whose objectives align with their own. By hiring ‘imperfect’ managerial agents (for example, managers that focus on maximizing market share), the ﬁrm’s owners might better protect their market power and thus realize greater proﬁts than would have been available with a proﬁt-maximizing manager.2 Gibbard (1973) and Satterthwaite (1975) show that the beneﬁts from this sort of strategic delegation are robust to the particulars of Schelling’s (1980

106

Governing money

[1960]) and Vickers’ (1985) examples. Roughly, the ‘Gibbard-Satterthwaite theorem’ says that truthfully revealing one’s preferences is almost never a dominant strategy (there almost always exist opponent-strategies that encourage players to act in a less than ‘truthful’ manner). But notice that delegating authority to a perfect agent can reveal the principal’s true preferences. And since making such a revelation is almost never ‘dominant’ (the best action that one can play, regardless of what one’s opponent does), then hiring a perfect agent can almost never be dominant. Conservative banker Even if it were possible for the principal to appoint an agent with the same objectives, it may not be optimal to do so. That is, in delegating the decision over a speciﬁc policy objective, a government may come closer to achieving its preferred objective by having an agent aim for a diﬀerent objective! (Drazen, 2000: 141, emphasis in original)

In a series of inﬂuential papers, Rogoﬀ (1985, 1987) showed that delegating monetary authority to a ‘conservative’ banker can mitigate the time inconsistency problem. Here, society hires an agent whose preference for inﬂation is considerably weaker than that of the median voter.3 By endowing this agent with the capacity to manage monetary aﬀairs, society can essentially tie its hands against acting on time-inconsistent preferences that would otherwise induce an inferior, discretionary equilibrium.4 Central bank independence To enjoy the superior ‘hand-tying’ equilibrium, however, society must not only delegate its monetary authority, it must do so in a manner that makes circumventing that delegation costly. Delegation per se does not change society’s preferences. Indeed, while the appointed banker might be conservative, the preferences of relevant constituents will not have changed. Hence, if the cost of circumventing the delegation of monetary authority is low enough, then society will continue to act on its time-inconsistent preference. Delegation without insulation does little to move a society away from the inferior, discretionary equilibrium.5

FISCAL VERSUS MONETARY AUTHORITIES The central bank independence (CBI) literature tends to focus on insulating monetary authorities from pressures to mitigate ﬂuctuations. However, a second type of insulation also deserves attention – insulation from ﬁscal pressures. The time inconsistency problem highlights how electoral

Organizing the production of monetary services

107

pressures can induce inferior equilibria via the time inconsistency dynamics. But also recall our model from Part II where a clearinghouse enhances welfare by passively exchanging a nominal currency for real goods and services. To achieve this end, the clearinghouse has to forgo real tax revenues from strategically ‘printing’ money. Optimal CBI thus appears to require insulation from pressures associated with ﬁscal issues, in addition to those associated with economic ﬂuctuations.6

CBI, INFLATION AND REAL ACTIVITY Drazen (2000, Chapter 5) surveys the literature on CBI and economic outcomes. In doing so, he highlights this literature’s tendency to agree on two ﬁndings. First, CBI negatively correlates with inﬂation (more independence is associated with less inﬂation), and this correlation is robust to how independence is measured. In addition, Drazen cites countries (for example, Belgium, the Netherlands) where this correlation persists despite CBI coinciding with high levels of political instability and national debt. This persistence increases conﬁdence that the CBI–low inﬂation correlation evidences causation.7 Second, contributors to this literature ﬁnd that low inﬂation from CBI does not increase output-volatility. This ﬁnding weakens concerns that price stability requires passivity in the face of shocks that active policy might productively address (for example, a liquidity crisis after terrorist attacks). Taking a more structural approach to developing this second type of evidence may be instructive. Recall, for example, our Chapter 8 discussion of Lucas (2003) and Barlevy (2004). In Lucas’s model, economic ﬂuctuations (evaluated at their current magnitudes) do not weigh much on economic well-being. But this model excludes a channel through which ﬂuctuations can inﬂuence growth. And Barlevy (2004) shows that, once we consider such a channel (for example, investment), dampening ﬂuctuations might signiﬁcantly improve welfare. Consequently, if a cost of central bank independence is, in the spirit of Rogoﬀ (1985), a heightened insensitivity of monetary authorities to the consequences of aggregate shocks, then we should observe a positive relationship between independence and aggregate ﬂuctuations. In turn, to the extent that Barlevy (2004) identiﬁes an empirically relevant channel through which ﬂuctuations inﬂuence economic growth, we should observe a negative relationship between such ﬂuctuations, investment and, ultimately, growth. Alesina and Summers (1993) oﬀer evidence that CBI and inﬂation share a negative relationship, but fail to ﬁnd a reduced form relationship between

108

Governing money

CBI and real ﬂuctuations or growth.8 On its face, this evidence increases conﬁdence in Lucas’s (2003) modeling assumptions – namely, that mechanisms for propagating ﬂuctuations into growth are suﬃciently unimportant to warrant exclusion from formal investigations. However, Alesina and Summers built their evidence from methods that ignore the potential for unobserved heterogeneity (and endogenous regressors more generally) to bias inference.9 They thus acknowledged that their results ‘are not conclusive’ and called for a more careful investigation (Alesina and Summers, 1993: 159).

NOTES 1. Ideas like this one helped Thomas Schelling win the 2005 Nobel Prize in Economics. 2. This example motivates Vickers’ (1985) analysis. 3. In Chapter 12, we’ll see how delegating some of the Federal Reserve System’s authority to ‘district banks’ is consistent with the prescription. 4. Like any other beneﬁt, however, those associated with delegation to a conservative banker come with a cost (for example, agency costs). In particular, Rogoﬀ (1985) shows that, while delegation can reduce inﬂation, it also sub-optimally raises the variance of employment when supply shocks are large. 5. Driﬃll and Rotondi (2003) show that re-appointment costs can facilitate commitment in this regard. Persson and Tabellini (1999) argue that forgone reputational beneﬁts represent another such cost. In this light, delegation appears to accomplish more than relocating the time inconsistency problem. 6. Dixit and Lambertini (2003) investigate the nature of this necessity, while Calvo and Guidotti (1993) consider how a consequent loss of ﬂexibility might move an economy away from an optimal inﬂation tax. 7. See, however, Posen (1993, 1995, 1998), who argues that CBI may simply reﬂect deeper societal forces (for example, those associated with the distributional consequences of inﬂation). 8. Grilli et al. (1991) oﬀer corroborating evidence. Other authors who examine how the organization of central banking relates to economic outcomes include Cukierman (1995), Eijﬃnger and de Haan (1996) and Falaschetti (2002b). 9. Posen (1995) argues that measures of CBI could reﬂect congealed preferences, and addresses this source of endogeneity by controlling for ‘ﬁnancial sector opposition to inﬂation.’ In doing so, he ﬁnds that the relationship between CBI and inﬂation becomes insigniﬁcant.

12.

The case of the ‘Fed’

INTRODUCTION Our previous chapters show how an optimal policy can be time inconsistent in the sense that its normative properties change over time. They also argue that delegating authority over relevant policy domains can mitigate this problem. To be sure, delegation does not obviate the time inconsistency problem. Rather, it enhances welfare by increasing the cost for political actors to act on time-inconsistent objectives. In this light, delegation appears to be a good ‘solution’ only to the extent that otherwise opportunistic political principals cannot circumvent the delegation of policy-making authority. The US Federal Reserve System (the ‘Fed’) exhibits several features that increase the cost of circumventing delegation. We’ll examine these features in the present chapter with the objective of better understanding the paradoxical phenomenon that insulating a delegated authority can enhance welfare by diminishing the ‘accountability’ of proximate decision-makers.

DELEGATION The Federal Reserve System includes 12 district banks, each of which has a president, and the Federal Reserve Board (FRB).1 District bank presidents receive nominations from their respective boards of directors and conﬁrmations from the FRB to renewable ﬁve-year terms. The FRB, on the other hand, consists of seven ‘governors,’ each of whom receives a nomination from the President and conﬁrmation from the Senate to nonrenewable 14-year terms.2 On a rotating basis, four district bank presidents combine with all seven Board members and the New York district president to become the Federal Open Market Committee’s (FOMC) voting membership. The FOMC, in turn, directly determines ‘monetary policy.’3

109

110

Governing money

INSULATION In addition to receiving delegated authority over monetary policy, the Fed enjoys considerable insulation from political pressures. For example, it funds operations from interest earned on a portfolio of government securities.4 But rather than being subject to the appropriations process, this portfolio largely accumulates from the Fed’s ‘open market operations’ (that is, the trading of US Treasury securities through which the Fed implements monetary policy). As a consequence, Congress cannot (easily) circumvent its delegation of monetary authority by strategically manipulating relevant bureaucratic budgets. This organizational feature thus diminishes an important channel through which more ‘accountable’ political agents might otherwise inﬂuence monetary policy. The Fed receives additional insulation from its governors serving relatively long, non-renewable terms. One governor’s term expires every other year. Each of the seven governorships thus lasts for 14 years. Moreover, governors who serve a full term cannot be reappointed. Consequently, not only does delegation to the Fed receive protection from governors’ terms lasting longer than do those of potentially inﬂuential political overseers, it also receives protection from those agents lacking another familiar bureaucratic control mechanism – that is, strategically manipulating the prospect of reappointment.

PERSISTENCE OF POLITICAL INFLUENCE Despite this insulation, we should recognize that relevant institutions do not (and indeed cannot) completely neutralize opportunistic political forces. Rather, potentially important channels remain through which interested actors can breach the delegation of monetary authority. For example, while the prospect of reappointment seldom sways Fed governors, the original appointment process may still be inﬂuential. Interested principals might, for example, aﬀect monetary policy by supporting the appointment of governors whose policy preferences are close to their own. In this manner, the policy decisions of a subset of the FOMC (governors more so than district presidents) may be directly subject to ‘ex-ante’ inﬂuence. The prospect of Congressional oversight might also be salient since, as Willett and Keen (1990: 17) argue, while changing the ‘rules of the game’ may be costly for Congress, the prospect is large enough for Fed oﬃcials to be considerate.5 Recognizing the potential for inﬂuence to ﬂow through these ‘residual channels,’ scholars have found evidence that strategic appointments to the

The case of the ‘Fed’

111

FRB and Congressional oversight of the FOMC signiﬁcantly inﬂuence monetary policy. This evidence also suggests, however, that the nature of such inﬂuence is sensitive to the party of relevant political agents (see Chappell et al., 1993; Tootell, 1996; Caporale and Grier, 1998). As such, it implies that time inconsistency is problematic for only a subset of political principals (for example, Democrats, but not Republicans, maintain an inﬂationary bias). Recall the implication of our time inconsistency game, however – to the extent that political agents maintain the same objective as do electoral principals, all agents will maintain an inﬂationary bias. Here, instead of varying with who is in oﬃce, the prospect of realizing an inﬂationary bias varies with the cost for political agents to act on that bias. Falaschetti (2002b) developed evidence to this eﬀect. In particular, he showed that elected political agents’ (for example, presidents, senators) capacity to inﬂuence US monetary policy can be sensitive to whether the same party controls both the executive oﬃce and Senate. When either party controls both branches, ‘looser’ FRB governors receive appointments, and incumbent governors tend to favor looser policy. District bank presidents, on the other hand, appear to be relatively insulated from these pressures. This extension is important because it highlights how potentially reproducible organizational features,6 such as increasing the number of ‘veto players’ through which an appointment must pass, can facilitate the insulation of monetary policy from opportunistic political forces.

NOTES 1. District banks reside in Boston, New York, Philadelphia, Cleveland, Richmond (VA), Atlanta, Chicago, St Louis, Minneapolis, Kansas City, Dallas and San Francisco. 2. Board members cannot be ﬁred, forced to resign, or voted out of oﬃce (Waller, 1992: 415). 3. The FOMC meets eight times per year to determine monetary policy. ‘Monetary policy’ refers to actions that ‘inﬂuence the availability and cost of money and credit to help promote national economic goals’ (the Federal Reserve Board, accessed 21 September 2004 at http://www.federalreserve.gov/FOMC/default.htm). 4. Residual earnings from this portfolio represent seignorage in the sense that the Fed transfers them to the Treasury (the US ﬁscal authority). 5. This potential for ex-post inﬂuence appears to gain strength from the incentives of oversight committee members. For example, Gilligan and Krehbiel (1987) argue that, if not for the ability to parlay strategic advantages into disproportionate inﬂuence, legislators would have little incentive to spend time on committee work. 6. Notice that having one party or the other control a legislature or president is not an easily reproduced institutional feature.

PART V

Intermediation, governance and economic performance

We learned in Parts II–IV that a nominal asset, ‘money,’ can inﬂuence real activity (levels of wealth and ﬂuctuations therein), and thus began to see why economic performance is sensitive to how societies govern their monetary services. These services can, for example, expand consumption possibilities by economizing on the cost of transacting at a point in time – the cost of trading in spot markets. To realize this potential, however, a society must organize its monetary authority in a manner that discourages strategically acting on changes in bargaining positions to either raise real taxes or dampen economic ﬂuctuations. We’ll extend this analytical approach in the following chapters to understand how societies produce ‘ﬁnancial intermediation services,’ and why strong economic performance further depends on whether these services govern individually rational actions in a manner that expands social opportunities. Recall that we motivated our examination of money by critically evaluating an assumption on which a principles-level model of households rests – transacting in spot markets negligibly exhausts resources so that bartering is costless. In motivating our examination of ﬁnancial intermediation, we’ll examine a similar assumption on which a principles-level model of the ﬁrm rests – transacting over time negligibly exhausts resources. A principles-level model characterizes ﬁrms as maximizing proﬁts by choosing inputs from an unbounded domain (see our review of the ﬁrm’s problem in Chapter 3). But just as households incur costs when trading endowments for more preferable consumption bundles (see Part II), ﬁrms incur costs when attempting to employ production factors that can increase proﬁts. We’ll thus see that, just as money and its governance are important for mutually beneﬁcial trades at a point in time, ﬁnancial intermediation

114

Intermediation, governance and economic performance

and its governance are important for mutually beneﬁcial trades across diﬀerent periods. We’ll develop these insights by extending our simple model of the ﬁrm to see what happens when ﬁrms cannot freely choose production factors. This investigation will teach us that, when ﬁrms are ‘cash constrained’ (that is, they lack direct control over resources that are necessary to hire production factors), they have an incentive to borrow from those who are ‘resource rich,’ but ‘project poor.’ (We’ll characterize economic actors as being ‘project poor’ if employing resources in immediately accessible projects would generate a smaller product than is technically possible from external projects.) Moreover, to the extent that borrowing ﬁrms can credibly share consequent increases in output, resource-rich but project-poor ﬁnanciers have an incentive to lend resources. A principles-level model implicitly assumes that governing this type of exchange is costless. This assumption cannot hold, however, in a world of maximizers. Notice that maximizing borrowers have an incentive to upwardly bias their disclosure of a project’s expected productivity. In addition, once these borrowers obtain a capital market’s proceeds, they have an incentive to pursue projects that are more risky or less productive than those in which their own resources might be invested. Unless an economy can mitigate these diﬃculties (‘adverse selection’ and ‘moral hazard,’ respectively), prospective borrowers and lenders will rationally forgo otherwise mutually beneﬁcial trades. Just as we learned in Part II that bartering costs foreclose mutually beneﬁcial trades when welldeveloped monetary systems are absent, we’ll learn in Part V that the cost of transacting across time discourages mutually beneﬁcial trades when ﬁnancial development is lagging. Financial development can thus expand a society’s economic opportunities. But while such development creates widespread beneﬁts, we should also appreciate its capacity to create politically powerful losers. This political-economic phenomenon has an analog in forces that work against eﬃciency in the monetary sector (see, for example, Part IV). Indeed, just as politically feasible mechanisms for producing monetary services must respect distributional constraints, so must politically feasible mechanisms for producing ﬁnancial services. These constraints can be obstinate, however, since even an improved education about ﬁnancial services may not change a) individual preferences over the distribution of more general progress or b) technologies with which interested individuals produce political pressure. We’ll thus conclude that, while well-developed systems of money and ﬁnancial intermediation can facilitate strong economic performance, even the most advanced economies must resist political forces that constantly work against expanding opportunities.

13.

Asymmetric information

INTRODUCTION Consider, once more, our most simple model of the ﬁrm (see Part 1). In this model, ﬁrms maximize proﬁts by employing production factors from an unbounded set. By positing that choice-sets are unconstrained, however, this model implicitly assumes that production factors costlessly move to their highest valued use, and thus abstracts from the potential for ﬁnancial markets to inﬂuence economic performance. Just as we saw in Part II that goods and services do not costlessly move to their highest valued use (that is, bartering is costly), we’ll see in the present chapter that ﬁnancial capital does not costlessly move to its highest valued use. Indeed, just as bartering deters even mutually beneﬁcial trades by creating frictions in goods markets, asymmetric information discourages such trades by introducing frictions to ﬁnancial markets. Likewise, just as money can enhance economic welfare by mitigating costs associated with bartering, ﬁnancial intermediation and corporate governance can enhance welfare by mitigating costs associated with asymmetric information. A slightly richer model of the ﬁrm will help us develop this insight. Suppose that, instead of choosing from an unbounded set of factors, ﬁrms face a ‘cash constraint’ where limited internal resources preclude the hiring of factors that would maximize proﬁts. In Figure 13.1, for example, the ﬁrm’s current assets might limit the employment of inputs to X , even though hiring inputs up to X* would increase proﬁts. Even a cash-constrained ﬁrm can be a maximizer, however. Consequently, if it could expand production (and thus increase proﬁts) by borrowing resources, it would. Moreover, if it could employ those resources more productively than could a prospective lender, then the lender would be better oﬀ employing its resources in the ﬁrm and receiving a share of the increased total product (that is, the diﬀerence between proﬁt levels ‘1’ and ‘0’ in Figure 13.1). Both the cash-constrained ﬁrm and project-constrained lender can thus enjoy superior economic opportunities by having ﬁnancial capital ﬂow to its highest productivity use. A principles-level model implicitly assumes that this migration occurs without consuming resources. We’ll see in the present chapter, however, that even though trading capital can create mutual beneﬁts, prospective 115

116

Intermediation, governance and economic performance

Outputs Profit (1) + Input price * X Production possibilities frontier Y* Profit (0) + Input price * X Profit (1) Y– Profit (0) Figure 13.1

X–

X*

Inputs

Cash-constrained proﬁt maximization

beneﬁciaries must overcome considerable frictions. Notice that maximizing borrowers have an incentive to both misrepresent their projects’ potential productivity (that is, ‘adversely select’ themselves into the pool of attractive borrowers) and employ borrowed capital in a less than optimal manner (that is, act in a ‘morally hazardous’ manner). Unless institutions exist to discourage such actions, lenders will thus hold back on supporting what might otherwise be proﬁtable projects.

ADVERSE SELECTION (THE LEMONS PROBLEM) Left unchecked, two types of information asymmetries will leave ﬁrms with a ‘binding’ cash constraint – that is, having to choose factors from a bounded set, like X in Figure 13.1. We’ll refer to this ﬁrst type of asymmetry as ‘adverse selection’ and the second as ‘moral hazard.’ Adverse selection refers to a consequence of information being asymmetrically distributed before a transaction takes place. To see how this asymmetry can discourage mutually beneﬁcial trades (for example, those that would relax Figure 13.1’s cash constraint), let’s consider a celebrated example from the automobile market. Suppose that only two types of automobiles exist, ‘high quality’ and ‘low quality’ (or ‘lemon’), and to ease reference, let’s label these types h and l,

Asymmetric information

117

respectively.1 In addition, let’s assume that the ‘true value’ of h-types is $100, the true value of l-types is $0, and there exists an equal number of hand l-types. Finally, notice that an automobile’s owner likely enjoys better information about his or her auto than does a prospective buyer – information asymmetrically distributes itself before the transaction takes place. Hence, let’s also assume that the seller’s information advantage lets him or her know the true value of any automobile that he or she might sell, while it lets the buyer see only the average value of automobiles that a market makes available. From this set-up, we can see how information asymmetries promote the ‘adverse selection’ of l-types into the market (and thus discourage what could be mutually beneﬁcial trades of h-types). To begin, consider how buyers calculate the average value of automobiles and how suppliers respond to that valuation. If the distribution of marketable automobiletypes is independent of how prospective buyers value the ‘average automobile,’ then the average value of tradable automobiles would be $50 in our example (since we assumed that an equal number of h- and l-types exists). But also notice that this independence opposes the axiom of maximizingbehavior. Indeed, if automobile demanders are willing to pay $50 for any automobile that a seller presents, then owners will only sell l-types. Under such conditions, the market for h-types must unravel. Notice that selling an h-type for $50 creates a $50 loss for its owner. But the owner is a maximizer and ‘knows’ the type of car that he or she is attempting to sell. Consequently, only owners of l-types are willing to sell autos when demanders are willing to pay $50. Moreover, if only l-types ultimately make their way to the market, then the average value of automobiles that are available for trade is not $50, but rather $0. In the end, demanders are willing to pay $0 for any car that makes its way to the market and suppliers are only willing to sell lemons! In this light, the popular apprehension about used cars’ quality appears well-founded. But used cars are not special in making adverse selection problematic. Indeed, adverse selection creates diﬃculties in numerous markets, including those for labor, insurance, and yes, ﬁnancial capital. For example, notice that just as a used car’s owner likely enjoys superior information about that car’s quality, a ﬁrm’s manager likely enjoys superior information about the ﬁrm’s prospects. In particular, managers can likely access superior information about how productive are new inputs (that is, the production function’s slope when evaluated at Figure 13.1’s cash constraint X ). This informational advantage, in turn, can encourage ‘low quality’ projects to adversely select themselves into capital markets. As in the automobile example, relatively well-informed sellers will not trade highly remunerative

118

Intermediation, governance and economic performance

ﬁnancial assets (rights to share in the product of newly employed inputs) at prices that reﬂect only average quality. To conclude this discussion, let’s be careful about why this insight into the lemons problem is so important. The lemons problem does not condemn markets to trading only lemons. Rather, it shows that institutions are necessary to facilitate mutually beneﬁcial trades when persistent pressures exist for markets to unravel. In our following two chapters, we’ll investigate how societies can organize ﬁnancial intermediaries and govern corporations to enjoy the expanded opportunities that such trades make feasible.

MORAL HAZARD Before moving in this direction, we conclude the present chapter by establishing how a second type of asymmetric information can create (welfare-reducing) capital market frictions. Adverse selection refers to how information asymmetrically distributes itself before a transaction takes place. But information remains asymmetric even after a transaction takes place, and can thus continue to discourage mutually beneﬁcial trades. We’ll refer to this second type of asymmetry as ‘moral hazard’ and see that it too can leave ﬁrms with binding cash constraints. Like adverse selection, moral hazard aﬀects myriad economic sectors. Consider, for example, the insurance industry and notice that, after an individual buys insurance, the insurer cannot perfectly monitor its client’s actions. Insurance-buyers have better information than do sellers after the transaction takes place. Moreover, equipped with this informational advantage and insulated from insured losses, buyers will act in a more risky manner than they would have without coverage. For example, to the extent that monitoring others’ driving habits is costly, maximizers drive faster with automobile insurance than without it. Insured drivers beneﬁt from driving faster (for example, the value of time that is saved), but pass some of the associated costs onto insurers (for example, the value of an expected claim). A similar phenomenon plays out in ﬁnancial markets. Just as insured clients have an incentive to play morally hazardous actions (actions that strategically exploit ex post informational advantages), borrowers have an incentive to pass their actions’ costs onto ﬁnanciers. Figure 13.2 illustrates how opportunistically investing in risky projects works toward this objective. Borrowers can enjoy the upside potential of projects while passing downside risks oﬀ to lenders. Moreover, to discourage borrowers from acting on this incentive, lenders must forgo alternative opportunities. Absent institutions that eﬃciently weaken borrowers’ opportunistic incentives, or constrain

119

Figure 13.2

Moral hazard

Demander of loanable funds ignores increased cumulative probability of realizing a ‘large’ loss

0

Probability

‘Safe’ payoffs

Project payoff

‘Risky’ payoffs

Demander of loanable funds enjoys increased cumulative probability of realizing a ‘large’ payoff

120

Intermediation, governance and economic performance

the ability to act on those incentives, ﬁnanciers will thus forgo otherwise mutually beneﬁcial transactions. Here, again, the problem of asymmetric information can leave ﬁrms with a binding cash constraint, and thus shrink an economy’s opportunities.

NOTE 1. A recent Nobel Prize-winner, George Akerlof from the University of California at Berkeley, is frequently cited as bringing the problem of asymmetric information to social scientists’ attention. Indeed, Akerlof’s (1970) seminal work resulted in the adverse selection problem being referred to as the ‘lemons problem.’

14.

Financial intermediaries and their governance

INTRODUCTION Recall from Part II that the cost of bartering discourages goods and services from moving to their highest valued uses. And while ‘money’ can economize on these costs, a sound monetary system must continually resist political forces that threaten eﬃciency (for example, those associated with time inconsistency problems). By carefully identifying these inﬂuences, we set the stage for investigating how societies organize their monetary authorities and how these organizational mechanisms relate to levels of wealth and ﬂuctuations therein. Chapter 13 similarly sets the stage for our investigation of how societies organize their ﬁnancial markets, and why success on this margin strongly inﬂuences economic performance. Just as Part II highlights the costs of barter to motivate the importance of money, Chapter 13 highlights the costs of ‘lemons’ adversely selecting themselves into ﬁnancial transactions, or individuals acting in a morally hazardous manner, to motivate the importance of ﬁnancial intermediation and corporate governance. In that chapter, we saw that reducing information asymmetries can expand economic opportunities for both suppliers and demanders of ﬁnancial capital. Strong incentives can thus exist for market participants (suppliers and demanders of loanable funds) to mitigate asymmetries in a cost-eﬀective manner. In the present chapter, we’ll evaluate how the incentive for proximate suppliers to mitigate asymmetries relates to the organization of ﬁnancial intermediaries, and what this incentive implies for the eﬃcacy of ﬁnancial regulation.1 This investigation will show how suppliers can facilitate the productive ﬂow of ﬁnancial capital by encouraging demanders to reveal otherwise hidden information. It will also carefully identify opportunities for regulation to improve upon such market discipline. Finally, we’ll see that while ﬁnancial intermediation can expand economic opportunities, it can also aﬀect the distribution of realized beneﬁts. These distributional consequences can (and frequently do) create political forces 121

122

Intermediation, governance and economic performance

that discourage societies from pursuing eﬃcient ﬁnancial mechanisms. While intermediaries may enhance economic welfare by mitigating information asymmetries, fundamental political obstacles can thus restrict societies to economically inferior outcomes.

ADVERSE SELECTION AND FINANCIAL STRUCTURE Intermediaries employ a number of ‘mechanisms’ to mitigate information asymmetries and thus encourage ﬁnancial capital to migrate toward more productive uses. While varied in their empirical realization, these mechanisms share a fundamental property – they encourage maximizers to truthfully reveal what would otherwise be hidden information. In short, successful mechanisms ask prospective borrowers to take actions that are cost-prohibitive for lemons, thereby creating an observable ‘separation’ between lemons and non-lemons. At least in principle, mechanisms can treat the source of adverse selection – that is, the propensity for ‘good’ and ‘bad’ types to pool themselves into apparently homogeneous groups. To see how such mechanisms work in practice, let’s consider the widespread requirement that borrowers post collateral. Suppose that you manage an intermediary in which two potential borrowers, A and B, present themselves. Suppose also that agent B maintains a greater appetite for risk than does A, but information about risk-appetites is ‘private’ (you cannot readily observe it). This model lets us see how asking for collateral encourages demanders of loanable funds to truthfully reveal their risk-appetites. Notice that, ceteris paribus, the probability that the high-risk agent (B) ultimately defaults exceeds that of the low-risk agent (A). Consequently, B’s expected cost of posting collateral is greater than that for agent A – that is, (PB C) (PA C) where Pi denotes the probability that agent i defaults and C denotes the cost of defaulting. Low-risk agents can thus distinguish themselves by being more likely to accept collateral requirements than are highrisk agents. At least on their face, mechanisms such as collateral requirements can appear punitive. Prospective borrowers frequently argue, for example, that they wouldn’t demand loanable funds if they already commanded the necessary resources for collateral. We see now, however, that such mechanisms can promote eﬃciency. To be sure, empirically interesting mechanisms like collateral must err in characterizing some truly productive borrowers as lemons. However, they can more than oﬀset such mistakes by identifying

Financial intermediaries and their governance

123

would-be lemons, and thus facilitate mutually beneﬁcial trades in markets that can unravel from adverse selection.

MORAL HAZARD AND FINANCIAL STRUCTURE The contract between shareholders and the manager leaves the latter a lot of discretion because the manager has the knowledge and the ability to run the company. The consequence is that the manager may engage in all kinds of behavior that are detrimental to the ﬁrm: pure theft (for example, by setting up a company and using transfer pricing to appropriate funds); enjoying private beneﬁts of control (perks, pet projects, empire building, and favoring friends and family); entrenchment (to protect the private beneﬁts of control); exerting insuﬃcient eﬀort; and taking biased decisions (too much or too little risk taking, for example). (Vives, 2000: 4)

As Vives (2000) suggests, a principal-agent problem emerges from separating a ﬁrm’s ownership (by stockholder principals) from its control (by managerial agents). Indeed, given our axiom of maximizing-behavior, we expect managers to exploit information asymmetries for their own beneﬁt – that is, to play ‘hidden actions’ that disproportionately beneﬁt themselves at other stakeholders’ cost. In Chapter 15, we’ll look at how markets can discipline corporations (demanders of ﬁnancial capital) to check this type of moral hazard. In the present chapter, we want to understand how suppliers of ﬁnancial capital can structure transactions to check the consequences of this type of asymmetric information. Recall that moral hazard emerges from the incentive for demanders to take on ‘too much’ risk, or play unproductive actions, after acquiring the proceeds of a ﬁnancial market transaction. Here, managers can participate in a project’s net beneﬁts while limiting their exposure to any losses. To mitigate this problem, intermediaries may require borrowers to maintain a minimum level of net worth, and thus expose demanders of ﬁnancial capital to their activities’ downside risk.2 They may also impose ‘restrictive covenants’ that increase the marginal cost for demanders to act in a morally hazardous manner.3 Even the cost of restructuring debt agreements can help in this regard. For example, while institutions that simplify debt restructurings can reduce ﬁnancial distress, they also encourage moral hazard by reducing the expected cost of such behavior. Eichengreen and Mody (2004) ﬁnd interesting evidence of this eﬀect – in the presence of ‘renegotiationfriendly loan provisions,’ borrowing costs are signiﬁcantly lower for ‘creditworthy’ demanders than they are for less worthy borrowers. This relationship supports the hypothesis that the increased risk of moral

124

Intermediation, governance and economic performance

hazard among less worthy demanders oﬀsets the lower cost of settling realized renegotiations.

INTELLECTUAL PROPERTY PROTECTION AND FINANCIAL STRUCTURE In principle, ﬁnancial intermediaries can bolster economic performance by reducing information asymmetries and thus freeing loanable funds to ﬁnd more productive uses. To generate this expansion of economic opportunities, however, intermediaries must also privately beneﬁt.4 In addition to developing mechanisms for checking information asymmetries, then, an intermediary must organize itself with an eye toward protecting the product of its eﬀort (for example, information about the demanders of loanable funds). The observed tendency for intermediaries to ‘privately place’ loans reﬂects this organizational strategy. Notice that, because restricting informational beneﬁts to those who produced them is costly, the incentive for decentralized maximizers to mitigate the lemons problem can be weak. Indeed, faced with the prospect of producing information about borrowers and having that information become readily available to others, maximizers will attempt to free ride on each other’s eﬀorts.5 Intermediaries can structure themselves in a manner that reduces this problem. Suppliers of loanable funds, for example, frequently make those funds available via ‘privately placed’ debt (as well as ‘private equity’). In doing so, they insulate themselves from the disclosure requirements of public exchanges, and can thus internalize more of the beneﬁts that their governance services produce (see James, 1987; Lummer and McConnell, 1989; Metrick, 2007). This incentive goes far in rationalizing why privately placed debt constitutes an overwhelming share of even large ﬁrms’ external funds (Mayer, 1990; Houston and James, 2001).

ILLUSTRATION: BANKS AS DISTINCTIVE INTERMEDIARIES We now have a ﬁrmly grounded model of how intermediaries can check information asymmetries. To more clearly see what this abstraction implies for ﬁnancial organization, let’s evaluate how it explains the organization of an important type of intermediary – banks. This application will illustrate how the organization of banks (and related intermediaries) creates advantages for mitigating information asymmetries and thus why economic

Financial intermediaries and their governance

125

performance appears so sensitive to how societies produce and regulate banking services. The Firm’s Financing Decision Diﬀerent forms of ﬁnancing may not be equivalent when managers have inside information about a ﬁrm’s prospects (that is, when asymmetric information exists). Consider what happens when a ﬁrm announces its desire to issue new securities. On one hand, this announcement might signal a ﬁrm’s proﬁtable investment opportunities. Alternatively, it might represent the ﬁrm’s attempt to exploit an informational advantage (for example, managers, but not outsiders, may have information that earnings are less than expected). Market participants may thus impose a ‘lemons’ discount on the issuance of new securities. James and Wier (1988) oﬀer evidence that just such a problem plagues attempts at direct ﬁnance – markets react negatively to announcements of new security issues. An important question thus becomes how ‘good’ ﬁrms can separate themselves from such reactions when attempting to ﬁnance new projects.6 Bank Loans’ Informative Role The organization of banks can facilitate this separation. By repeatedly putting bankers in close contact with demanders of loanable funds, the process of ‘relationship’ banking can generate information that is not readily available to others. A bank’s decision of whether to lend may thus convey better information about a ﬁrm’s investment prospects and creditworthiness than would the ﬁrm’s own announcement to issue new securities. This theory of banks as ‘special’ generators of information creates several observable implications. For example, the value of a ﬁrm’s securities should rise when it reaches a bank loan agreement. James and Wier (1988) ﬁnd evidence of just such a reaction – client stock prices signiﬁcantly increase after banks announce new loan agreements. No such increase is evident, however, after non-bank intermediaries announce new agreements. Moreover, consistent with a willingness to pay for such ‘certiﬁcation services,’ banks appear to command signiﬁcantly higher (risk-adjusted) prices for loanable funds (Fama, 1985). Are Banks Better Evaluators of Information? Evidence that we’ve summarized so far does not diﬀerentiate between two hypotheses about the special role that banks play in capital markets – that

126

Intermediation, governance and economic performance

is, are banks simply better than other intermediaries at evaluating ﬁrms’ prospects, or do banks have access to information that other outsiders do not? Lummer and McConnell (1989) address this question by looking at how reactions to new and revised loan agreements diﬀer, and ﬁnding evidence of more information in announcements of revisions.7 This evidence is particularly strong for cancelled or reduced agreements where lenders initiate revisions (Lummer and McConnell, 1989: 107, Table 3). Lummer and McConnell (1989) thus conclude that banks have no comparative advantage in evaluating risky lending opportunities, but do have special access to private information that emerges from ongoing relationships.

GOVERNING FINANCIAL INTERMEDIATION In Parts II and III, we learned that well-developed monetary systems ease the way for goods to ﬁnd higher-valued uses, and then investigated in Part IV how an economy’s governance features should and do contribute to the production of monetary services. We now understand that economic wellbeing also depends on a society’s capacity to produce intermediation services, and are thus similarly interested in how those services should be governed and why societies frequently depart from our normative prescriptions. We conclude our investigation of ﬁnancial intermediation by addressing this important political economy question. Asymmetric Information, Systemic Risk and Prudent Bank Regulation If information is asymmetric between demanders and suppliers of ﬁnancial capital in general, then asymmetries must also exist between intermediaries and their ﬁnanciers (for example, depositors). Moreover, because this type of asymmetry can create risks for the ﬁnancial system (that is, risks that are ‘external to particular intermediaries’), market discipline may leave open welfare-enhancing governance opportunities. We’ll consider such opportunities in this section’s remainder. We’ll also uncover a discouraging analog to the governance of money – ﬁnancial regulators, like their monetary counterparts, face persistent pressures to pursue redistributive policies. In this light, regulatory crises appear capable of repeating themselves despite ‘lessons learned’ from past episodes. Adverse selection Following numerous ‘bank runs’ during the Great Depression, the US adopted a system of deposit insurance. To see this system’s economic rationale, suppose that you own uninsured deposits. Suppose also that a bank

Financial intermediaries and their governance

127

similar to yours experiences a ‘liquidity crisis’ – that is, demands to return deposits exceed the bank’s immediate capacity. Observing this crisis, you might rationally downgrade the expectation of your own bank’s integrity, and thus withdraw your own deposits. If enough depositors take such actions, however, then the original ‘crisis’ can spread to your and other banks – even if they are solvent. Moreover, if the cost of liquidating assets is high enough, then liquidity crises can turn into solvency crises. In this case, banks cannot make depositors whole, even if they sell all of their assets. Depression-era economies may have realized just such a scenario. The resulting loss of intermediation services, in turn, appears to have been an important factor in producing such a large economic downturn. Deposit insurance can check this source of economic ﬂuctuations. To see how, return to our example, but suppose that your deposits are insured. In this case, new information about another bank’s liquidity crisis is less likely to encourage a ‘run.’ After all, even if you suspect that your bank is less sound, the existence of deposit insurance reduces your exposure to realizing that expectation. In this manner, deposit insurance can enhance economic performance by checking the potential for information asymmetries to act as seeds for ‘ﬁnancial contagion’ in otherwise sound systems. Moral hazard This beneﬁt comes with several costs, however. For example, notice that deposit insurance not only checks the incentive for depositors to ‘run’ on banks, it also discourages them from monitoring banks’ investment activities. Indeed, to the extent that deposits are insured, maximizers will shop for the highest yields without regard to how risky are the underlying assets. In turn, intermediaries may enjoy expanded opportunities to play morally hazardous actions. Financial regulations, such as those that restrict the scope of an intermediary’s operations, require minimum levels of capital, and subject covered intermediaries to audits, may thus be necessary for a sound system of deposit insurance. Repeated Regulatory ‘Mistakes’ The regulatory prescriptions outlined above are well known. Nevertheless, even well-developed ﬁnancial systems repeatedly experience crises from ignoring them. This pattern suggests that persistent political forces not only work against ﬁnancial advances, they also constantly threaten development once it is achieved. Instead of relying on ‘lessons learned,’ sound ﬁnancial policy appears more likely to emerge from addressing obstinate political constraints.

128

Intermediation, governance and economic performance

The US savings and loan crisis To see how regulation can (and often does) promote something other than an eﬃcient outcome, let’s look at how US regulators should have addressed problematic savings and loan organizations (‘thrifts’) in the 1980s. Our investigation will beneﬁt from modeling deposit insurance as a ‘put option.’8 A put option gives its holder the right to sell an asset at a prespeciﬁed price (the ‘strike price’). Acting on this right (‘putting’ the asset to the contract’s seller) constitutes an ‘exercise’ of the option. Deposit insurance oﬀers ﬁnancial institutions an option to transfer liabilities to the insurance contract’s underwriter (for example, the Federal Deposit Insurance Corporation). This option reaches its strike price when the value of a ﬁnancial institution’s assets decreases to that of its liabilities. Past this ‘price,’ the ﬁnancial institution cannot fulﬁll its depositors’ claims, and thus puts those claims to the insurance agency. Figure 14.1 illustrates this feature of the deposit insurance contract, as well as those that we introduce below. When the deposit insurance contract is ‘in the money’ (that is, after the price of the underlying asset reaches its strike), the option clearly creates value for depositors. This value emerges from the insurance contract Value of put option to low-risk thrift

Value of insurance Value of put option to high-risk thrift Increased premium under capital-based pricing

Value of put option to depositors

45º Increased net worth requirement under fixed price contracts

Figure 14.1

Put option analysis of public insurance crises

Net worth

Financial intermediaries and their governance

129

making depositors whole for any shortfall in the ﬁnancial institution’s assets. The 45-degree curve in Figure 14.1 represents this value – the insurance contract pays oﬀ, dollar for dollar, for each dollar of insolvency that the intermediary realizes. The contract also creates value for the ﬁnancial institution, whether or not it is in the money. This value emerges from letting intermediaries enjoy the upside potential of their investments while mitigating the downside risk. Absent such protection, depositors would demand an increased default risk premium, and thus increase an intermediary’s cost of funds. The intermediary’s option value is greatest when its net assets equal zero (that is, when the option is ‘at the money’). At this point in Figure 14.1, the intermediary puts all losses to the insurer, but retains the right to all gains. As net worth decreases below zero, the put option’s value also decreases. Over this range, the option’s value (to the intermediary) depends on how likely the intermediary is to ‘resurrect’ itself. As net worth becomes more negative, this likelihood diminishes. In a symmetric manner, the option’s value also diminishes as an intermediary’s net worth grows more positive, since the chance of putting depositors’ claims decreases. Finally, holding all else constant, the intermediary’s valuation of the put grows with the risk of its asset portfolio. For any level of insurance premium or net worth requirement, the ﬁnancial organization can increase the value of insurance by taking on more risk. Indeed, such a strategy lets organizations enjoy the increased potential for upside gain while transferring increased downside risks to the insurer. Figure 14.2 illustrates this phenomenon. We can now employ our model to see why regulatory responses to the 1980s’ savings and loan crisis diminished economic performance and, more importantly, why eﬃciency-enhancing regulation must constantly resist political forces that promote such crises. The proximate cause of the crisis appears to be a rapid increase in nominal interest rates. Because savings and loans tended to fund long-term investments with short-term deposits, this increase caused the value of loans from low interest rate periods to plummet and the cost of capital to increase. The net worth of savings and loans rapidly decreased as a consequence, pushing their put option contracts toward or into ‘the money.’ As Figure 14.2 illustrates, an optimal regulatory response would have raised deposit insurance premiums, increased capital requirements, and in any event, restricted the moral hazard that deposit insurance encourages.9 Instead, regulators lowered net worth requirements, maintained deposit insurance premiums while increasing coverage and weakened both accounting standards and monitoring eﬀorts. Moreover, rather than being attributable to remediable ignorance, this failure appears to have emerged from more

130

Figure 14.2

0 K

A put option’s value increases with the underlying asset’s risk

The high-risk intermediary is more likely to put the depositors’ claims than is its low-risk counterpart

Probability

Net worth

Distribution of probabilities for low-risk intermediary

Distribution of probabilities for high-risk intermediary

Financial intermediaries and their governance

131

obstinate political pressures. Indeed, regulators’ actions (and those of their overseers) tended to concentrate beneﬁts on politically inﬂuential constituents while diﬀusing associated costs. While such transfers, and even their prospect, diminish economic performance, their political attractiveness lets them resist even considerable public outrage. A looming pension crisis?10 To better appreciate this resistance, let’s apply our model to an emerging US crisis with deﬁned-beneﬁt pension programs. Deﬁned-beneﬁt plans compensate retirees through speciﬁed monthly beneﬁts, which tend to vary with salary and years of service. The Pension Beneﬁt Guaranty Corporation (PBGC) guarantees most of these plans, and funds itself with premiums from the plans’ private-sector ‘sponsors’ (that is, employers). When an employer experiences ﬁnancial distress, the PBGC may take control of the plan’s management and use the plan’s assets and its own funds to pay retirees a capped portion of promised beneﬁts. The same types of problems that plagued savings and loans, however, can also create diﬃculties for the public insurance of pensions. In particular, market ﬂuctuations appear to have interacted with rules that govern how employers participate in the deﬁned-beneﬁt system to weaken PBGC’s ability to fulﬁll its obligations. Decreasing interest rates and stock market valuations, coupled with the exposure of pension plan assets to market ﬂuctuations, coincided with a marked increase in the underfunding of deﬁned-beneﬁt plans. For privately sponsored pensions, the value of assets to fund retirement obligations thus began to decrease in 2000 while the value of promised beneﬁts began to increase. The total underfunding of private pension plans grew from less than $50 billion at the end of 2000 to over $400 billion in 2006. At the same time, PBGC’s capacity to insulate workers from employer defaults turned from a $10 billion surplus in 2000 into a deﬁcit that now totals more than $20 billion. The same types of regulatory reforms that could have mitigated the savings and loan crisis should also be a part of pension reforms. Recent legislation, however, loosely ties pension insurance premiums to relevant risks, weakly encourages sound levels of net assets and ignores restrictions on the risks that insured portfolios can take. Rather than reﬂecting any ignorance of economics, this crisis may instead reﬂect more obstinate political forces that work against economic eﬃciency (for example, those that emanate from the incentive to pass private costs onto others). Glass-Steagall’s eﬃcacy and implications for corporate governance reforms In addition to advancing the ‘public’s interest’ (for example, by checking systemic risks), centrally produced governance services can be captured to

132

Intermediation, governance and economic performance

advance narrow interests at the public’s expense. Given this realization, one might be curious about the extent to which market discipline, by itself, can encourage private agents to govern themselves in a socially optimal manner. Kroszner and Rajan (1997) address this issue by looking at the organization of historical ‘universal banks.’ Universal banks can enjoy scope economies from jointly producing ‘commercial’ and ‘investment’ services.11 In doing so, however, they confront an important conﬂict, which Kroszner and Rajan (1997) characterize as follows: If a bank had private bad news about a ﬁrm it had lent to, it could use its underwriting arm to certify and distribute securities on behalf of the ﬁrm to an unsuspecting public and have the ﬁrm use the proceeds to repay the outstanding bank loan.

Left unchecked, this conﬂict would curb a ﬁnancial system’s productivity by creating a lemons problem. The Glass-Steagall Act of 1933 attempted to diminish this conﬂict by prohibiting the joint production of commercial and investment banking services. In doing so, it facilitated mutually beneﬁcial trades of ﬁnancial capital by increasing the cost for intermediaries to strategically act on informational advantages. Indeed, by checking the lemons problem that might have otherwise plagued primary markets for underwritten securities, the Act encouraged loanable funds to migrate toward more productive applications. But like all beneﬁts, this one comes with a (perhaps undue) cost. For example, Glass-Steagall’s coarse partitioning of commercial and investment banking services not only mitigated information asymmetries, it also increased costs by discouraging organizational forms that generate scope economies (for example, universal banks). By imposing a complete (as opposed to marginal) separation in this regard, the public interest motivation for Glass-Steagall relies on forgone economies being small (relative to the beneﬁts from weakening the lemons problem). If these economies are considerable, however, then markets might encourage a less than complete separation of commercial and investment services. Here, universal banks would separate constituent operations until the (private) marginal beneﬁt from reducing information asymmetries equaled the marginal cost from forgoing scope economies. Kroszner and Rajan (1997) ﬁnd evidence that markets disciplined the governance choices of universal banks in this manner. They begin by citing evidence (including that from Kroszner and Rajan, 1994) that default rates on securities underwritten by pre-Glass-Steagall universal banks were less than those on securities underwritten by contemporary investment banks. This relationship exactly opposes what one would expect if private organizational structures did not address the commercial–investment bank

Financial intermediaries and their governance

133

conﬂict of interest. They then develop evidence that banks recognized signiﬁcantly lower prices on underwritten securities when they lacked internal controls on the commercial–investment conﬂict of interest (for example, placing underwriting activities in a separate aﬃliate and placing independent directors on the aﬃliate’s board). To the extent that ﬁrms fully internalize their organizational implications, this evidence supports the hypothesis that Glass-Steagall’s formal separation went too far.12 Market Discipline Versus Public Regulation Kroszner and Rajan’s (1997) question remains an important one for contemporary regulators – that is, do markets go far enough in disciplining private governance decisions? One might ask this question, for example, when evaluating the Sarbanes-Oxley Act of 2002 (SOX). Here, the salient conﬂict occurs between auditing and consulting divisions within accounting service ﬁrms. To the extent that jointly producing auditing and consulting services creates scope economies, regulations (like SOX) that coarsely partition production processes appear ineﬃcient (at least from a technological perspective). However, this partition also creates commitment beneﬁts since jointly producing these services can exacerbate the principal-agent problem between auditors and ﬁnancial statement readers. Whether regulatory intervention can advance the public interest thus turns on whether markets adequately discipline decentralized choices of auditintegrity. We’ll turn to this issue, and those of corporate governance more generally, in the following chapter.

NOTES 1. 2.

3. 4. 5.

We’ll look at demanders in Chapter 15. The capacity for collateral and net worth requirements to mitigate asymmetric information is central to arguments for strong creditor rights. Manove et al. (2001), however, oﬀer an interesting argument for why creditor protections can be ‘too strong.’ In short, they suggest that collateral and screening are substitutes from the creditor’s perspective – that is, both check the risk of debtors playing morally hazardous actions. Because some types of creditors may have a comparative advantage in evaluating entrepreneurial projects, however, these tools are not equivalent from ‘society’s’ perspective – that is, collateral represents a contingency transfer, whereas screening can more productively allocate resources by checking bad projects before they get started. Such restrictions may even be necessary for ﬁnancial market eﬃciency – see Chan et al. (1992). A maximizing-intermediary’s objective is its own proﬁt, not that of society more generally. Frank et al. (2004) develop interesting evidence of this phenomenon for ‘copy cat’ mutual funds. Here, requiring fund managers to disclose portfolio holdings can mitigate moral hazard (on behalf of fund managers), but can also exacerbate information asymmetries by strengthening the ability to free ride on others’ eﬀorts.

134 6.

7. 8. 9. 10. 11. 12.

Intermediation, governance and economic performance Note well the source of this question’s importance. If capital markets cannot distinguish ‘good’ ﬁrms from ‘lemons,’ the market for good ﬁrms will unravel (just as that for h-type autos in Chapter 13). In this case, the economy will forgo truly productive projects – that is, individuals will realize an inferior level of economic well-being. Here, we are beginning to see the importance of ﬁnancial development for social welfare. We’ll address this topic more fully in Chapter 16. See Lummer and McConnell (1989: 104–5 and 108–9) for examples of credit agreement revisions. Merton (1977) is frequently cited as seminal in showing that deposit insurance can be modeled as a put option. In a model like that of Chan et al. (1992), portfolio restrictions appear necessary for sound deposit insurance. This sub-section heavily draws from the Economic Report of the President (Bush, 2006: 74–7). Commercial banking refers to an intermediary’s production of lending services, while investment banking refers to an intermediary’s production of underwriting services. Kroszner and Rajan (1997) acknowledge the potential for spillover eﬀects to rationalize regulation as enhancing eﬃciency, but their research design limits their investigation to own-ﬁrm eﬀects.

15.

Corporate governance

INTRODUCTION Asymmetric information constrains the set of mutually beneﬁcial trades that an economy can support. Consider, for example, the ﬁrm’s proﬁtmaximization problem, and recall that our most simple model (see Part I) implicitly assumes that acquiring inputs is costless – that is, the act of transacting negligibly consumes resources. But when a ﬁrm’s internal resources cannot fund input choices, it must demand capital from ﬁnancial markets. This demand, in turn, must address constraints from asymmetric information. This asymmetry can emerge from those who might supply ﬁnancial resources having less information about a ﬁrm’s prospects than does the ﬁrm itself. In addition, potential suppliers likely have little information about how a ﬁrm would, after receiving a ﬁnancial market’s proceeds, actually employ those funds. Absent organizational features that check such asymmetries, suppliers will rationally forgo investing, even in technically valuable projects. In Chapter 14, we examined how intermediaries can economize on such costs and thus push economies toward the superior levels of performance that more simple treatments of the ﬁrm imply. We restricted our attention, however, to how proximate suppliers of ﬁnancial capital can mitigate information asymmetries. But the axiom of maximizing-behavior applies to all economic actors. Consequently, if suppliers of ﬁnancial capital want to mitigate information asymmetries, then so should demanders. They do, and recent corporate governance scandals (such as those associated with Adelphia, Enron, Tyco, WorldCom) highlight the important role that acting (or not acting) on this incentive plays in an economy’s performance. Absent productive actions from ﬁrms (that is, demanders of ﬁnancial capital) on this dimension, ﬁnanciers will confront a less than optimal level of information about both a ﬁrm’s investment prospects and the manner in which a capital market’s proceeds will ultimately be employed. An economy in which ﬁrms do not economize on information costs must thus forgo proﬁtable projects. Our objective for the present chapter is to understand how organizational features that inﬂuence a ﬁrm’s investment and distribution decisions address this problem, and the extent to which markets can be expected to discipline ﬁrms in this regard. 135

136

Intermediation, governance and economic performance

To facilitate reference, we’ll refer to this set of features as an economy’s ‘corporate governance’ system. Our examination will let us see that, just as suppliers of loanable funds can mitigate information asymmetries by strategically organizing themselves, and thus push an economy away from inferior equilibria, demanders can structure their governance systems so as to facilitate the same type of welfare gains. We’ll learn, for example, that proﬁt-maximizing ﬁrms can have an incentive to mitigate the adverse selection problem by increasing their ﬁnancial transparency (and thus revealing their true ‘types’). Here, accounting rules, and those rules’ enforcers (for example, ﬁnancial statement auditors), strongly inﬂuence a society’s economic opportunities. We’ll also see that proﬁt-maximizing ﬁrms have an incentive to constrain managers’ capacity to play morally hazardous actions, and pay particular attention to how a ﬁrm’s capital structure (mix of debt and equity ﬁnancing) can produce this governance service. This approach will let us consider other salient features as well, including the manner in which ownership- and directorship-structures inﬂuence an economy’s capacity to produce governance services. Finally, it will further our consideration of how market forces relate to economic performance, and the conditions that are necessary for centrally produced governance services (for example, those that stem from public regulation) to facilitate superior outcomes.

ACCOUNTING SYSTEMS AND ASYMMETRIC INFORMATION Study after study has shown that better accounting standards help make ﬁrms more transparent, making it easier for them to inspire conﬁdence in investors. Of course, ﬁrms can adopt a policy of better disclosure by choosing a transparent accounting standard on their own. . . . But without sanctions for improper disclosure or omissions, this would just be cheap talk. (Rajan and Zingales, 2003: 161)

Recall our discussion of how Akerlof (1970) motivated the ‘lemons’ problem – that is, to the extent that used-car sellers enjoy informational advantages, they will bring only ‘lemons’ to the market. Also recall, however, that our axiom of maximizing-behavior ultimately rationalizes the behavior of all economic agents, whether they’re selling used cars or ﬁnancial securities. Hence, just as we expect lemon-automobiles when information asymmetries remain unchecked, we expect lemon-ﬁnancial assets when demanders of ﬁnancial capital (that is, suppliers of ﬁnancial assets) enjoy informational advantages. Absent forces to the contrary,

Corporate governance

137

ﬁnancial markets should unravel, leaving both demanders and suppliers of loanable funds at an ‘inferior’ equilibrium (stable outcomes from which both prefer to move). Private Incentives to Truthfully Disclose Financial Performance Wealthy economies tend to maintain well-developed and stable markets for ﬁnancial assets. One reason for this correlation is that demanders of loanable funds have an incentive to credibly commit to ﬁnancial transparency. Absent this transparency, ﬁnanciers can require such a high price for loanable funds that even productive demanders rationally forgo technically sound projects. The very incentive to maximize proﬁts can thus encourage ﬁrms to adopt reporting rules that increase transparency. But also note that maximizers have an incentive to enhance conﬁdence in reported results while strategically obscuring information about fundamental performance. To the extent that potential suppliers of loanable funds recognize this incentive, they will continue to charge demanders a relatively high price for ﬁnancial capital. Consequently, proﬁt-maximizing ﬁrms not only have an incentive to adopt accounting rules that facilitate transparency, they have an incentive to credibly bind themselves to following such rules – for example, by employing ‘independent’ auditors to certify ﬁnancial disclosures. To What Extent can Regulation Improve upon Market Discipline? Understanding how, and the extent to which, market incentives address this governance issue helps us evaluate the potential for ﬁnancial regulations to improve economic performance. Consider, for example, popular arguments that motivated the Sarbanes-Oxley Act of 2002 (SOX). Here, coincident increases in (a) ﬁnancial restatements and (b) accountants consulting for audit clients bolstered popular calls to de-couple the production of audit and non-audit services. This (reduced form) evidence suggested to many that regulation is necessary to make auditors better agents of ﬁnancial stakeholders and less susceptible to managerial pressure. These arguments ignore, however, the private incentives for ﬁrms to economize on information costs. Scholars such as Holmstrom and Kaplan (2003) suggest that such incentives are considerable. They even hypothesize that expanding regulations on how private parties govern themselves (measures that SOX embodies) risks an over-reaction to ‘extreme events.’ To see Holmstrom and Kaplan’s (2003) concern, notice that if the United States was producing a less than optimal level of governance services, or ineﬃciently producing a given level of such services, then its securities

138

Intermediation, governance and economic performance

market performance should have lagged that of foreign substitutes. But US markets have outperformed their foreign competitors over several diﬀerent time periods.1 The relatively high rate at which US aggregate output has grown corroborates this evidence. Any diﬃculties that the US corporate governance system is experiencing may thus weigh less heavily on economic performance than do related diﬃculties in other economies. Others oﬀer a more favorable evaluation of recent regulatory reforms. Kane (2003) even calls for a stronger response, arguing that the more frequent use of ‘high-powered’ compensation schemes (for example, stock options) strengthens managers’ incentive to strategically act on asymmetric information. Moreover, Kane observes, substitute governance services have not kept pace with this increased risk. Rather than maintaining the status quo, or even reigning in ‘excessive’ regulations, Kane thus argues for stronger regulations, especially those that address the ‘self-interested ethical codes of the auditing industry and accounting profession’ (Kane, 2003: 24). Both received and developing empirical research can help us evaluate these theoretical arguments’ merits. Consider, for example, the GlassSteagall Act, a Depression-era banking regulation that precluded intermediaries from simultaneously producing investment and commercial banking services. And recall that the concern for public welfare in this case stemmed from universal banks’ potential to opportunistically act on informational advantages – that is, having investment arms exploit commercial information to adversely select ﬁnancial assets into primary capital markets (see Chapter 14). Glass-Steagall hoped to improve welfare by separating these services’ production. Contemporary interventions, such as SOX, hold out an essentially identical promise. Policy proscriptions like these tend to ignore, however, the potential for factors other than auditor independence to produce corporate governance services, and thus ﬁnancial statement integrity. For example, director independence, audit committee experience and executive compensation structures create only a few of the many forces that researchers cite as inﬂuencing the credibility of reported ﬁnancial performance. Moreover, building on Kroszner and Rajan’s (1997) research, Falaschetti and Orlando (2004a, 2004b) and Brown et al. (2007) ﬁnd evidence that the extent of substitution in governance factors is considerable, and that proscriptive regulations like SOX may have sacriﬁced scope economies without improving upon market discipline. This evidence supports the hypothesis that auditors’ dependence on non-audit fees compromises earnings quality, but market discipline (perhaps enhanced by disclosure mandates) left little room for the SOX proscription on non-audit services to expand ﬁnancial market opportunities.

Corporate governance

139

CAPITAL STRUCTURE AND GOVERNANCE SERVICES Cash-constrained ﬁrms ﬁnance their operations by selling debt and equity securities to outsiders. We’ll refer to the particular mix of securities employed in this manner as a ﬁrm’s ‘capital structure.’ Simultaneous to addressing cash constraints, this structure inﬂuences the types of obligations (both formal and informal) that managers owe ﬁnanciers, and thus produces corporate governance services as well. Equity Finance and the Production of Governance Services Absent mitigating institutional structures, conventional equity securities appear to be inferior factors in producing governance services.2 To be sure, note that equity investments are sunk in the sense that, once a ﬁnancier commits equity capital to a ﬁrm, managers confront little in the way of formal obligations to share that investment’s product (for example, managers do not face an explicit obligation to pay dividends). Moreover, informal obligations may not fully substitute for this shortfall since, once a ﬁrm receives ﬁnancing, the cash constraint that originally motivated its capital market demands may be less binding. The prospect of having to repeatedly interact with equity market suppliers may thus only weakly encourage the development of reputations for sharing an equity investment’s product. Faced with the prospect of enjoying relatively little in the way of formal control rights, equity ﬁnanciers may be willing to supply capital only if they are relatively concentrated. When the centralized production of governance services discourages eﬃciency, owners have an incentive to independently produce those services. Collectively acting on this incentive may not be easy, however. For example, relative to diﬀuse owners, concentrated owners are more likely to be pivotal in producing the general beneﬁts that come from governance services (for example, monitoring). Indeed, since these services tend to beneﬁt all owners, concentrating ownership can discourage the free riding to which individually inconsequential investors might otherwise be attracted. But, while concentrating ownership can strengthen control rights (and thus mitigate agency costs that discourage ﬁnancial transactions), it can also constrain access to equity ﬁnancing by exposing minority shareholders to the prospect of expropriation (that is, a ‘tyranny of the majority’). Similarly, large owners may be unable to commit against expropriating the product of other stakeholders’ ﬁrm-speciﬁc eﬀorts (for example, see Shivdasani, 1993 and Falaschetti, 2002a). As a substitute for concentrating ownership, ﬁnanciers may thus employ a board of directors to act as a

140

Intermediation, governance and economic performance

‘monitoring intermediary.’ Boards, in turn, might address managerial agencies while protecting managers from the moral hazard of owners. In any event, ﬁrms can beneﬁt from addressing managerial agencies via incentive compensation schemes. But while these schemes can elicit productive activity, they can also encourage managers to strategically maximize the performance measure, rather than pursue an investor’s objective per se. Managers can, for example, strategically manipulate the accounting data from which their compensation derives, control the release of information in a manner that favors stock option payments, or even ‘capture’ the very process through which the compensation mechanism develops (Vives, 2000: 5). In this last case, the ‘regulator’ (directors) acts in the interest of the ‘regulated’ (managers) rather than that of its principal (equity investors).3 Debt Finance and the Production of Governance Services By far the dominant form of lending around the world is bank lending. (Shleifer and Vishny, 1997: 763)

Confronted with these diﬃculties, investors have tended to supply loanable funds via debt securities. Debt constitutes an overwhelming share of even large ﬁrms’ external funds (Mayer, 1990; Houston and James, 2001). This large share, in turn, can be rationalized by looking at debt’s comparative governance beneﬁts. Free cash ﬂows The incentive for managerial agents to play morally hazardous actions always exists – after all, they’re maximizers. To act on this incentive, however, agents must have access to requisite resources. One source of such resources is ‘free cash ﬂows.’4 Debt contracts strongly encourage the disbursement of otherwise free cash ﬂows, and thereby check even loosely monitored managers’ capacity to engage in morally hazardous behavior (Jensen, 1986). Indeed, absent funding for such actions, managerial agents can more credibly commit to ﬁnanciers that loanable funds will ﬁnd productive employment. Maintaining a debt-heavy capital structure can, in this manner, create a comparative governance advantage. Even if a ﬁrm completely distributes its free cash ﬂows, however, it can never fully inform capital market participants about its operations. In other words, a residual level of asymmetric information always persists. And if this residual is considerable, then it will preclude the ﬁrm from attracting capital with which to fund projects (even those that are technically proﬁtable). Had

Corporate governance

141

the ﬁrm maintained its free cash ﬂows, it would have increased its asymmetric information problem (ceteris paribus). At the same time, however, it would have also internally maintained resources for funding projects that capital markets may have otherwise dismissed. In this manner, free cash ﬂows oﬀer a ﬁnancial beneﬁt and governance cost. The cost emerges from the potential for free cash ﬂows to fund morally hazardous actions. The beneﬁt comes from the fact that asymmetric information can never be fully eliminated – ﬁrms always face a ‘cash constraint.’ And to the extent that they ﬁnd this constraint binding, free cash ﬂows will let them adopt (potentially proﬁtable) projects that do not ﬁnd support from capital market participants.5 Monitoring services Holders of privately placed debt appear to maintain a comparative advantage in producing monitoring services. Recall, for example, that capital suppliers face an important free-rider problem when attempting to produce evaluation and monitoring services. By (largely) internalizing the beneﬁts that their governance services produce, however, banks and other private lenders face a relatively small problem in this regard (see James, 1987 and Lummer and McConnell, 1989). In addition, debt contracts tend to be written over relatively short terms, and thus allow for repeated interaction with demanders of loanable funds. Finally, violations of debt contracts are rather transparent, and may thus facilitate remedies at a relatively low cost.6 Compared to public equity, debt appears to be a ‘governance-rich’ form of capital on these dimensions. Firm-speciﬁc investment Finally, debt may enjoy a comparative governance advantage in encouraging non-ﬁnancial stakeholders (for example, employees, suppliers) to make ‘speciﬁc investments.’ Speciﬁc investments confer disproportionate beneﬁts onto the ﬁrm in which they are made – that is, they are diﬃcult to externally market. And once stakeholders sink resources into such investments, residual claimants (for example, ﬁnanciers) are tempted to expropriate associated returns. Here, again, equity ﬁnance can exacerbate governance problems – this time by increasing productive investments’ exposure to the prospect of opportunistic expropriation. To see this temptation more clearly, consider the case of equity shareholders accepting a hostile takeover bid. Contributors to academic and popular media frequently interpret associated increases in target share prices as evidence that takeovers expand ﬁrms’ production possibilities. Shleifer and Summers (1989) recognize, however, that the tendency for target share prices to increase with hostile bids may also reﬂect the capacity for takeovers

142

Intermediation, governance and economic performance

to redistribute (rather than increase) a ﬁrm’s product.7 The post-takeover ﬁrm may, for example, enjoy an increased ability to renegotiate wage contracts and thus redistribute residual earnings from employees to owners. This feature of equity ﬁnancing can retard eﬃciency since, confronted with the prospect of opportunistic redistributions, employees, suppliers and other non-equity stakeholders have relatively little incentive to make otherwise optimal ﬁrm-speciﬁc investments.8 Debt capital, on the other hand, shelters stakeholders from this exposure (see Burkart et al., 1997). By limiting themselves to ﬁxed payments from residual earnings, debt-holders pose relatively little threat to expropriating proceeds that are necessary for motivating speciﬁc investments. Illustration – Tax Policy and Capital Structure Economists tend to agree that the United States’ recent reduction of dividends taxes maintains considerable expansionary potential. By treating dividends as taxable distributions and interest payments as deductible expenses, the pre-cut code biased capital structures toward debt. This bias arguably reduced the economy’s productive capacity by distorting investment decisions and increasing ﬁnancial distress costs. Federal Reserve Board chair Alan Greenspan thus ‘spoke warmly’ in congressional testimony ‘about the beneﬁts of eliminating the double taxation of dividends’ (The Economist, 2003: 31). But these beneﬁts may be oﬀset (at least partially) by an increase in associated governance costs. Consider, for example, the relatively rich set of information that emerges from debt-holders’ relationship with ﬁnancial capital demanders. As we learned above, holders of privately placed debt (for example, banks) can face a lesser free-rider problem than do those of publicly placed equity when attempting to produce evaluation and monitoring services. By removing a bias toward debt-heavy capital structures, the dividends tax cut can thus exacerbate information asymmetries between suppliers and demanders of loanable funds, thereby foreclosing at the margin otherwise mutually beneﬁcial capital market trades. Information asymmetries may grow even further if, in the face of a tax cut, dividend distributions do not fully oﬀset associated interest payment reductions. Recall, for example, that debt contracts create a relatively strong incentive for managers to disburse ‘free cash ﬂows,’ and can thus check even loosely monitored managers’ capacity to employ residual earnings in a manner that opposes shareholders’ interests. Finally, cutting the dividends tax can magnify the potential for owners to play morally hazardous actions. By expanding the capacity for residual claimants to expropriate returns from ‘ﬁrm-speciﬁc’ investments, removing

Corporate governance

143

a policy-bias for debt ﬁnance can discourage non-equity stakeholders from optimally employing their eﬀorts. Relative to equity claim holders, debtholders pose a small threat since their contracts essentially cap the residual earnings that can feasibly be extracted from ﬁrms. Debt-holders’ potential to opportunistically expropriate the product of ﬁrm-speciﬁc investments is thus likely to be smaller than that of associated equity holders. By removing the bias for debt obligations, and thus increasing (non-owner) stakeholders’ exposure to opportunistic expropriation, cutting the dividends tax might again oﬀset gains to productive economic activity. Debates over how the dividends tax cut might aﬀect such activity tend to focus on the technical merits of removing capital structure distortions. In doing so, however, they ignore how these same distortions can inﬂuence the economy’s stock of governance services. Ultimately, this ignorance may prove innocuous. Indeed, beneﬁts from reduced distortions and ﬁnancial distress may very well overwhelm any associated governance costs. If they do not, however, then the cut’s welfare consequences may rest on how strongly private governance production responds. Here, we recognize the incentive for ﬁrms to substitute for governance services that may have been produced via biased capital structures (for example, move to private equity). An open question, however, is whether this incentive has the right magnitude. This question is important since cutting the dividends tax can work against capital structures producing governance services, and ﬁrms may not fully or quickly respond to this negative force. First, if privately produced governance contributes to broader market integrity, then policies that bias capital structures toward producing governance services can sustain superior levels of social welfare. Indeed, absent such a bias, ﬁrms face a reduced (if any) incentive to produce governance services at levels that exceed those that are privately optimal. Second, to the extent that ﬁrms substitute for forgone features of governance-rich capital structures, they are likely to do so slowly. An unbiased capital structure represents a new technology with which to produce governance services, and ﬁrms will recognize adjustment costs while learning to employ this alternative mechanism. Governance costs associated with the tax cut may thus be especially important in the short run. In either case, productivity gains from the tax cut are likely to be less positive than those that are popularly reported.

NOTES 1. Here, ‘performance’ is measured by the rate of return on assets being traded on a particular exchange.

144

Intermediation, governance and economic performance

2. This discussion focuses on conventional equity securities, as opposed to those that venture capital and private equity channels might employ. 3. In various and inﬂuential writings, Lucian Bebchuk argues that increasing the boards’ accountability to shareholders can mitigate this type of agency problem. Recall from our discussion above, however, that such an increase can also threaten the product of other stakeholders’ eﬀorts. Absent a more careful evaluation, adopting prescriptions like those of Bebchuk may thus create ‘unintended,’ and adverse, consequences. 4. Jensen (1986) oﬀers a seminal examination of the governance implications of free cash ﬂows. Burkart and Ellingsen (2004) oﬀer a related rationalization for the prominence of ‘trade credit’ (that is, a loan bundled with an exchange of goods or services). Here, managers can more easily divert cash than they can physical inputs. 5. Carpenter et al. (1994), among others, develop evidence on how private investment relates to free cash ﬂows. 6. ‘Legal protection of creditors is often more eﬀective than that of the shareholders, since default is a reasonably straightforward violation of a debt contract that a court can verify’ (Shleifer and Vishny, 1997: 752). 7. Stout (2002) reviews and extends this type of argument. 8. Falaschetti (2002a) oﬀers evidence that golden parachutes strengthen commitments against such opportunistic behavior.

16.

Financial development and economic performance

INTRODUCTION I think the question, ‘Why isn’t the whole world rich?’ is the most important question facing economists. (Edward C. Prescott, quoted by Rolnick, 1996)

In Parts II–IV, we argued that monetary systems strengthen economic performance by facilitating spot transactions. We now understand that, in addition, ﬁnancial systems facilitate transactions across time. To the extent that such systems economize on transactions costs, they free up resources that the process of trading might have exhausted, and thus ease the ﬂow of resources to higher-valued uses. Rather than beneﬁting a narrow group of ﬁnancial market participants, ﬁnancial development can thus expand economic opportunities more generally. We conclude the present set of chapters with a more careful evaluation of this last implication – that is, ﬁnancial development strongly inﬂuences general economic opportunities. In doing so, we’ll also address this chapter’s opening quotation by arguing that, just as persistent political forces can work against eﬃciency in monetary systems, they can also constrain ﬁnancial systems from realizing their potential. Unless a society’s political organization weakens these inﬂuences, individuals must forgo the expanded opportunities that ﬁnancial development oﬀers. Variation in societies’ success on this political dimension says a lot about why the whole world is not rich.

CHANNELS FROM FINANCE TO ECONOMIC PERFORMANCE Several channels exist through which ﬁnancial development can inﬂuence economic performance. Recall from this part’s motivating example that ‘cash-constrained’ ﬁrms must forgo fundamentally proﬁtable projects (see Chapter 13). Economies that enjoy productive ﬁnancial systems, however, need not rest at the resulting inferior equilibrium. Indeed, ﬁnancial 145

146

Intermediation, governance and economic performance

development makes superior outcomes feasible by facilitating mutually beneﬁcial trades between ‘idea-rich’ entrepreneurs and ‘liquidity rich’ ﬁnanciers. Production factors encounter relatively little friction in such settings when moving from lower- to higher-valued uses. Financial development also creates opportunities for ﬁrms to employ relatively skilled managers. Indeed, the pool of managers from which diﬀusely held ﬁrms can draw is larger than that from which closely held ﬁrms draw (the latter must, by deﬁnition, restrict management to owners). Moreover, by allowing for the independent production of management and ownership services, separating ownership from control can expand opportunities by encouraging individuals to specialize. Financial development thus not only expands economic opportunities by relaxing constraints on a given set of projects (through the cash constraint channel), it also enlarges the set of projects that are technically feasible (through the managerial expertise channel). Finally, ﬁnancial development can strengthen competitive forces. Notice that, as ﬁnancial development wanes, information asymmetries discourage the separation of management from control. A reduction in separation, in turn, can insulate ineﬃcient incumbents from competition that might have otherwise drawn resources to more productive sectors or ﬁrms. In cases like this, potential entrants must overcome barriers of having to embody both technical competence and ﬁnancial capital. By lowering such barriers, ﬁnancial development can strengthen competitive forces that select for productive ﬁrms and thus free resources from inferior employment. We’ll refer to this third channel from ﬁnancial development to economic performance as the ‘competition channel.’1 Figure 16.1 summarizes how ﬁnancial development can promote economic performance through each of these channels. As our previous chapters show, ﬁnancial intermediation and corporate governance can increase ﬁnanciers’ conﬁdence that (a) managers will productively employ an investment’s proceeds and (b) distribute the consequent product in a manner that meets or exceeds the investors’ opportunity cost. Likewise, such systems can protect the product of managerial (and other stakeholders’) eﬀorts from being expropriated, and thus strengthen incentives to productively employ a ﬁnancier’s capital from the beginning. The consequent reduction in friction that capital experiences when attempting to move to highervalued uses thus relaxes cash constraints, facilitates specialization and eases market entry. Working through these channels, ﬁnancial development can expand economic opportunities by making feasible more productive uses of a given technology, expanding technological frontiers and selecting for proﬁt maximizers.

Financial development and economic performance

147

Output Production technology with specialized managers

Profit** Profit*

Production technology with manager-owners

Capacity to separate facilitates specialized management

Share of output to production factors

Profit

–

K

K*

K**

Capacity to separate ownership from control relaxes cash constraint

Input Capacity to separate increases competition, leaving firms that act ‘as if’ they are maximizers

Figure 16.1 Channels from ﬁnancial development to economic performance

EVIDENCE FOR FINANCIAL DEVELOPMENT AND ECONOMIC PERFORMANCE Consistent with the hypothesis that ﬁnancial development encourages superior outcomes, economies that enjoy productive ﬁnancial systems tend to perform more strongly than those where ﬁnancial development lags. Indeed, increased ﬁnancial development shares a strong association with dampened and less frequent economic ﬂuctuations, as well as relatively high rates of economic growth. That plausible channels exist for ﬁnancial development to exert this inﬂuence, and associated empirical support for this hypothesis addresses diﬃculties with reduced form evidence,2 increases conﬁdence that this association is more than a simple correlation. In an inﬂuential article, Rajan and Zingales (1998) report that industries that heavily rely on external ﬁnance (that is, those that tend to face binding cash constraints) grow signiﬁcantly faster in countries that maintain welldeveloped ﬁnancial systems. A qualitatively identical relationship appears at higher levels of aggregation, where the rate of macroeconomic growth increases with ﬁnancial development.3 This type of evidence also emerges from cross-state investigations. The relaxation of multi-state branch banking restrictions since the mid-1970s,

148

Intermediation, governance and economic performance

for example, appears to have improved the quality of bank lending (as measured by a decline in non-performing loans). This improvement, in turn, appears to have increased entrepreneurial activity and thus pushed state-level economies onto higher and more stable growth paths (Jayaratne and Strahan, 1996). The breadth of this evidence is impressive. Looking at the (micro) ﬁrm and (macro) economy levels, as well as across countries and US states, researchers ﬁnd evidence that living standards and growth prospects strongly depend on ﬁnancial development. Moreover, achieving these longterm beneﬁts may not risk short-term stability. Indeed, ﬁnancial development may also help reduce economic volatility.4

WHY ISN’T THE WHOLE WORLD RICH? The models reviewed above are ‘good’ in that they build on a ﬁrm ground (that is, the axiom of maximizing-behavior) and ‘predict’ well (that is, they rationalize empirical regularities without trivializing problems of interest). They leave open, however, an important question: if ﬁnancial development really expands economic opportunities, then ‘why isn’t the whole world rich?’ Recall a related problem from Parts II and III – if price stability is so important for economic performance, then why do monetary authorities persistently deviate from price-stabilizing policies? We argued there, and in Part IV, that if the axiom of maximizing-behavior is a reasonable starting point for modeling human sociality, then it must also be reasonable for modeling monetary authorities.5 ‘Maximizing’ monetary authorities, however, constantly face incentives to fund public expenditures through inﬂation (see Part II) and opportunistically boost economic performance (see Part III). In this light, rather than being the rule, societies that organize political aﬀairs in a manner that mitigates such unproductive forces appear exceptional.6 For similar reasons, societies that productively organize their ﬁnancial systems also appear exceptional. In our present set of chapters, we have so far argued that well-developed systems mitigate information asymmetries that discourage capital from reaching productive employment. While such a migration can beneﬁt societies in general, however, it can also lessen the economic opportunities of politically (but not economically) powerful constituents. And since a society’s capacity to check ineﬃcient political forces depends on these constituents’ strength, inferior economic performance can persist (even if everyone understands such eﬀects!).

Financial development and economic performance

149

Saving Capitalism from the Capitalists Financial development appears to be so beneﬁcial that it seems strange that anyone would be opposed to it. However, ﬁnancial development is not always a win-win. It could pose a threat to some. (Rajan and Zingales, 2003: 166)

Rajan and Zingales (2003) show that even capitalists can beneﬁt from sabotaging ﬁnancial development. We’ve already learned that ‘idea-rich’ but ‘capital-poor’ innovators have a strong incentive to support ﬁnancial development. However, as those innovators grow to be incumbents, the dimensions on which they can compete expand. In particular, not only can continual innovation strengthen their market position, so can political barriers to new (and potentially more productive) entrants. Sabotaging the ﬁnancial sector can be an eﬀective strategy for pursuing this latter objective. Notice that ‘losers’ from ﬁnancial regulations can be rather anonymous – for example, try to identify innovators who failed from a lack of ﬁnancial access rather than good ideas. In addition, while deteriorating ﬁnancial systems can create widespread costs, those costs can disproportionately burden ﬁrms for which information asymmetries are greatest – potentially innovative entrants. Acting in their own self-interest, political agents may thus ﬁnd attractive policies that impede (rather than promote) ﬁnancial development. And since these forces emerge from individually rational actions, they can persistently discourage ﬁnancial systems from developing in a productive manner, or even reverse the development of successful systems. Social Democracy Market mechanisms have to compete for votes with alternative resource allocation schemes more favorable to the underdogs; and in this competition, fairness to me is my primary concern, eﬃciency is someone else’s problem. (McFadden, 2006: 6)

Roe (2003) oﬀers an additional rationalization of why, despite its potential to create widespread economic beneﬁts, ﬁnancial development frequently faces strong political resistance. For Roe, an important constraint on societies’ ﬁnancial opportunities is the extent to which governments can be characterized as ‘socially democratic.’ Here, ‘social democracy’ refers to institutions (formal and informal) that deﬁne property rights in a manner that favors labor over capital in distributional considerations – that is, ‘a nation in which employee pressures are strong’ (Roe, 2003: 24). By favoring manager–labor coalitions, social democracies can widen the agency gap between suppliers of ﬁnancial capital and their managerial agents. The

150

Intermediation, governance and economic performance

resulting increase in moral hazard, in turn, can discourage what might otherwise be mutually beneﬁcial ﬁnancial market trades. In particular, the relatively high ‘systemic’ level of moral hazard in social democracies can constrain an economy’s set of organizational opportunities – that is, the capacity to separate ownership from control. Roe (2003) ﬁnds empirical support for this conjecture from cross-country evidence that ownership concentration increases with social democracy. Here, again, we observe a channel through which persistent pressures for ‘accountability’ can, paradoxically, work against economic performance.

CONCLUSION Roe’s (2003) argument, like that of Rajan and Zingales (2003), does not require individual irrationality for its validity. Politically dominant coalitions can strictly prefer lesser-developed ﬁnancial systems. For Rajan and Zingales, this coalition includes incumbent capitalists who beneﬁt from discouraging innovative newcomers. Roe identiﬁes a related constituency – laborers whose economic opportunities shrink as ﬁnancial development fuels the process of ‘creative destruction.’ In both cases, educating individuals about the merits of ﬁnancial development may do little to encourage productive reforms (or discourage unproductive policies). Indeed, unless reforms address fundamental distributional issues (for example, by mitigating adjustment costs), even the most (technically) productive ﬁnancial developments must go unrealized, and those that already exist will face considerable political risks.

NOTES 1. Rajan and Zingales (2003) examine this channel in an authoritative book-length manuscript. Levine (1997) reviews a broader set of such channels. 2. Chapter 2 reviews these diﬃculties. 3. See Levine (1997) for an authoritative review of this and related research. 4. Jayaratne and Strahan (1996) and the Economic Report of the President (Bush, 2006: 203–4) oﬀer suggestive evidence to this eﬀect. 5. And recall from Part I that we must model. 6. Recall from Part II, for example, that ﬁat currency has only enjoyed success during the last 35 years or so.

Questions Whether in a classroom or other setting, our readers might enjoy a better understanding of money, ﬁnancial intermediation and governance by contemplating the following questions.

PART I 11. True, false, explain: We must, either explicitly or implicitly, employ ‘models’ to understand social phenomena. 12. True, false, explain: Even if individuals do not consciously ‘maximize,’ assuming that they do can facilitate insight to how economies behave. 13. Under what conditions is the assumption of maximizing-behavior reasonable? Do the markets in which monetary and ﬁnancial services trade satisfy this condition? 14. A ‘good’ model builds on self-evident truths to generate accurate predictions: A. True, false, explain: Models that logically derive from the axiom of maximizing-behavior can be ‘good.’ B. Please describe a model built on the axiom of maximizingbehavior that predicts well. C. Did the decision-maker in your example (human or non-human) explicitly engage in maximizing-behavior? D. Even if decision-makers do not explicitly engage in maximizingbehavior, why can we gain insight from models that assume maximizing-behavior? 15. Congratulations! Congress just approved your nomination to the Federal Reserve Board of Governors (that is, the United States’ monetary authority). As a Board member, you will be presented with numerous models of the economy. How will you evaluate these competing models’ merits? 16. Why does the axiom of maximizing-behavior imply that the marginal cost of one’s actions equals the marginal beneﬁt? 17. ‘Intellectual honesty’ has been characterized as requiring the precise speciﬁcation of ‘conditions under which one is willing to give up one’s

151

152

Questions

position’ (Lakatos, 1978: 8–9). How does being transparent about modeling encourage ‘intellectual honesty’? 18. How does a principles-level model of an economy (that is, one that ignores transactions costs) diminish money and banking topics? Why is it nevertheless important for us to investigate these phenomena? How can models build on the axiom of maximizing-behavior to advance this investigation?

PART II 11. Why is bartering costly? 12. What is money? 13. How can money economize on bartering costs and therefore improve economic performance? 14. True, false, explain: If bartering is costless, then maximizers have little incentive to develop monetary systems. 15. If equilibriums in monetary economies can be superior to those in barter economies, then why don’t all societies employ well-developed monetary systems? 16. What distinguishes ﬁat currency from other assets that can produce transactions services? (Hint: what is ﬁat currency’s non-monetary value?) 17. Why is a well-developed system of ﬁat money superior to that of commodity money? Why did societies take so long to adopt ﬁat money (that is, about 11 000 years!)? 18. How can agreeing to use a particular form of money (for example, the euro or US dollar) improve everyone’s economic well-being? What diﬃculties might we encounter in attempting to achieve this end?

PART III 11. Congratulations! You are an elected oﬃcial, and your prospect for reelection is sensitive to the economy’s performance (that is, the level, variance and growth of wealth). Suppose that the classical model is a ‘good’ one. Do you want the monetary authority to pursue an ‘easier’ policy (that is, increase the money supply)? Why or why not? 12. How do market forces mitigate deviations from ‘full employment’ (that is, the level of employment that clears the classical labor market) and what does their eﬀectiveness imply for activist policy opportunities? 13. What does the axiom of maximizing-behavior imply for how individuals form expectations? Given this implication, please evaluate the

Questions

14.

15.

16.

17.

18.

153

potential for ‘active’ monetary policy to improve economic performance. Suppose that only ‘classical’ forces inﬂuence an economy’s evolution: A. Given this assumption, carefully evaluate the potential for activist monetary policy to mitigate economic ﬂuctuations. B. How conﬁdent are you in this evaluation (that is, do the insights that you developed come from a good model)? Suppose that only Keynesian forces inﬂuence an economy’s evolution: A. How can involuntary unemployment arise in this model? Why might this strictly positive level of unemployment ﬂuctuate? B. How conﬁdent are you in this evaluation (that is, do the insights that you developed come from a good model)? Congratulations on your conﬁrmation to the Federal Reserve Board of Governors (that is, the United States’ monetary authority)! As a Governor, you are receiving pressure from elected politicians (for example, the President, Senators) to ‘ﬁne tune’ the economy’s performance: A. Suppose that you are a ‘classical Governor’ – that is, one who sees the classical model of the aggregate economy as a ‘good model.’ Given this assumption, please carefully argue to our elected politicians that you maintain little (if any) capacity to inﬂuence ﬂuctuations in real economic activity (for example, employment, output). B. Now, suppose that you are a ‘Keynesian Governor’ – that is, one who sees the Keynesian model of the aggregate economy as a ‘good model.’ Relative to your incarnation as a classical Governor, are you more or less conﬁdent about your ability to inﬂuence real economic activity? Why? C. If individuals form expectations rationally, and holding all else constant, would the welfare consequences of your policy prescriptions tend to be better as a classical or Keynesian Governor? Why? D. Finally, why would the same elected political oﬃcials who are attempting to inﬂuence your monetary policy decisions have delegated authority to you in the ﬁrst place? In other words, how might those oﬃcials have beneﬁted from abdicating authority over this policy domain? True, false, explain: If individuals form expectations rationally, then the capacity for monetary policy to inﬂuence real activity (for example, output) is relatively small. Congratulations on your continued service as a Federal Reserve Board (FRB) Governor! While preparing you for the next Federal Open Markets Committee (FOMC) meeting, a staﬀ member notes that the

154

19. 10.

11.

12. 13.

Questions

US economy is experiencing a negative supply shock (that is, the relative price of oil is increasing), but that economic performance appears relatively unaﬀected. This staﬀ member further argues that this robust performance reﬂects an oﬀset to the supply shock from an increased foreign demand for US goods and services (that is, an increase in US aggregate demand). Please carefully consider whether your staﬀ member’s argument is a good one as follows: A. Begin by considering the ‘pre-shock’ economy. In particular, please show how the labor market equilibrates. Your answer should reference how the choices of (a) proﬁt-maximizing ﬁrms and (b) utilitymaximizing households ultimately determine the economy’s equilibrium real wage (that is, (wp) E ) and employment level (LE). B. Sticking with the ‘pre-shock’ economy, please show how the goods market equilibrates. Your answer should reference how (a) equilibrium employment from part A (LE), (b) the manner in which maximizers form expectations, and (c) the level of aggregate demand (before the increased foreign demand) determine the economy’s equilibrium price level (PE) and level of output (YE). C. Now show how the supply shock and increased demand from abroad ( AD ↑) together aﬀect each of the variables that you developed in parts A and B (that is, (wp) E , LE, PE and YE). D. In light of how (wp) E , LE, PE and YE ultimately respond in part C to the supply shock and increased foreign demand, please assess the validity of your staﬀ member’s rationalization of robust economic performance. In light of Lucas’s critique, should policy makers evaluate structural or reduced form evidence? Why? Leaving considerations of how expectations are formed aside, please evaluate the potential for activist monetary policy to improve welfare. Your argument should draw on our discussion of how ﬂuctuations aﬀect welfare when we hold growth rates constant (that is, Lucas’s argument) and when we let ﬂuctuations aﬀect growth rates (that is, Barlevy’s argument). Please critically evaluate Edward Prescott’s assertion that ‘What we should be worrying about is increasing the average rate of increase in economywide productivity and not smoothing business cycle ﬂuctuations’ (quoted by Rolnick, 1996). What does it mean for a policy to be ‘time inconsistent’? Applied to the case of monetary policy, why is time inconsistency problematic? Recall from Part II that systems of ﬁat money are a recent development. How might the problem of time inconsistency have slowed societies’ move from commodity-based to ﬁat systems?

Questions

155

14. Why might elected political agents, despite their interest in monetary policy’s real consequences, delegate authority over monetary policy to an insulated bureaucracy? What incentives do political agents (for example, legislators, presidents) face to circumvent this delegation? What do these incentives imply for the optimal organization of monetary authorities?

PART IV 11. Congratulations! You have been appointed to a select committee with the task of improving economic data. One measure in which you are interested is the economy’s stock of monetary services. What types of assets should this measure include, and how should this measure aggregate those assets? 12. In which direction (that is, upward or downward) do simple sum measures bias measures of monetary services? Why? What diﬃculties might this bias create for policy makers? How can a weighted aggregate address this bias? 13. Why do monetary authorities face a time inconsistency problem? How can the organization of these authorities mitigate this problem? Why don’t all governments act on this organizational prescription? 14. Why might elected political agents, despite their interest in monetary policy’s real consequences, delegate authority over monetary policy to an insulated bureaucracy? What incentives do political agents (for example, legislators, presidents) face to circumvent this delegation? How does the Federal Reserve System’s design mitigate the potential for such circumvention? 15. Conventional regulatory agencies are subject to political control via several mechanisms. For instance, such agencies may depend on legislatures for budgets or reappointments. Please describe how controls over the Federal Reserve System are diﬀerent. In particular, how does the Fed ﬁnance its operations? How does the term structure for Federal Reserve Board Governors diﬀer from more conventional structures? What incentives does this institutional framework provide for individuals who directly inﬂuence monetary policy? 16. True, false, explain: The Federal Reserve System has too much autonomy from the rest of the government. 17. Nobel Prize-winner Edward Prescott once said that, ‘An independent Fed, I think, is something that is a valuable commitment technology for Congress, for the same reason an independent judiciary is a good arrangement’ (see Rolnick, 1996). Please refer to the ‘time inconsistency

156

Questions

problem’ to carefully identify the source of value for this ‘commitment technology.’ 18. Suppose that society prefers higher output to lower output and lower inﬂation to higher inﬂation, and note that the Federal Reserve System insulates monetary authorities from these preferences. Can such an ‘undemocratic’ organization nevertheless improve social welfare? Why? 19. Nobel Prize-winner James Tobin argues that there should not be ‘votes on the Federal Open Market Committee [FOMC] for people who are not appointed as public servants by the president and who are not subject to conﬁrmation by the Senate. I think either the bank presidents should have no votes, or, to achieve voting status, they should be appointed and conﬁrmed in the same manner as the governors . . . I just think it’s contrary to democratic politics to have private citizens voting on the most important questions of macroeconomic policy’ (quoted from Fettig, 1996): A. How would monetary policy change if the US adopted Tobin’s recommendation to make the FOMC more ‘accountable’? B. Does maximizing ‘democracy’ appear to be a dominant strategy? 10. Tobin also recommends that ‘we go back to the practice, as before 1933, of having the Secretary of the Treasury on the [Federal Reserve] Board’ (see Fettig, 1996). How would monetary policy change if the US adopted this recommendation?

PART V 11. Two types of asymmetric information exist: adverse selection and moral hazard. Please deﬁne these asymmetries and explain how they restrict economic performance (that is, levels, ﬂuctuations, or growth of consumption possibilities). 12. True, false, explain: If information is symmetric, then maximizers have little incentive to develop systems of intermediation (ﬁnancial or otherwise). 13. Why might a used car dealer be more truthful about an automobile’s quality than would a direct seller? 14. How can collateral or net worth let borrowers ‘signal’ ﬁnancial markets about the quality of their prospective projects? Put more formally, how might the availability of (asymmetrically) costly actions induce a separating equilibrium and thus help channel loanable funds to more productive uses? 15. Why do demanders of loanable funds tend to acquire those funds primarily through intermediaries in general and banks in particular?

Questions

157

16. True, false, explain: Holding all else constant, economic performance strengthens with intermediaries’ capacity to economize on transaction costs. 17. Is market discipline suﬃcient for optimally organizing a ﬁnancial market? If not, how can public regulation improve upon decentralized outcomes? What obstacles must public regulation overcome to achieve such improvements? How likely is public regulation to overcome these obstacles, and what does this likelihood say about the eﬃciency of regulations that we observe? 18. Suppose that information technology can reduce information asymmetries. What do advances in this technology imply for (a) the importance of traditional intermediaries (for example, banks), (b) the cost of protecting intellectual property in ﬁnancial information via traditional organizational forms, and (c) systemic risks, and thus the scope for productive ﬁnancial regulation? 19. Systems of ‘micro ﬁnance’ have emerged in a number of developing economies. Because these systems’ lending-terms frequently include high interest rates or severe covenants (for example, strong collateral requirements), they risk being characterized as usurious (that is, extortive). If policy makers respond by sanctioning associated intermediaries, will (holding all else constant) productive capacity increase or decrease in these economies? Why? 10. Why might economic performance be sensitive to the potential for bank panics? What does your answer say about the potential for ﬁnancial market regulations to enhance welfare? 11. In a model that characterizes insurance as a put option, why should net worth requirements increase with the risk of the underlying asset? Why shouldinsurancepremiumsincreasewiththeriskof theunderlyingasset? Why are portfolio restrictions necessary for a sound insurance system, even if capital requirements and premium prices are risk sensitive? 12. Public insurance for US deﬁned beneﬁt pensions is experiencing considerable diﬃculties, and these diﬃculties parallel those from the 1980s’ saving and loan crisis. True, false, explain: If policy makers understood economics, then regulatory crises like these would not repeat themselves. 13. You are a corporate executive and demand ﬁnancial capital to fund a new project. If interest on bond obligations can be deducted for tax purposes, why would decreasing dividend tax rates reduce economic distortions and thus enhance productivity? What ‘unintended consequences’ might emerge from such a policy? 14. True, false, explain: Policies that increase auditor independence enhance ﬁnancial statement transparency, and thus ultimately bolster economic performance.

158

Questions

15. An economy’s system of corporate governance strongly inﬂuences its productive capacity. For example, an economy’s production of ﬁnancial audit services can help check the problem of adverse selection (that is, the ‘lemons’ problem): A. Confronted with a spike in corporate governance scandals, US policy makers implemented the Sarbanes-Oxley Act of 2002, part of which separates the production of audit and consulting services. Why might this separation be superﬂuous – that is, why might private ﬁnancial service ﬁrms have already organized themselves (without regulation) in a manner that checked the adverse selection problem? B. Why, even if private ﬁrms maintain an incentive to check the adverse selection problem, might you still want to encourage regulations that separate the production of audit and consulting services? 16. Why is ﬁnancial development important for economic performance? Why, nevertheless, do so many economies remain undeveloped, and why do even developed economies face a persistent threat of ‘reversal’? 17. Congratulations! You have just been named President of the World Bank. As President, your objective is to raise developing economies’ aggregate output (that is, the variable Y in our models from Part III). You decide to work toward this objective by reforming markets for ﬁnancial capital. To implement your plan, however, you must gain the support of the World Bank’s backers. Please carefully argue to these backers that an economy’s capacity to generate wealth rests, in large part, on its capacity to produce intermediation and corporate governance services, and thus mitigate information asymmetries between suppliers and demanders of loanable funds. 18. Having addressed an important technical issue in Question 17, you must now confront what can be even more fundamental political issues: A. While implementing your reform-proposals would expand economic opportunities, doing so might also diminish the welfare of certain constituents. Please explain why these constituents (for example, incumbent ﬁrms) might oppose performance-enhancing reforms. B. Does this opposition imply that individuals are not always ‘maximizers’? Please explain. C. How might you craft policies that are not only economically sound, but also address relevant political constraints?

Bibliography Admati, Anat and Paul Pﬂeiderer (2000), ‘Forcing ﬁrms to talk: ﬁnancial disclosure regulation and externalities,’ Review of Financial Studies, 13 (3), 479–519. Akerlof, George (1970), ‘The market for lemons: quality uncertainty and the market mechanism,’ Quarterly Journal of Economics, 84 (3), 488–500. Akerlof, George and Janet Yellen (1985), ‘A near-rational model of the business cycle, with wage and price inertia,’ Quarterly Journal of Economics, 100 (supplement), 823–38. Akerlof, George and Janet L. Yellen (1986), Eﬃciency Wage Models of the Labor Market, Cambridge, New York and Melbourne: Cambridge University Press. Alchian, Armen (1950), ‘Uncertainty, evolution, and economic theory,’ Journal of Political Economy, 58 (3), 211–21. Alesina, Alberto (1987), ‘Macroeconomic policy in a two-party system as a repeated game,’ Quarterly Journal of Economics, 102 (3), 651–78. Alesina, Alberto and Lawrence Summers (1993), ‘Central bank independence and macroeconomic performance: some comparative evidence,’ Journal of Money, Credit, and Banking, 25 (2), 151–62. Alesina, Alberto and Guido Tabellini (2004), ‘Bureaucrats or politicians?,’ NBER working paper no. 10241. Alesina, Alberto, Nouriel Roubini and Gerald Cohen (1997), Political Cycles and the Macroeconomy, Cambridge, MA: MIT Press. Backus, David and John Driﬃll (1985), ‘Inﬂation and reputation,’ American Economic Review, 75 (3), 530–38. Bajari, Patrick and Steven Tadelis (2001), ‘Incentives versus transactions costs: a theory of procurement contracts,’ RAND Journal of Economics, 32 (3), 387–407. Barlevy, Gadi (2004), ‘The cost of business cycles under endogenous growth,’ American Economic Review, 94 (4), 964–90. Barnett, William (1987), ‘The microeconomic theory of monetary aggregation,’ in William Barnett and Kenneth Singleton (eds), New Approaches to Monetary Economics, Cambridge, New York and Melbourne: Cambridge University Press, pp. 115–68.

159

160

Bibliography

Barnett, William and Kenneth Singleton (eds) (1987), New Approaches to Monetary Economics, Cambridge, New York and Melbourne: Cambridge University Press. Barro, Robert and David Gordon (1983a), ‘A positive theory of monetary policy in a natural rate model,’ Journal of Political Economy, 91(4), 589–610. Barro, Robert and David Gordon (1983b), ‘Rules, discretion, and reputation in a model of monetary policy,’ Journal of Monetary Economics, 12 (1), 101–21. Bawn, Kathleen (1995), ‘Political control versus expertise: congressional choices about administrative procedures,’ American Political Science Review, 89 (1), 62–73. Bawn, Kathleen (1997), ‘Choosing strategies to control the bureaucracy: statutory constraints, oversight, and the committee system,’ Journal of Law, Economics, and Organization, 13 (1), 101–26. Baye, Michael (2000), Managerial Economics and Business Strategy, 3rd edn, Boston, MA: Irwin McGraw-Hill. Bebchuck, Lucian and Mark Roe (1999), ‘A theory of path dependence in corporate governance and ownership,’ Columbia Law School, Center for Law and Economics Studies working paper no. 131. Becker, Gary (1962), ‘Irrational behavior and economic theory,’ Journal of Political Economy, 70 (1), 1–13. Becker, Gary (1976a), ‘Altruism, egoism and genetic ﬁtness: economics and sociobiology,’ Journal of Economic Literature, 14 (3), 817–26. Becker, Gary (1976b), The Economic Approach to Human Behavior, Chicago, IL: University of Chicago Press. Beiner, Stefan, Wolfgang Drobetz, Markus Schmid and Heinz Zimmermann (2004), ‘An integrated framework of corporate governance and ﬁrm valuation – evidence from Switzerland,’ WWZ Department of Finance working paper no. 10/03. Benassy, Jean-Pascal (1990), ‘Non-Walrasian equilibria, money, and macroeconomics,’ in Benjamin Friedman and Frank Hahn (eds), Handbook of Monetary Economics, New York: North-Holland, pp. 103–69. Berman, Eli (2003), ‘Hamas, Taliban, and the Jewish underground: an economist’s view of radical religious militias,’ NBER working paper no. 10004. Bernanke, Ben (1983), ‘Non-monetary eﬀects of the ﬁnancial crisis in the propagation of the Great Depression,’ American Economic Review, 73 (3), 257–76. Bernanke, Ben (1995), ‘The macroeconomics of the Great Depression: a comparative approach,’ Journal of Money, Credit, and Banking, 27 (1), 1–28.

Bibliography

161

Bernanke, Ben and Kevin Carey (1996), ‘Nominal wage stickiness and aggregate supply in the Great Depression,’ Quarterly Journal of Economics, 111 (3), 853–83. Bernanke, Ben and Mark Gertler (1989), ‘Agency costs, net worth, and business ﬂuctuations,’ American Economic Review, 79 (1), 14–31. Besanko, David, David Dranove and Mark Shanley (2000), Economics of Strategy, 2nd edn, New York: John Wiley & Sons, Inc. Bharath, Sreedhar (2002), ‘Agency costs, bank specialness and renegotiation,’ http://papers.ssrn.com/so13/papers.cfm?abstract_id=301778, accessed 26 February 2008. Bharath, Sreedhar, Jayanthi Sunder and Shyam Vallabhajosyula Sunder (2004), ‘Accounting quality and debt contracting,’ http://papers.ssrn.com/ so13/papers.cfm?abstract_id=591342, accessed 28 Febrauary 2008. Blanchard, Olivier Jean (1990), ‘Why does money aﬀect output? A survey,’ in Benjamin Friedman and Frank Hahn (eds), Handbook of Monetary Economics, New York: North-Holland, pp. 779–835. Blanchard, Olivier Jean and Stanley Fischer (1989), Lectures on Macroeconomics, Cambridge, MA and London: MIT Press. Blume, Lawrence and David Easley (1993), ‘Economic natural selection,’ Economics Letters, 42 (2–3), 281–9. Blume, Lawrence and David Easley (1998), ‘Optimality and natural selection in markets,’ Cornell University working paper. Bordo, Michael and Finn Kydland (1995), ‘The gold standard as a rule: an essay in exploration,’ Explorations in Economic History, 32 (4), 423–64. Bordo, Michael and Anna Schwartz (1999), ‘Policy regimes and economic performance: the historical record,’ in John Taylor and Michael Woodford (eds), Handbook of Macroeconomics, Amsterdam: Elsevier, pp. 149–234. Brock, William (1990), ‘Overlapping generations models with money and transactions costs,’ in Benjamin Friedman and Frank Hahn (eds), Handbook of Monetary Economics, New York: North-Holland, pp. 263–95. Brown, James, Dino Falaschetti and Michael Orlando (2007), ‘Auditor independence and earnings quality: evidence for market discipline vs. Sarbanes-Oxley proscriptions,’ http://papers.ssrn.com/so13/papers.cfm? abstract_id=940847, accessed 26 February 2008. Brunner, Karl and Allan H. Meltzer (eds) (1976), The Phillips Curve and Labor Markets, Amsterdam, New York and Oxford: North-Holland Publishing Company. Burkart, Mike and Tore Ellingsen (2004), ‘In-kind ﬁnance: a theory of trade credit,’ American Economic Review, 94 (3), 569–90. Burkart, Mike, Denis Gromb and Fausto Panunzi (1997), ‘Large shareholders, monitoring, and the value of the ﬁrm,’ Quarterly Journal of Economics, 112 (3), 693–728.

162

Bibliography

Bush, George W. (2006), The Economic Report of the President 2006, Washington, DC: US Government Printing Oﬃce. Bushman, Robert and Abbie Smith (2001), ‘Financial accounting information and corporate governance,’ Journal of Accounting and Economics, 32 (1–3), 237–333. Calomiris, Charles and Joseph Mason (2003), ‘Consequences of bank distress during the Great Depression,’ American Economic Review, 93 (3), 937–47. Calvo, Guillermo (1978), ‘On the time inconsistency of optimal policy in a monetary economy,’ Econometrica, 46 (November), 1411–28. Calvo, Guillermo and Pablo Guidotti (1993), ‘On the ﬂexibility of monetary policy: the case of the optimal inﬂation tax,’ Review of Economic Studies, 60 (3), 667–87. Caporale, Tony and Kevin Grier (1998), ‘A political model of monetary policy with application to the real Fed Funds rate,’ Journal of Law and Economics, 41 (2), 409–28. Carlin, Wendy and Colin Mayer (2003), ‘Finance, investment and growth,’ Journal of Financial Economics, 69 (1), 191–226. Carpenter, Robert, Steven Fazzari and Bruce Petersen (1994), ‘Inventory investment, internal ﬁnance ﬂuctuations, and the business cycle,’ Brookings Papers on Economic Activity, 2, 75–138. Chan, Yuk-Shee, Stuart Greenbaum and Anjan Thakor (1992), ‘Is fairly priced deposit insurance possible?’ Journal of Finance, 47 (1), 227–45. Chappell, Henry, Thomas Havrilesky and Rob Roy McGregor (1993), ‘Partisan monetary policies: presidential inﬂuence through the power of appointment,’ Quarterly Journal of Economics, 108 (1), 185–218. Chari, V.V. and Patrick Kehoe (2004), ‘On the desirability of ﬁscal constraints in a monetary union,’ NBER working paper no. 10232. Chari, V.V., Patrick Kehoe and Edward Prescott (1989), ‘Time consistency and policy,’ in Robert Barro (ed.), Modern Business Cycle Theory, Cambridge, MA: Harvard University Press, pp. 265–305. Chow, Chee (1982), ‘The demand for external auditing: size, debt and ownership inﬂuences,’ Accounting Review, 57 (2), 272–91. Christiano, Lawrence and Terry Fitzgerald (1998), ‘The business cycle: it’s still a puzzle,’ Economic Perspectives, 22 (4), 56–83. Christiano, Lawrence, Martin Eichenbaum and Charles Evans (1997), ‘Sticky price and limited participation models: a comparison,’ European Economic Review, 41 (6), 1201–49. Christiano, Lawrence, Martin Eichenbaum and Charles Evans (1999), ‘Monetary policy shocks: what have we learned and to what end?’ in John Taylor and Michael Woodford (eds), Handbook of Macroeconomics, Amsterdam: Elsevier, pp. 65–148. Clement, Douglas (2003), ‘Price signals,’ The Region, 17 (3), 8–11.

Bibliography

163

Coase, Ronald (1982), ‘How should economists choose?’ The G. Warren Nutter Lecture in Political Economy presented to The American Enterprise Institute for Public Policy Research, Washington, DC. Cohen, Alma and Rajeev Dehejia (2004), ‘The eﬀect of automobile insurance and accident liability laws on traﬃc fatalities,’ Journal of Law and Economics, 47 (2), 357–93. Cole, Harold and Lee Ohanian (1999), ‘The Great Depression in the United States from a neoclassical perspective,’ Federal Reserve Bank of Minneapolis, Quarterly Review, 23 (1), 2–24. Cukierman, Alex (1995), Central Bank Strategy, Credibility, and Independence: Theory and Evidence, Cambridge, MA and London: MIT Press. Dee, Thomas and Rebecca Sela (2003), ‘The fatality eﬀects of highway speed limits by gender and age,’ Economics Letters, 79 (3), 401–8. Denhaan, Wouter, Steven Sumner and Guy Yamashiro (2004), ‘Bank loan components and the time-varying eﬀects of monetary policy shocks,’ CEPR discussion paper no. 4724. Descartes, René (2007), Discourse on Method and the Meditations, London: Penguin. Devlin, Keith (1998), The Language of Mathematics: Making the Invisible Visible, New York: Freeman. Diamond, Douglas (1991), ‘Monitoring and reputation: the choice between bank loans and directly placed debt,’ Journal of Political Economy, 99 (4), 689–721. Diamond, Douglas and Raghuram Rajan (2006), ‘Money in a theory of banking,’ American Economic Review, 96 (1), 30–53. Dixit, Avinash and Luisa Lambertini (2003), ‘Interactions of commitment and discretion in monetary and ﬁscal policies,’ American Economic Review, 93 (5), 1522–42. Dotsey, Michael and Peter Ireland (1996), ‘The welfare cost of inﬂation in general equilibrium,’ Journal of Monetary Economics, 37 (1), 29–47. Drazen, Allan (2000), Political Economy in Macroeconomics, Princeton, NJ: Princeton University Press. Drazen, Allan and Paul Masson (1994), ‘Credibility of policies versus credibility of policymakers,’ Quarterly Journal of Economics, 109 (3), 735–54. Driﬃll, John and Zeno Rotondi (2003), ‘Delegation of monetary policy: more than a relocation of the time-inconsistency problem,’ CEPR working paper. Dubey, Pradeep, John Geanakoplos and Martin Shubik (2002), ‘Is gold an eﬃcient store of value?’ Cowles Foundation discussion paper no. 1031R. The Economist (2003), ‘Alan Greenspan and the economy: irking the right,’ 8 March, 31, 34.

164

Bibliography

Eichengreen, Barry and Ashoka Mody (2004), ‘Do collective action clauses raise borrowing costs?’ Economic Journal, 114 (495), 247–64. Eijﬃnger, Sylvester and Jakob de Haan (1996), ‘The political economy of central bank independence,’ Princeton University Department of Economics Studies Papers in International Economics no. 19. Einzig, Paul, (1966), Primitive Money in its Ethnological, Historical and Economic Aspects, New York: Pergamon Press. Elster, Jon (1984), Ulysses and the Sirens: Studies in Rationality and Irrationality, Cambridge: Cambridge University Press. Falaschetti, Dino (2002a), ‘Golden parachutes: credible commitments or evidence of shirking?’ Journal of Corporate Finance, 8 (2), 159–78. Falaschetti, Dino (2002b), ‘Does partisan heritage matter? The case of the federal reserve,’Journal of Law, Economics, & Organization, 18 (2), 488–510. Falaschetti, Dino and Gary Miller (2001), ‘Constraining Leviathan: moral hazard and credible commitment in constitutional design,’ Journal of Theoretical Politics, 13 (4), 389–411. Falaschetti, Dino and Michael Orlando (2003), ‘Cutting the dividends tax . . . and corporate governance too?’ Corporate Ownership & Control, 3 (2), 31–4. Falaschetti, Dino and Michael Orlando (2004a), ‘Optimization in the production of governance services: lessons for the regulation of auditor independence,’ Financiele Studievereniging Rotterdam (FSR) Forum, 6 (2), 16–23. Falaschetti, Dino and Michael Orlando (2004b), ‘Is auditor independence endogenous? Evidence and implications for public policy,’ Federal Reserve Bank of Kansas City Research Division working paper no. 03–13. Fama, Eugene (1985), ‘What’s diﬀerent about banks?’ Journal of Monetary Economics, 15 (1), 29–39. Farmer, Roger (2002), Macroeconomics, 2nd edn, London and New York: South-Western. Fatas, Antonio and Ilian Mihov (2003), ‘The case for restricting ﬁscal policy discretion,’ Quarterly Journal of Economics, 118 (4), 1419–47. Fazzari, Steven, R. Glenn Hubbard and Bruce Petersen (1988), ‘Financing constraints and corporate investment,’ Brookings Papers on Economic Activity, 1, 141–95. Federal Reserve Bank of St Louis (2003), http://research.stlouisfed.org/ fred2/series/CPILFENS?&cid=9, accessed 28 Febrauary 2008. Fettig, David (1996), ‘Interview with James Tobin,’ The Region, http://www.minneapolisfed.org/pubs/region/96-12/tobin.cfm, accessed 28 February 2008. Foster, George (1980), ‘Externalities and ﬁnancial reporting,’ Journal of Finance, 35 (2), 521–33.

Bibliography

165

Frank, Mary Margaret, James Poterba, Douglas Shackelford and John Shoven (2004), ‘Copycat funds: information disclosure regulation and the returns to active management in the mutual fund industry,’ Journal of Law and Economics, 47 (2), 515–42. Frank, Robert (2006), Microeconomics and Behavior, 6th edn, New York: McGraw-Hill Irwin. Friedman, David (1996), Hidden Order: The Economics of Everyday Life, New York: HarperBusiness. Friedman, Milton (1953), ‘The methodology of positive economics,’ in Milton Friedman (ed.), Essays in Positive Economics, Chicago, IL: University of Chicago Press, pp. 3–43. Friedman, Milton (1992), Money Mischief: Episodes in Monetary History, New York, San Diego, CA and London: Harcourt Brace Jovanovich. Friedman, Milton and Anna Jacobson Schwartz (1963), A Monetary History of the United States, 1867–1960, Princeton, NJ: Princeton University Press. Giannetti, Mariassunta and Andrei Simonov (2003), ‘Which investors fear expropriation? Evidence from investors’ stock picking,’ Stockholm School of Economics working paper. Giavazzi, Francesco and Marco Pagano (1988), ‘The advantage of tying one’s hands: EMS discipline and central bank credibility,’ European Economic Review, 32 (5), 1055–75. Gibbard, Alan (1973), ‘Manipulation of voting schemes: a general result,’ Econometrica, 41 (4), 587–601. Gilligan, T.W. and K. Krehbiel (1987), ‘Collective decision-making and standing committees: an informational rationale for restrictive amendment procedures,’ Journal of Law, Economics, & Organization, 3(2), 287–335. Gorton, Gary and Lixin Huang (2004), ‘Liquidity, eﬃciency, and bank bailouts,’ American Economic Review, 94 (3), 455–83. Grilli, Vittorio, Donato Masciandaro and Guido Tabellini (1991), ‘Political and monetary institutions and public ﬁnancial policies in the industrial countries,’ Economic Policy, 6 (13), 341–92. Grossman, Sanford and Oliver Hart (1980), ‘Takeover bids, the free-rider problem, and the theory of the corporation,’ Bell Journal of Economics, 11 (1), 42–62. Hall, Robert (2003), ‘Wage determination and employment ﬂuctuations,’ NBER working paper no. 9967. Hall, Robert (2005), ‘Employment ﬂuctuations with equilibrium wage stickiness,’ American Economic Review, 95 (1), 50–65. Hamilton, William (1964), ‘The genetical evolution of social behavior, I and II,’ Journal of Theoretical Biology, 7 (1), 1–52.

166

Bibliography

Haslag, Joseph (1998), ‘Monetary policy, banking, and growth,’ Economic Inquiry, 36 (3), 489–500. Hetzel, Robert (1990), ‘The political economy of monetary policy,’ in Thomas Mayer (ed.), The Political Economy of American Monetary Policy, Cambridge: Cambridge University Press. Hibbs, Douglas (1987), The Political Economy of Industrial Democracies, Cambridge, MA and London: Harvard University Press. Hilton, Peter (1997), ‘Foreword: mathematics in our culture,’ in Jan Gullberg (ed.), Mathematics: From the Birth of Numbers, New York: W.W. Norton & Company, pp. xvii–xxii. Hirshleifer, Jack (1977) ‘Economics from a biological viewpoint,’ Journal of Law and Economics, 20 (1), 1–52. Holmstrom, Bengt (1982), ‘Moral hazard in teams,’ Bell Journal of Economics, 13 (2), 324–40. Holmstrom, Bengt (1999), ‘The ﬁrm as a subeconomy,’ Journal of Law, Economics & Organization, 15 (1), 74–102. Holmstrom, Bengt and Steven Kaplan (2003), ‘The state of US corporate governance: what’s right and what’s wrong?’ NBER working paper no. 9613. Holthausen, Robert and Richard Leftwich (1983), ‘The economic consequences of accounting choice: implications of costly contracting and monitoring,’ Journal of Accounting and Economics, 5 (2), 77–117. Hoshi, Takeo and Anil Kashyap (2004), ‘Japan’s ﬁnancial crisis and economic stagnation,’ Journal of Economic Perspectives, 18 (1), 3–26. Houston, Joel and Christopher James (2001), ‘Do relationships have limits? Banking relationships, ﬁnancial constraints, and investment,’ Journal of Business, 74 (3), 347–74. James, Christopher (1987), ‘Some evidence on the uniqueness of bank loans,’ Journal of Financial Economics, 19 (2), 217–35. James, Christopher and Peggy Wier (1988), ‘Are bank loans diﬀerent? Some evidence from the stock market,’ Journal of Applied Corporate Finance, 1 (2), 46–54. Jayaratne, Jith and Philip Strahan (1996), ‘The ﬁnance–growth nexus: evidence from branch banking deregulation,’ Quarterly Journal of Economics, 111 (3), 639–70. Jensen, Michael (1986), ‘Agency costs of free cash ﬂow, corporate ﬁnance, and takeovers,’ American Economic Review, Papers and Proceedings, 76 (2), 323–9. Jensen, Michael and William Meckling (1976), ‘Theory of the ﬁrm: managerial behavior, agency costs, and ownership structure,’ Journal of Financial Economics, 3 (4), 305–60. John, Kose, Lubomir Litov and Bernard Yeung (2004), ‘Corporate governance, risk-taking, and growth,’ Stern School of Business working paper.

Bibliography

167

Kane, Edward (2003), ‘Continuing dangers of disinformation in corporate accounting reports,’ NBER working paper no. 9634. Keefer, Philip and David Stasavage (2003), ‘The limits of delegation: veto players, central bank independence, and the credibility of monetary policy,’ World Bank working paper. King, Robert and Ross Levine (1993a), ‘Financial intermediation and economic development,’ in Colin Mayer and Xavier Vives (eds), Capital Markets and Financial Intermediation, Cambridge: Cambridge University Press, pp. 156–89. King, Robert and Ross Levine (1993b), ‘Finance and growth: Schumpeter might be right,’ Quarterly Journal of Economics, 108 (3), 717–37. King, Robert and Ross Levine (1993c), ‘Finance, entrepreneurship and growth: theory and evidence,’ Journal of Monetary Economics, 32 (3), 513–42. King, Robert and Charles Plosser (1986), ‘Money as the mechanism of exchange,’ Journal of Monetary Economics, 17 (1), 93–115. Koopmans, Tjalling (1957), Three Essays on the State of Economic Science, New York: McGraw-Hill Book Company. Kortum, Samuel and Josh Lerner (2000), ‘Assessing the contribution of venture capital to innovation,’ Rand Journal of Economics, 31 (4), 674–92. Kotter, John and James Heskett (1992), Corporate Culture and Performance, New York: The Free Press. Krainer, Robert (1985), The Financial System and the Coordination of Economic Activity, New York: Praeger. Krainer, Robert (2003), Corporate Finance, Governance and Business Cycles: Theory and International Comparisons, Amsterdam: Elsevier. Krebs, J.R. and N.B. Davies (1993), An Introduction to Behavioural Ecology, 3rd edn, Oxford: Blackwell Scientiﬁc Publications. Kroszner, Randall and Raghuram Rajan (1994), ‘Is the Glass-Steagall Act justiﬁed? A study of the US experience with universal banking before 1933,’ American Economic Review, 84 (4), 810–32. Kroszner, Randall and Raghuram Rajan (1997), ‘Organization structure and credibility: evidence from commercial bank securities activities before the Glass-Steagall Act,’ Journal of Monetary Economics, 39 (3), 475–516. Kydland, Finn and Edward Prescott (1977), ‘Rules rather than discretion: the inconsistency of optimal plans,’ Journal of Political Economy, 85 (3), 473–91. La Porta, Rafael, Florencio Lopez-de-Silanes and Andrei Shleifer (1999), ‘Corporate ownership around the world,’ Journal of Finance, 54 (2), 471–517.

168

Bibliography

La Porta, Rafael, Florencio Lopez-de-Silanes Andrei Shleifer and Robert Vishny (1997), ‘Legal determinants of external ﬁnance,’ Journal of Finance, 52 (3), 1131–50. La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer and Robert Vishny (1998), ‘Law and ﬁnance,’ Journal of Political Economy, 106 (6), 1113–55. Lakatos, Imre (1978), The Methodology of Scientiﬁc Research Programmes, Philosophical Papers, vol 1, Cambridge, New York and Melbourne: Cambridge University Press. Lazonick, William and Mary O’Sullivan (2000), ‘Perspectives on corporate governance, innovation, and economic performance,’ Corporate Governance, Innovation, and Economic Performance in the EU (CGEP) working paper. Levine, Ross (1997), ‘Financial development and economic growth: views and agenda,’ Journal of Economic Literature, 35 (2), 688–726. Levine, Ross (1998), ‘The legal environment, banks, and long-run economic growth,’ Journal of Money, Credit, and Banking, 30 (3), 596–613. Levine, Ross and Sara Zervos (1998), ‘Stock markets, banks, and economic growth,’ American Economic Review, 88 (3), 537–58. Lucas, Robert E. Jr. (1976), ‘Econometric policy evaluation: a critique,’ in Karl Brunner and Allan Meltzer (eds), The Phillips Curve and Labor Markets, Amsterdam, New York and Oxford: North-Holland Publishing Company, pp. 19–46. Lucas, Robert E. Jr. (1980), ‘Methods and problems in business cycle theory,’ Journal of Money, Credit, and Banking, 12 (4), 696–715. Lucas, Robert E. Jr. (1994), ‘On the welfare cost of inﬂation,’ Federal Reserve Bank of San Francisco working papers in applied economic theory no. 94–07. Lucas, Robert E. Jr. (2003), ‘Macroeconomic priorities,’ American Economic Review, 93 (1), 1–14. Lummer, Scott and John McConnell (1989), ‘Further evidence on the bank lending process and the capital market response to bank loan agreements,’ Journal of Financial Economics, 25 (1), 99–122. Manove, Michael, A. Jorge Padilla and Marco Pagano (2001), ‘Collateral versus project screening: a model of lazy banks,’ RAND Journal of Economics, 32 (4), 726–44. Mayer, Colin (1990), ‘Financial systems, corporate ﬁnance, and economic development,’ in R. Glenn Hubbard (ed.), Asymmetric Information, Corporate Finance, and Investment, Chicago, IL and London: University of Chicago Press, pp. 307–32. McFadden, Daniel (2006), ‘Free markets and fettered consumers,’ American Economic Review, 96 (1), 5–29.

Bibliography

169

Meltzer, Allan (2003), A History of the Federal Reserve, Volume 1: 1913–1951, Chicago, IL: University of Chicago Press. Merton, Robert (1977), ‘An analytical derivation of the cost of deposit insurance and loan guarantees,’ Journal of Banking and Finance, 1 (1), 3–11. Metrick, Andrew (2007), Venture Capital and the Finance of Innovation, Hoboken, NJ: Wiley. Miller, Gary (1991), Managerial Dilemmas: The Political Economy of Hierarchy, Cambridge, New York and Melbourne: Cambridge University Press. Miller, Preston (1975), ‘On macroeconomic theories and models,’ in Preston Miller (ed.) A Prescription for Monetary Policy: Proceedings from a Seminar Series, Minneapolis, MN: Federal Reserve Bank of Minneapolis. Minsky, Hyman (1964), ‘Longer waves in ﬁnancial relations: ﬁnancial factors in the more severe depressions,’ American Economic Review: Papers and Proceedings, 54 (3), 324–35. Mints, Lloyd (1975), A History of Banking Theory, Chicago, IL: University of Chicago Press. Mishkin, Frederic (2004), The Economics of Money, Banking, and Financial Markets, 7th edn, New York: Addison Wesley. Morck, Randall, Daniel Wolfenzon and Bernard Yeung (2004), ‘Corporate governance, economic entrenchment and growth,’ NBER working paper no. 10692. Morgan, Donald, Bertrand Rime and Philip Strahan (2004), ‘Bank integration and state business cycles,’ Quarterly Journal of Economics, 119 (4), 1555–84. Mundell, Robert (1961), ‘A theory of optimum currency areas,’ American Economic Review, 51 (4), 657–65. Muth, John (1961), ‘Rational expectations and the theory of price movements,’ Econometrica, 29 (3), 315–35. Myers, Stewart and Nicholas Majluf (1984), ‘Corporate ﬁnancing and investment decisions when ﬁrms have information that investors do not have,’ Journal of Financial Economics, 13 (2), 187–221. Nietzsche, Friedrich (1887), On the Genealogy of Morals, Oxford: Oxford University Press. North, Douglass (1990), Institutions, Institutional Change and Economic Performance, Cambridge, New York and Melbourne: Cambridge University Press. North, Douglass (1994), ‘Economic performance through time,’ American Economic Review, 84 (3), 359–68. Nickell, Stephen and Glenda Quintini (2003), ‘Nominal wage rigidity and the rate of inﬂation,’ Economic Journal, 113 (490), 762–81.

170

Bibliography

Orphanides, Athanasios and Robert Solow (1990), ‘Money, inﬂation and growth,’ in Benjamin Friedman and Frank Hahn (eds), Handbook of Monetary Economics, New York: North-Holland, pp. 223–61. Ostroy, Joseph and Ross Starr (1990), ‘The transactions role of money,’ in Benjamin Friedman and Frank Hahn (eds), Handbook of Monetary Economics, New York: North-Holland, pp. 3–62. Pagano, Marco (1993), ‘Financial markets and growth: an overview,’ European Economic Review, 37 (2–3), 613–22. Pagano, Marco and Paolo Volpin (2000), ‘The political economy of corporate governance,’ Center for Studies in Economics and Finance working paper no. 29. Patinkin, Don (1965), Money, Interest, and Prices, New York: Harper and Row. Payne, J.W., J.R. Bettman and E.J. Johnson (1993), The Adaptive DecisionMaker, Cambridge: Cambridge University Press. Persson, Torsten and Lars Svensson (1989), ‘Why a stubborn conservative would run a deﬁcit: policy with time inconsistent preferences,’ Quarterly Journal of Economics, 104 (2), 325–45. Persson, Torsten and Guido Tabellini (1992), ‘Growth, distribution, and politics,’ European Economic Review, 36 (2–3), 593–602. Persson, Torsten and Guido Tabellini (1993), ‘Designing institutions for monetary stability,’ Carnegie-Rochester Conference Series on Public Policy, 39, 53–89. Persson, Torsten and Guido Tabellini (1999), ‘Political economics and macroeconomic policy,’ in John Taylor and Michael Woodford (eds), Handbook of Macroeconomics, Amsterdam: Elsevier, pp. 1397–482. Poirier, Dale (1995), Intermediate Statistics and Econometrics: A Comparative Approach, Cambridge, MA and London: MIT Press. Posen, Adam (1993), ‘Why central bank independence does not cause low inﬂation: there is no institutional ﬁx for politics,’ in Richard O’Brien (ed.), Finance and the International Economy, vol 7, Oxford: Oxford University Press. Posen, Adam (1995), ‘Declarations are not enough: ﬁnancial sector services of central bank independence,’ in Ben Bernanke and Julio Rotemberg (eds), NBER Macroeconomics Annual 1995, Cambridge, MA: MIT Press, pp. 253–74. Posen, Adam (1998), ‘Do better institutions make better policy?’ International Finance, 1 (1), 173–205. Public Broadcasting Service (2004), The History of Money online tutorial, http: //www.pbs.org/wgbh/nova/moolah/history.html, accessed 28 February 2008. Radford, R.A. (1945), ‘The economic organization of a POW camp,’ Economica, 12 (48), 189–201.

Bibliography

171

Rajan, Raghuram (1992), ‘Insiders and outsiders: the choice between relationship and arm’s-length debt,’ Journal of Finance, 47 (4), 1367–400. Rajan, Raghuram and Luigi Zingales (1998), ‘Financial dependence and growth,’ American Economic Review, 88 (3), 559–86. Rajan, Raghuram and Luigi Zingales (2003), Saving Capitalism from the Capitalists: Unleashing the Power of Financial Markets to Create Wealth and Spread Opportunity, New York: Crown Business. Redman, Deborah (1997), The Rise of Political Economy as a Science: Methodology and the Classical Economists, Cambridge, MA and London: MIT Press. Richerson, Peter and Robert Boyd (2001), ‘The evolution of subjective commitment to groups: a tribal instincts hypothesis,’ in Randolph Nesse (ed.), Evolution and the Capacity for Commitment, New York: Russell Sage Foundation, pp. 186–220. Robson, Arthur (2001a), ‘The biological basis of economic behavior,’ Journal of Economic Literature, 39 (1), 11–33. Robson, Arthur (2001b), ‘Why would nature give individuals utility functions?’ Journal of Political Economy, 109 (4), 900–914. Roe, Mark (1994), Strong Managers, Weak Owners: The Political Roots of Corporate Finance, Princeton, NJ: Princeton University Press. Roe, Mark (2003), Political Determinants of Corporate Governance: Political Context, Corporate Impact, New York: Oxford University Press. Rogoﬀ, Kenneth (1985), ‘The optimal degree of commitment to an intermediate monetary target,’ Quarterly Journal of Economics, 100 (4), 1169–89. Rogoﬀ, Kenneth (1987), ‘Reputational constraints on monetary policy,’ in Karl Brunner and Allan Meltzer (eds), Carnegie-Rochester Conference Series on Public Policy, vol 26, Amsterdam, New York and Oxford: North-Holland Publishing Company, pp. 141–81. Rogoﬀ, Kenneth and Anne Sibert (1988), ‘Elections and macroeconomic policy cycles,’ Review of Economic Studies, 55 (1), 1–16. Rolnick, Arthur (1996), ‘Interview with Edward C. Prescott,’ The Region, http://www.minneapolisfed.org/pubs/region/96-09/prescott.cfm;s=0, accessed 28 Febrauary 2008. Romer, Christina and David Romer (1989), ‘Does monetary policy matter? A new test in the spirit of Friedman and Schwartz,’ in Olivier Jean Blanchard and Stanley Fischer (eds), NBER Macroeconomics Annual, 4, 121–70. Romer, David (2001), Advanced Macroeconomics, 2nd edn, New York: McGraw-Hill/Irwin. Satterthwaite, Mark (1975), ‘Strategy-proofness and Arrow’s conditions,’ Journal of Economic Theory, 10 (2), 187–217.

172

Bibliography

Schelling, Thomas (1980 [1960]), The Strategy of Conﬂict, Cambridge and London: Harvard University Press. Schreft, Stacey (1992), ‘Transaction costs and the use of cash and credit,’ Economic Theory, 2 (2), 283–96. Schreft, Stacey and Bruce Smith (2000), ‘The evolution of cash transactions: some implications for monetary policy,’ Journal of Monetary Economics, 46 (1), 97–120. Selgin, George (2003), ‘Adaptive learning and the transition to ﬁat money,’ Economic Journal, 113 (484), 147–65. Shapiro, Carl and Joseph Stiglitz (1984), ‘Equilibrium unemployment as a worker discipline device,’ American Economic Review, 74 (3), 433–44. Shimer, Robert (2005), ‘The cyclical behavior of equilibrium unemployment and vacancies,’ American Economic Review, 95 (1), 25–49. Shivdasani, Anil (1993), ‘Board composition, ownership structure, and hostile takeovers,’ Journal of Accounting and Economics, 16 (1–3), 167–98. Shleifer, Andrei (2000), Ineﬃcient Markets: An Introduction to Behavioral Finance, New York: Oxford University Press. Shleifer, Andrei and Lawrence H. Summers (1989), ‘Breach of trust in hostile takeovers,’ in Alan J. Auerbach (ed.), Corporate Takeovers: Causes and Consequences, Chicago and London: University of Chicago Press, pp. 33–68. Shleifer, Andrei and Robert Vishny (1986), ‘Large shareholders and corporate control,’ Journal of Political Economy, 94 (3), 461–88. Shleifer, Andrei and Robert Vishny (1997), ‘A survey of corporate governance,’ Journal of Finance, 52 (2), 737–83. Shubik, Martin (1990), ‘A game theoretic approach to the theory of money and ﬁnancial institutions,’ in Benjamin Friedman and Frank Hahn (eds), Handbook of Monetary Economics, New York: North-Holland, pp. 171–219. Sidrauski, Miguel (1967a), ‘Rational choice and patterns of growth in a monetary economy,’ American Economic Review, Papers and Proceedings, 57 (2), 534–44. Sidrauski, Miguel (1967b), ‘Inﬂation and economic growth,’ Journal of Political Economy, 75 (6), 796–810. Silberberg, Eugene and Wing Suen (2001), The Structure of Economics: A Mathematical Analysis, 3rd edn, New York: Irwin McGraw-Hill. Singh, Ajit (2003), ‘Competition, corporate governance and selection in emerging markets,’ Economic Journal, 113 (491), F443–F464. Skaperdas, Sergios (2003), ‘Restraining the genuine Homo Economicus: why the economy cannot be divorced from its governance,’ Economics and Politics, 15 (2), 135–62.

Bibliography

173

Solow, Robert (1956), ‘A contribution to the theory of economic growth,’ Quarterly Journal of Economics, 70 (1), 65–94. Stock, James and Mark Watson (1999), ‘Business cycle ﬂuctuations in US macroeconomic time series,’ in John Taylor and Michael Woodford (eds), Handbook of Macroeconomics, Amsterdam: Elsevier, pp. 3–64. Stout, Lynn (2002), ‘Do antitakeover defenses decrease shareholder wealth? The ex post/ex ante valuation problem,’ Stanford Law Review, 55 (December), 845–61. Tabellini, Guido and Alberto Alesina (1988), ‘Voting on the budget deﬁcit,’ American Economic Review, 80 (1), 37–49. Timberlake, Richard (1993), Monetary Policy in the United States: An Intellectual and Institutional History, Chicago, IL: University of Chicago Press. Tobin, James (1965), ‘Money and economic growth,’ Econometrica, 33 (4), 671–84. Tootell, Geoﬀrey (1996), ‘Appointment procedures and FOMC voting behavior,’ Southern Economic Journal, 63 (1), 191–204. Trivers, Robert (2002), Natural Selection and Social Theory: Selected Papers of Robert Trivers, New York: Oxford University Press. Velde, Francois (1998), ‘Lessons from the history of money,’ Economic Perspectives, 22 (1), 2–16. Vickers, John (1985), ‘Delegation and the theory of the ﬁrm,’ Economic Journal, 95 (issue supplement: conference papers), 138–47. Vives, Xavier (2000), ‘Corporate governance: does it matter?’ in Xavier Vives (ed.), Corporate Governance: Theoretical and Empirical Perspectives, Cambridge: Cambridge University Press, pp. 1–21. Wald, Abraham (1940), ‘The ﬁtting of straight lines if both variables are subject to error,’ Annals of Mathematical Statistics, 11 (3), 284–300. Waller, Christopher J. (1992), ‘A bargaining model of partisan appointments to the central bank,’ Journal of Monetary Economics, 29 (3), 411–28. Walsh, Carl (1995), ‘Optimal contracts for central bankers,’ American Economic Review, 85 (1), 150–67. Weber, Axel (1992), ‘The role of policymakers’ reputation in the EMS disinﬂations: an empirical evaluation,’ European Economic Review, 36 (7), 1473–92. White, Michelle (2004), ‘The “arms race” on American roads: the eﬀect of sport utility vehicles and pickup trucks on traﬃc safety,’ Journal of Law and Economics, 47 (2), 333–56. Wicker, Elmus (1996), The Banking Panics of the Great Depression, Cambridge: Cambridge University Press. Willett, Thomas D. and Edward Keen (1990), ‘Studying the Fed: toward a broader public choice perspective,’ in Thomas Meyer (ed.), The Political

174

Bibliography

Economy of American Monetary Policy, Cambridge: Cambridge University Press. Williamson, Oliver (1985), The Economic Institutions of Capitalism, New York: Free Press. Williamson, Stephen and Randall Wright (1994), ‘Barter and monetary exchange under private information,’ American Economic Review, 84 (1), 104–23. Woodford, Michael (1994), ‘Notes on dynamic eﬃciency wage models,’ http://www.columbia.edu/%7EMW2230/EfficiencyWages.pdf, accessed 28 February 2008. Woodford, Michael (1999), ‘Revolution and evolution in twentieth-century macroeconomics,’ http://www.columbia.edu/%7EMW2230/macro20C. pdf, accessed 28 February 2008. Woolley, John (1984), Monetary Politics: The Federal Reserve and the Politics of Monetary Policy, Cambridge: Cambridge University Press. Yap – Land of Stone Money (2003), http://www.fm/yap.htm, accessed 28 February 2008. Yellen, Janet (1984), ‘Eﬃciency wage models of unemployment,’ American Economic Review, Papers and Proceedings, 74 (2), 200–205.

Index accountability 150 accounting service ﬁrms 133 accounting systems 136–8, 140 activist monetary policy case against 63, 65 inﬂation–unemployment trade-oﬀ policy 70, 80–82 Lucas critique and policy evaluation 80–84 and stabilization policy, beneﬁts of 84–5 and time inconsistency problem 87–8, 89–93 adverse selection 116–18, 126–7, 136–7 adverse selection mitigation 122–3, 126–7, 132, 136 agency costs 139 agency problems 131–3, 140 see also separation aggregate demand in classical economy’s demand side 59–60, 61–2, 64 and money stock changes in classical economy 61–2 and wage stickiness 76 aggregate demand shocks 75 aggregate output, and time inconsistency problem 90 aggregate supply in classical economy’s supply side 56–8, 60, 61, 62–3, 64 in Keynesian economy 67–8, 69–70 Lucas critique and Phillips Curve breakdown 81 and negative technology shocks in classical economy 62–3, 64 Akerlof, George 75, 79, 136 Alchian, Armen 17 Alesina, Alberto 107–8 animal evidence, in maximizingbehavior axiom 18–19

appointed political agents, and time inconsistency problem 111 ‘as if’ principle 16–17, 18–19, 69 asset prices 68–9 asset weights, in monetary service measurement 100–101 assets, ‘moneyness’ of 35–7 assumptions, in models 10, 11 see also axiomatic approach; maximizing-behavior axiom asymmetric information and accounting systems 136–8, 140 and adverse selection 116–18, 136 managers’ incentives for strategic use of 138, 140–41, 142 and moral hazard 118–20, 140–41, 142 and tax policy 142 asymmetric information mitigation and adverse selection in ﬁnancial intermediaries 122–3, 125 and banks 124–6, 138 and governance of ﬁnancial intermediaries 124–6 and moral hazard in ﬁnancial intermediaries 123–4 auditing 133, 137, 138 axiomatic approach 15–16 see also maximizing-behavior axiom backward induction, and time inconsistency problem 90 banks asymmetric information mitigation 124–6, 138 debt ﬁnance 140, 141, 142 in model economy 28–31 monitoring 141, 142 restriction reductions and economic performance 146–7

175

176

Index

see also central bank independence (CBI); checkable deposits, ‘moneyness’ of; commercial banks; conservative banker; investment banks; relationship banking; universal banks bargaining sets, and eﬃciency wage theories 75 Barlevy, Gadi 85, 107 Barnett, William 100 barter economy 42–5, 46, 47, 49, 50 see also commodity money Baye, Michael 22 Becker, Gary 15, 22, 23 benchmark assets, in monetary service measurement 100, 101 benevolent dictators, and time inconsistency problem 90–93 Berman, Eli 20, 21, 24 Besanko, David 23, 31 bias, omitted variable 12 Blanchard, Olivier Jean 68, 71 boards of directors 139–40 Brown, James 138 Brunner, Karl 70 budget constraints barter economy 43, 44 and maximizing-households in model economy 25, 26, 27, 30 see also cash constraints business cycles, real (RBCs) 62–3, 64, 65, 66 see also cyclical output; cyclical unemployment capital, higher-value uses 115, 146 capital markets 30–31, 136, 137 capital structure debt ﬁnance and governance services 140–43 equity ﬁnance and governance services 139–40, 141–2 and tax policy 142–3 capitalists 149 ‘capture’ 131–3, 140 see also separation cash constraint channel 146, 147 cash constraints and asymmetric information 120 and debt ﬁnance 140–43

and dividends tax reduction 142–3 and equity ﬁnance 139–40, 141–2 and ﬁnancial development 149 and maximizing-ﬁrms in model economy 30–31, 115 and project-constrained lenders 115, 145–6 see also budget constraints cash ﬂows 140–41 causation 12, 13, 107 central bank independence (CBI) 106–8 CH (clearinghouse), in formal model of superior monetary equilibrium 46, 47, 49, 50, 107 checkable deposits, ‘moneyness’ of 37 Christiano, Lawrence 77, 78, 79 cigarettes, as currency 46, 50 classical economy, money in activist monetary policy, case against 63, 65 demand side 57–60 equilibrium 60–61 money, prices and real activity 61–3, 64 supply side 53–7, 58 clearinghouses (CH), in formal model of superior monetary equilibrium 46, 47, 49, 50, 107 collateral requirements, and adverse selection mitigation 122 commercial banks 132–3, 134, 138 commitment 21, 24, 39, 40, 89–90, 102–6 commodity money 36–7, 38, 39 compensation schemes, managers’ 138, 140 competition 146, 147, 149 concentration of ownership 139, 150 conﬁdence of ﬁnanciers in ﬁrms 136, 137, 146 in money as exchange medium 49 conﬂicts of interest 131–3, 140 see also separation Congress 110, 111 conservative banker 106

Index constituents’ rational expectations, and time inconsistency problem 87, 89–90, 91, 92–3, 106–7 constraints 18, 21 see also budget constraints; cash constraints; ‘No shirking constraint’ curve; project constraints; technology constraints consultation 133 consumer prices 39, 40, 53 consumption 84–5 consumption bundles 42, 43, 44 control, separation from ownership 139–40, 146, 147, 150 corporate governance accounting systems and asymmetric information 136–8 capital structure and governance services 139–43 reforms, and eﬃcacy of GlassSteagall Act (1933) 131–3 correlation 12, 107 costs 49, 141 see also agency costs; currency service costs; governance costs; information costs; marginal costs; opportunity costs; resource costs; transaction costs; transportation costs; user costs CPI (consumer price index) 53 creative destruction 150 credibility 21, 24, 103, 138 credit services 78 see also banks; debt ﬁnance currency 35, 100–101 see also barter economy; exchange and exchange media; monetary economy; money currency service beneﬁts 38–9 currency service costs 38, 39 cyclical output 82, 83 cyclical unemployment 75 debt ﬁnance 140–43 see also banks; credit services; debt restructuring debt restructuring 123–4 debt violations 141

177

delegation 105–6, 109–11 demand in barter versus money endowment economy 44 in barter versus money production economy 48 in classical economy 57–62 law of 26, 31 and Lucas critique and Phillips Curve breakdown 82 and maximizing-households in model economy 25–7 see also aggregate demand; aggregate demand shocks; labor demand deposit insurance, United States 126–7, 128–31 Devlin, Keith 15 disclosure of ﬁnancial performance 136, 137, 138 distributional issues 149, 150 dividends tax reductions 142–3 ‘double coincidence of wants,’ in barter economies 42–3, 44, 45, 46 Drazen, Allan 102, 106, 107 dynamic economy, money in see classical economy, money in; Keynesian economy, money in; monetary policies economic ﬂuctuations see aggregate demand shocks; cyclical output; cyclical unemployment; ‘liquidity crises’; monetary policy shocks; negative supply shocks; real activity; real business cycles (RBCs); technology shocks economic method of inquiry advantages of 22–3 described 15–16 foundation see foundation of economic method of inquiry importance 8–9 model economy see model economy modeling 9–14 see also maximizing-behavior axiom economic models see model economy; modeling economic performance 137–8, 143, 145–50

178

Index

economic stability 148 economic well-being see welfare eﬃciency, in barter versus money endowment economy 48 eﬃciency wage theories 70–75, 79 eﬃcient market hypothesis 68–9 eﬀort, and eﬃciency wage theories 71, 72–4 egoists, in models 9 see also maximizing-behavior axiom Eichengreen, Barry 123–4 elected political agents, time inconsistency problem 111 employment 67–8 see also eﬃciency wage theories; employment equilibrium; employment policies; labor demand; labor-manager coalitions; labor markets; labor prices; labor supply; nominal wages; real wages; unemployment; wage agreements; wage stickiness employment equilibrium 56–8, 60, 74 employment policies 68, 70, 80–82 endowment economies 42–5, 46–7 enforceable rules, monetary services 103 entrepreneurs 146, 148, 149 ‘envelope theorem’ 75 equity ﬁnance 139–40, 141–2 evaluation, of models 10–11, 17–21 evaluation services 141, 142 evidence, in modeling 11–13, 18–21 exchange and exchange media 35–40, 50, 99–101 see also barter economy; monetary economy; money expectations-augmented Phillips Curve 69–70, 80–81, 90–91 Falaschetti, Dino 111, 138, 139 Fama, Eugene 125 Federal Open Market Committee (FOMC) 109, 110, 111 Federal Reserve Board (FRB) 109, 110, 111, 142 Federal Reserve System 99–100, 109–11

fei 36–7 ﬁat money 37–9, 40 ﬁnancial crises see ‘liquidity crises’; monetary policy shocks; pensions crisis, emerging (US); savings and loan crisis (US); ﬁnancial development, and economic performance 145–50 ﬁnancial intermediaries adverse selection 117, 126–7, 136–7 adverse selection mitigation 122–3, 126–7, 132 and disclosure of ﬁnancial performance 136, 137 ﬁnancial structure 122–4 governance see governance of ﬁnancial intermediaries monetary policy shocks, eﬀect on credit services 78 moral hazard 118–20 moral hazard mitigation 123–4, 127, 140 project constraints 115 see also accounting service ﬁrms; banks; credit services; debt ﬁnance; equity ﬁnance; monetary services ﬁnancial performance disclosure, and ﬁnancial intermediation 136, 137, 138 ﬁnancial statements 133, 137, 138 ﬁnancial structure of ﬁnancial intermediaries 122–4 ﬁrm-speciﬁc investment 141–3 ﬁrms, maximization see proﬁtmaximization ﬁscal versus monetary authorities 106–7 see also tax policy Fischer, Stanley 71 ﬂexible rules, monetary services 103 FOMC (Federal Open Market Committee) 109, 110, 111 foundation of economic method of inquiry advantages of economic method 22–3 axiomatic approach 15–16

Index maximizing-behavior axiom 16–17 positive versus normative analysis 21–2 prediction from maximizingbehavior axiom 17–21 FRB (Federal Reserve Board) governors 109, 110, 111, 142 free cash ﬂows 140–41, 142 free riders 139, 141, 142 Friedman, David 7, 16 Friedman, Milton 10–11, 23, 36, 37, 41, 69–70, 77 full employment policies 68, 70 games, and time inconsistency problem 88–9 Gibbard, Alan 105–6 Glass-Steagall Act (1933) 131–3, 138 governance, corporate see corporate governance governance costs 141, 142–3 governance of ﬁnancial intermediaries asymmetric information 126–7, 129, 131 market discipline versus public regulation 132–3 regulatory ‘mistakes,’ repetition of 127–33 governance of money and capital markets 28–31 governance services, and capital structure 139–43 Great Depression (US) 126–7 Greenspan, Alan 142 Hall, Robert 75 hangman game, and timeinconsistency problem 88–9 higher-value uses, capital 115, 146 Holmstrom, Bengt 137–8 hostile take-overs 141–2 households, maximization see maximizing-households housing units, ‘moneyness’ of 35 human evidence, in maximizingbehavior axiom 19–20 hypothesis-testing 11–13, 14

179

‘idea rich’ entrepreneurs 146, 149 incentives 137, 138, 139, 140–41, 142 indiﬀerence curve 25–7 induction, backward, and time inconsistency problem 90 industrial services 38, 39 inferior barter equilibrium 42–5 inferior economic performance 148–50 inﬂation and central bank independence (CBI) 107–8 and ﬁat money supply 39, 40 and Keynesian economy’s supply side 67, 68 and Lucas critique and Phillips Curve breakdown 81, 82, 84 and money in classical economy 61–2 and Phillips Curve 68, 69, 70, 80–82, 83, 84 and time inconsistency problem 90, 91, 92–3, 111 inﬂation–unemployment trade-oﬀ policy 70, 80–82 information see asymmetric information; asymmetric information mitigation; information costs; information evaluation; information generation; price information; private information information costs 137 information evaluation 125–6 information generation 125 input prices 67 institutions, monetary services 102–6 insulation and delegation of ﬁscal authority 106–7 and delegation of monetary authority 105–6 Federal Reserve System 110, 111 intellectual property protection 124 interest payments 142 investment, and stabilization policy 85 investment, ﬁrm-speciﬁc 141–3 investment banks 132–3, 134, 138

180

Index

involuntary unemployment, and eﬃciency wage theories 72–4, 75 irrationality, and prediction 16, 17 iso-proﬁt curve, maximizing-ﬁrms in model economy 27, 28, 29, 30 James, Christopher 125, 141 Jayaratne, Jith 148 Jensen, Michael 140 Kane, Edward 138 Kaplan, Steven 137–8 Keen, Edward 110 Keynesian economy, money in eﬃciency wage theories 70–75, 79 monetary transmission, other mechanisms 76–8 nominal rigidities, evidence against 76 supply side 67–70 Klenow, Peter 76, 77 Kroszner, Randall 132–3, 134, 138 Kydland, Finn 103 labor see eﬃciency wage theories; employment; employment equilibrium; employment policies; labor demand; labor-manager coalitions; labor markets; labor prices; labor supply; nominal wages; real wages; unemployment; wage agreements; wage stickiness labor demand 53–6, 57, 62–3, 67–8 labor-manager coalitions 149–50 labor markets in classical economy’s supply side 53–6, 57, 62–3, 64 and eﬃciency wage theories 71–3 equilibrium 56–8, 60, 74 labor prices 53, 54–5, 56, 57, 67 labor supply 56, 57, 67, 72–4 law of demand 26, 31 laws, monetary services 102 see also Glass-Steagall Act (1933); Sarbanes-Oxley Act (2002) (SOX) leisure 53, 56, 72–3 ‘lemons’ problem see adverse selection ‘liquidity crises’ 127 living standards 148

loan and savings crisis (US) 128–31 Lucas, Robert E. Jr. 11, 80, 84, 85, 107, 108 see also Lucas critique Lucas critique 80–85, 107, 108 Lummer, Scott 126, 134, 141 manager–labor coalitions 149–50 managerial expertise channel 146, 147 managers incentives, for using asymmetric information 138, 140–41, 142 skills, and ﬁnancial development 146 and social democracies 149–50 marginal analysis, in maximizing-ﬁrms in model economy 27 marginal beneﬁts in classical economy’s demand side 59, 60 commodity money 38 and eﬃciency wage theories 72–3 ﬁat money 38 marginal costs in classical economy’s demand side 59, 60 in classical economy’s supply side 54 commodity money 38 and eﬃciency wage theories 72–3 ﬁat money 38 and maximizing-households in model economy 26 marginal productivity 54, 62–3 marginal utility, and maximizinghouseholds in model economy 26 market discipline 132–3, 137–8 market incentives 137, 139 market opportunities 146 maximizing-behavior axiom ‘as if’ principle 16–17, 18–19, 69 eﬃcient market hypothesis 68–9 in models 9 and Phillips Curve 68, 69–70 predictions 16, 17–21 self-evidence 11, 17 and time inconsistency problem 87, 88–93 see also maximizing-households; maximizing-managers; proﬁtmaximization

Index maximizing-ﬁrms see proﬁtmaximization maximizing-households in classical economy’s demand side 59–60 in classical economy’s supply side 56 and demand curve in model economy 25–7, 30 and eﬃciency wage theories 72–3 and money 30 maximizing-managers 138, 140–41 McConnell, John 126, 134, 141 McFadden, Daniel 149 meaningfulness, of models 11 Meltzer, Allan H. 70 mimicking, problem of 105 model economy cash constraints 30–31, 115 governing money and capital markets 28–31 maximizing-ﬁrms and supply curve 27–8, 29 maximizing-households and demand curve 25–7 project constraints 30, 115 modeling 9–14 see also model economy Mody, Ashoka 123–4 monetary economy 43–4, 46–7, 48, 49, 50 monetary policies 70, 76 see also activist monetary policy; inﬂation–unemployment tradeoﬀ policy; monetary policy shocks; money supply changes; optimal monetary policies; passive monetary policy; stabilization monetary policy; tax policy monetary policy shocks 77, 78 monetary services institutions and commitment 102–6 measurement 99–101 see also accounting service ﬁrms; auditing; banks; consultation; credit services; debt ﬁnance; equity ﬁnance; evaluation services; ﬁnancial intermediaries; monitoring services

181

monetary versus ﬁscal authorities 106–7 money conﬁdence in monetary system 49 in a dynamic economy see classical economy, money in; Keynesian economy, money in; monetary policies in model economy 28–31 monetary economy 43–4, 46–7, 48, 50 ‘moneyness’ 35–7 quantity theory 58–9 and real activity in classical economy 62–3, 64 in static economy see static economy, money in turnover 58–9 money stock in classical economy’s demand side 59, 61–2 in Keynesian economy’s supply side 67–8 simple sum measures of monetary services 99–100 and velocity of money 58–9 money supply changes in classical economy’s demand side 61–2 commodity versus ﬁat money 39, 40 costs 49 and eﬃciency wage theories 75 in Lucas critique and Phillips Curve breakdown 83 and output 77 money supply shocks 77, 78 monitoring services 139, 140, 141, 142 moral hazard and debt ﬁnance 140–41 and deposit insurance 127, 129, 131 described 118–20 and dividends tax reductions 142–3 and equity ﬁnance 140 and ﬁnancial intermediaries 118–20 and social democracies 150 moral hazard mitigation in ﬁnancial intermediaries 123–4, 127, 140 multiplicity, problem of 104 Muth, John 80

182

Index

negative monetary policy shocks 77, 78 negative supply shocks 63 negative technology shocks 62–3, 64, 66 net worth 123, 129, 130, 131 ‘No shirking constraint’ curve 73, 74 nominal prices, and time inconsistency problem 90 nominal rigidities deﬁned 51 and eﬃciency wage theories 71–5 evidence against 76, 77–8 in Keynesian economy’s supply side 67 nominal wages and eﬃciency wage theories 71–2, 73, 74, 75 and money in a classical economy 53, 54 non-audit services 137, 138 non-equity stakeholders 141, 143 non-human evidence, in maximizingbehavior axiom 18–19 normative analysis versus positive analysis 21–2 omitted variable bias 12 opportunism 88, 111, 143 see also free riders; moral hazard opportunity costs 21, 27 opportunity set 44, 46 optimal contracts, monetary services 102–3 optimal monetary policies 87, 89, 90, 91 Orlando, Michael 138 output in classical economy’s demand side 59, 60 in classical economy’s supply side 57, 58 and ﬁnancial development 146, 147 and Lucas critique and Phillips Curve breakdown 81, 82, 83 and money stock in classical economy 61–2 and negative technology shocks in classical economy 62–3, 64

and Phillips Curve breakdown 81 and time inconsistency problem 90, 91 and velocity of money 58–9 output prices 28, 29, 67 ownership concentration 139, 150 separation from control 139–40, 146, 147, 150 passive monetary policy 63, 65 Pension Beneﬁt Guaranty Corporation (PBGC) 131 pensions crisis, emerging (US) 131 permanent output, Lucas critique and Phillips Curve breakdown 82, 83 Phillips Curve described 67–8, 80 expectations-augmented Phillips Curve 69–70, 80–81, 90–91 and Lucas critique 80–85, 107, 108 and monetary policy 70, 80 and rational expectations 68, 69–70, 80–81 pieces of paper (POP), in formal model of superior monetary equilibrium 46, 47, 49, 50 Poirier, Dale 9 politics and Federal Reserve System 110–11 and ﬁnancial development 148–50 and regulatory ‘mistakes’, repeated 131 and rule enforcement of monetary services 103 and time inconsistency problem 111 see also employment policies; monetary policies POP (pieces of paper), in formal model of superior monetary equilibrium 46, 47, 49, 50 positive analysis versus normative analysis 21–2 POWs (prisoners of war), economy of 46, 50 predictions, from models 10–13, 16, 17–21 Prescott, Edward C. 87, 103, 145 price changes 81–2, 83, 90, 91 price indexes 53, 58–9

Index price information 68–9, 83 price stickiness 66, 67, 76 prices in classical economy’s demand side 59, 60, 61, 62 in classical economy’s supply side 53, 54–5, 56, 57 ﬁnancial intermediation and transparency of ﬁrms 137 in Keynesian economy’s supply side 68, 69 and Lucas critique and Phillips Curve breakdown 82, 83 and maximizing-ﬁrms in model economy 28, 29, 30 and maximizing-households in model economy 26–7 and money stock in classical economy 61–2 see also asset prices; consumer prices; input prices; labor prices; output prices; price changes; price indexes; price information; price stickiness; relative prices; strike prices principal-agent problem 131–3, 140 see also separation prisoners of war (POWs), economy of 46, 50 private information 124, 126 production economy, barter versus money 47–8 production function, in classical economy’s supply side 53, 54 proﬁt-maximization and cash constraints 30, 115, 116 in classical economy’s supply side 53–7, 62–3 and eﬃciency wage theories 71–2, 74 and ﬁnancial development 146, 147 as incentive for disclosure by ﬁnancial intermediary demander 137 and project constraints 115 supply curve 27–8, 30 proﬁts 53, 54, 75 see also iso-proﬁt curve, maximizingﬁrms in model economy; proﬁtmaximization

183

prohibitions, in maximizing-behavior axiom 21 project constraints 115, 146 ‘project rich’ ﬁrms 30, 115, 146 property rights 149 see also intellectual property protection; ownership public regulation see regulation, public ‘put options’ 128–31 quantity theory of money 58–9 Radford, R.A. 46, 50 Rajan, Raghuram 132–3, 134, 136, 138, 147, 149, 150 random inﬂation, and Lucas critique and Phillips Curve breakdown 81 rational expectations and Lucas critique and Phillips Curve breakdown 82, 83 and Phillips Curve 68, 69–70 and time inconsistency problem 87, 89–90, 91, 92–3, 106–7 rationality see maximizing-behavior axiom real activity and central bank independence (CBI) 107–8 deﬁned 113 and deposit insurance 127 and Lucas critique and Phillips Curve breakdown 81–2, 83 and money in classical economy 62–3, 64 and money supply 77 and Phillips Curve breakdown 68–70 and time inconsistency problem 90 real business cycles (RBCs) 62–3, 64, 65, 66 see also cyclical output; cyclical unemployment real cash balances, in classical economy’s demand side 59, 60 real wages in classical economy’s supply side 55–8, 60, 62–3, 64 and eﬃciency wage theories 71–2, 73, 74 in Keynesian economy’s supply side 67, 68, 69, 70

184

Index

and money supply shocks 77, 78 negative monetary policy shocks, eﬀect on 77, 78 reduced form evidence 11–12, 13, 14, 77–8 regulation, public 132–3, 137–8 see also laws, monetary services; regulatory ‘mistakes’, repetition of; rules, monetary services regulatory ‘mistakes’, repetition of 127–33 relationship banking 125, 126 relative price changes, Lucas critique and Phillips Curve breakdown 82, 83 relative prices 38, 82, 83 repeated interactions of monetary authorities 103–5, 141 repetition, regulatory ‘mistakes’ 127–33 reputation 104 resource costs 38, 39 restrictive covenants, and moral hazard mitigation 123 rigidities 51 see also nominal rigidities risk and deposit insurance 127, 128, 129, 130 and moral hazard 118–20, 123, 127 risk aversion, and stabilization policy 84–5 Roe, Mark 149–50 Rogoﬀ, Kenneth 106, 107 rules, monetary services 102–3 see also laws, monetary services; regulation, public sacriﬁces, and maximizing-behavior axiom 21 Sarbanes-Oxley Act (2002) (SOX) 133, 137, 138 Satterthwaite, Mark 105–6 savings and loan crisis (US) 128–31 Schelling, Thomas 105–6 Schleifer, Andrei 69, 140, 141–2 Schwartz, Anna Jacobson 77 securities markets 137–8 see also capital structure seignorage see money supply changes self-evident assumptions 10, 11, 17

separation auditing and consultation 133 commercial and investment services 132, 133, 138 ownership and control 139–40, 146, 147, 150 see also conﬂicts of interest shirking, and eﬃciency wage theories 72–4, 79 shocks 79 see also aggregate demand shocks; ‘liquidity crises’; monetary policy shocks; negative supply shocks; pensions crisis, emerging (US); savings and loan crisis (US); technology shocks simple sum measures, monetary services 99–100 skilled managers 146, 147 social democracy 149–50 SOX (Sarbanes-Oxley Act 2002) 133, 137, 138 specialization, in barter versus money production economy 48 specialized managers 146, 147 ‘speciﬁc’ investments 141–3 stabilization monetary policy 84–5 stagﬂation 70 stakeholder investments 141, 142–3 stakeholders, non-equity 141, 142–3 static economy, money in evolution of exchange media 37–9, 40 inferior barter equilibrium 42–5, 46, 47, 50 money as exchange facilitator 35–7 superior monetary equilibrium 46–7, 50 stickiness see price stickiness; wage stickiness Stock, James 76, 77 stock option payments 138, 140 store of value 40 Strahan, Philip 148 strike prices 128–9 structural evidence, in modeling 13, 14 sub-optimal monetary policies, and time inconsistency problem 87, 89, 90, 91, 92–3

Index Summers, Lawrence 107–8, 141–2 superior monetary equilibrium 46–7 supply in barter versus money endowment economy 44 in barter versus money production economy 48 in classical economy 53–7, 58, 60, 61, 62–3, 64 in a Keynesian economy 67–70 and maximizing-ﬁrms in model economy 27–8, 29 see also aggregate supply; labor supply; money supply changes; money supply shocks; negative supply shocks; Phillips Curve surprise inﬂation, and unemployment in Phillips Curve 69, 70 systematic inﬂation, in Lucas critique and Phillips Curve breakdown 81 systematic risk 127–31 take-overs, hostile 141–2 tax policy 142–3 see also ﬁscal versus monetary authorities technology in barter versus money endowment economy 48 in classical economy’s supply side 53, 60 and ﬁnancial development 146, 147 technology constraints 27, 30 technology shocks 62–3, 64, 66 time inconsistency problem and active monetary policy, reasons for 87–8, 89–93 and central bank independence (CBI) 107 described 87, 88–9 and Federal Reserve System 111 and inﬂation 90, 91, 92–3, 111 and repeated interaction of monetary authorities 103–4 and rule enforcement of monetary services 103 transaction costs 30, 43–4, 45, 46, 48 transparency 136, 137, 141 transportation costs 38

185

unemployment and inﬂation trade-oﬀ policy 70, 80–82 involuntary, and eﬃciency wage theories 72–4, 75 and Lucas critique and Phillips Curve breakdown 81, 90 and Phillips Curve 68, 69, 70, 80 and time inconsistency problem 90, 91 unexpected inﬂation, and Phillips Curve breakdown 81, 90 unexpected price changes, in Lucas critique and Phillips Curve breakdown 81–2 United States bank restriction reductions and economic performance 146–7 commitment problems, commodity versus ﬁat money 39, 40 corporate governance 137–8 deposit insurance 126–7, 128–31 money supply and output 77 repeated regulatory ‘mistakes’ 128–33 see also Congress; Federal Open Market Committee (FOMC); Federal Reserve Board (FRB); Federal Reserve System universal banks 132–3, 138 user costs 100, 101 validity, of models 10–11, 13 value 40, 128–9 velocity of money 58–9 veto players 111 Vickers, John 105, 106 Vishny, Robert 140 Vives, Xavier 123, 140 wage agreements 67, 78 wage stickiness 67, 68, 76 wages see eﬃciency wage theories; nominal wages; real wages; wage agreements; wage stickiness Watson, Mark 76, 77 weighted aggregates, in monetary service measurement 100–101

186

Index

welfare barter versus money in production economy 47–8, 49 ﬁnancial development and living standards 148 ﬁnancial development and politics 148–50 and Glass-Steagall Act (1933) 131–3, 138 inferior barter economy 42–5, 50 stabilization policy 84–5 superior monetary economy 46–7 and time inconsistency problem 90, 91, 92

welfare maximization, and time inconsistency problem 90, 91, 92, 93 well-being see welfare well-developed ﬁnancial systems 30–31, 137, 147–8 Wier, Peggy 125 Willett, Thomas D. 110 Yap 35–7 Yellen, Janet 75, 79 Zingales, Luigi 136, 147, 149, 150