Exploitation 9780691214511

What is the basis for arguing that a volunteer army exploits citizens who lack civilian career opportunities? How do we

118 82 34MB

English Pages 336 [325] Year 2020

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Exploitation
 9780691214511

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Exploitation

Exploitation Alan Wertheimer

PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY

Copyright © 1996 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, Chichester, West Sussex All Rights Reserved Second printing, and first paperback printing, 1999 Paperback ISBN 0-691-01947-9 The Library of Congress has cataloged the cloth edition of this book as follows Wertheimer, Alan Exploitation / Alan Wertheimer p. cm Includes index. ISBN 0-691-02742-0 (CL. alk. paper) 1. Exploitation. I. Title. BJ1474.5.W47 1996 170—dc20 95-52898

This book has been composed in Palatino

The paper used in this publication meets the minimum requirements of ANSI/NISO Z39.48-1992 (R1997) (Permanence of Paper) http: / /pup.princeton.edu Printed in the United States of America

2 4 6 8

10

9 7 5 3

For My Family, Susan, Rachel, Rebecca, and Joel

CONTENTS

ix

Preface

xiii

Acknowledgments CHAPTER ONE

Overview PART ONE:

Contexts

CHAPTER TWO

Unconscionable Contracts

37

CHAPTER THREE

The Exploitation of Student Athletes

77

CHAPTER FOUR

Commercial Surrogacy

96

CHAPTER FIVE

Unconstitutional Conditions

123

CHAPTER SIX

Sexual Exploitation in Psychotherapy PART TWO:

158

A Theory of Exploitation

CHAPTER SEVEN

Unfair Transactions

207

CHAPTER EIGHT

Consent

247

CHAPTER NINE

Moral Weight and Moral Force

278

Index

311

PREFACE

THIS BOOK has

its origins in ordinary moral and political discourse. We say that some transactions and relationships are exploitative whereas others are not. "But don't you think that 'exploitation' is so elastic that it really has no meaning?" So I was asked at a law school faculty seminar. I answered that people think they mean something when they employ "exploitation" and its cognates. The purpose of this book is to see whether such thoughts must be illusory. I believe that they need not be. The one-word title EXPLOITATION is somewhat misleading and presumptuous (and misleading because presumptuous). It is misleading if it implies that I am concerned with advancing a set of necessary and sufficient conditions for the proper use of the word "exploitation." The book does not undertake that kind of conceptual analysis. Rather, I seek to explore a certain class of phenomena. I seek to lay bare the morally relevant features of transactions and relationships in which it is claimed that one party is taking unfair advantage of another. The one-word title is presumptuous because my analysis may be both over- and underinclusive. It may be thought that I cast my net too widely because this project focuses on what I shall call mutually advantageous and consensual exploitation, on cases where it can at least be plausibly maintained that the exploitee gains from and has voluntarily consented to the transaction or relationship. With one major exception (sexual exploitation in psychotherapy), I have little to say about the forms of exploitation in which the exploiter gains by harming the exploitee or to which the exploitee does not (arguably) consent. Although mutually advantageous and consensual exploitation is (typically) less bad than harmful and nonconsensual exploitation, it raises more interesting theoretical questions— or so I shall argue. Still, it might be objected that even to consider mutually advantageous and consensual transactions as exploitative—much less focus upon them—is to fundamentally misconceive the nature of exploitation, that there can be no wrongful exploitation if both parties gain from and consent to a transaction. I do not want to quarrel over words or labels. If someone wishes to insist that exploitation must be harmful to the exploitee, then I shall say that this book is not about exploitation, but mexploitation or shmexploitation or whatever. I am inter-

X

PREFACE

ested in the moral character of certain sorts of transactions and relationships, whatever we want to call them. I believe that we call them exploitative. But it may also be thought that I have not cast my net widely enough. While I focus on cases in which it might be said that one party takes unfair advantage of another, it might be said that one party exploits another whenever one uses the exploitee as a means to one's own ends, even when one does not use the exploitee unfairly. I shall have a bit more to say about such "usings" along the way. But, once again, I do not want to worry about words. If it makes sense to describe usings that are not unjust as cases of exploitation, then I shall say that I am interested in the subset of exploitative transactions that are unjust. There is another sense in which I have restricted the analysis: I have chosen to say little directly about the Marxist view of exploitation. Indeed, with the exception of the discussion of student athletes in Chapter 3 and procreational labor in Chapter 4, I have little to say about the exploitation of labor per se. Given that Marxism is the only major tradition in political theory in which the concept of exploitation figures prominently, a few words are in order about what might appear to be a major lacuna. I have several reasons for taking a different tack. First, there is an extensive and sophisticated Marxist literature on exploitation, and I do not believe I can add much to it. Second, I believe that the important moral core of the Marxist view of exploitation is not unique to Marxism. When Marxism claims that the capitalist class exploits the proletariat, it employs the ordinary notion that one party exploits another when it gets unfair and undeserved benefits from its transactions or relationships with others. About that I hope to have something to say. On the other hand, I believe that what is unique to Marxism—its approach to measuring exploitation through calculations of surplus value—is very problematic. In any case, we know that Marxism takes exploitation seriously. It is, I believe, more important to show that one can take exploitation seriously from within the framework of a liberal egalitarian political theory, the tradition within which this work is firmly anchored. So much for scope. A few words about the conceptual apparatus that I sometimes employ. Although I have avoided extensive contact with Marxist economics, I have found it useful to employ some terminology from mainstream microeconomics. I know that this language—social surplus, reservation prices, zones of agreements, payoff matrices, etc.—can be off-putting and may appear to be reductionist. Because this language does not capture the psychodynamics and sociology of relationships, the analysis may seem insufficiently

PREFACE

xi

deep. That is a risk I shall run. I think that this language focuses attention on certain crucial elements in transactions and relationships and, when kept in proper perspective, illuminates rather than obscures. Do I think that marriage creates a social surplus? Yes. Do I think it is important to understand how that surplus is divided? Yes. Do I think that all there is to marriage can be perspicuously put in these terms? I trust that my wife thinks I do not. A point about the structure of the book. I have chosen to start reasonably close to the ground. In Part I, I work through several different contexts in which it is claimed that a party is engaged in exploitation. In Part II, I focus more directly on the philosophical issues that an account of exploitation must address. Although there is a clear distinction in tone and emphasis between the two parts, the division is somewhat artificial. Because much of the philosophical analysis is developed in Part I, I urge the more philosophically minded readers not to cheat by skipping ahead to Part II. If I am right, such readers will find Part I to be philosophically engaging. If I am wrong, they may find the analyses in Part I to be independently interesting in their own right. A final point. Although I think that this is a reasonably ambitious project, it may not provide the answers that some seek. If I am successful, I will have identified the crucial elements in exploitative transactions and relationships and will also have indicated the sorts of arguments one needs to make if we are to justifiably prohibit such exploitation. Still, I will not answer the questions of which transactions or relationships are exploitative and which exploitative transactions are justifiably prohibited. To answer those questions will often require empirical evidence and philosophical analyses that are beyond the scope of this project. If I can make good on my objectives, I believe that I will have done enough. Whether I make good on those objectives and whether it is, in fact, enough, is for the reader to decide.

ACKNOWLEDGMENTS

TRY AS I MIGHT,

it is simply too hard to resist some play on the subject of this book here. Although I am quite confident that the help I have received from numerous friends and colleagues has always been provided with their "free, voluntary, and undeceived consent," it does not follow, on the account herein, that I have not taken unfair advantage of their good graces. I hope that I have not. Because this project has taken (too) many years, I probably do not remember all those who have provided helpful comments on parts of the manuscript. To those whom I may have overlooked, I apologize in advance. But I do remember many. My departmental colleagues Pat Neal and Bob Pepperman Taylor commented on virtually every chapter. My colleagues in the Department of Philosophy at the University of Vermont have been both encouraging and constructively critical with their nonphilosopher colleague. I thank David Christensen and Derk Pereboom for their comments on several chapters. Arthur Kuflik and Don Loeb read even more chapters with truly extraordinary care and acuity while complying with my request that they try to identify problems that would not take me too long to fix. They found all too many. Arthur Kuflik also suggested the Vermeer painting that appears on the dust jacket. I began this project during the 1989-1990 academic year while on sabbatical leave from the University of Vermont and as a fellow in the Program in Ethics and the Professions at Harvard University. To both institutions I am grateful. I am particularly grateful to the director of the Program, Dennis Thompson, who wisely chose a wonderful group of fellows with whom I could share early versions of some of this work, and who also commented in detail on parts of the manuscript. Other members of the Program in Ethics and the Professions have made many valuable suggestions. They include Greg Dees, Deborah Hellman, and Ken Winston. While continuing to work on the book, I returned to Harvard as a visiting professor at the Kennedy School of Government. My KSG colleague Fred Schauer commented upon the entire manuscript, not only (as he feared readers might think from my footnotes) on Chapter Three. It is difficult to say just how much I owe to another KSG colleague and friend, Arthur Applbaum. It is not just that he commented in great detail on much of the manuscript at various points in its development, although it is certainly that as well. I have learned from

Xiv

ACKNOWLEDGMENTS

him virtually everything I know about the use of economic concepts in the sort of analysis I advance here. And more. Ann Himmelberger Wald and the staff of Princeton University Press have been very supportive. The reader can judge whether Lyn Grossman was sufficiently aggressive in editing the manuscript. From the author's perspective, she was both extremely helpful and mercifully nonintrusive. Professional colleagues are fair game. Family is not. I am afraid that I did exploit my family during the period in which I worked on this book. The entire family (minus a college-age daughter) relocated for my sabbatical year. Later on, I became a weekend husband and parent for two academic years. My wife, Susan, unselfishly encouraged all this, and my children accepted both arrangements (more or less) without complaint. This is the first small installment on my debt.

Exploitation

Chapter One OVERVIEW

PEOPLE WRONG ONE ANOTHER in numerous ways. They kill, steal, as-

sault, injure, abuse, oppress, dominate, enslave, coerce, extort, seduce, manipulate, subjugate, lie, deceive, defraud, pry, cheat, betray, neglect, discriminate. And they exploit.1 While I have been working on this book, I have looked and listened for instances in which it was claimed that persons or institutions were engaged in exploitation. Consider: 1. It was said that the R. J. Reynolds Company was exploiting African Americans by planning to test-market a new cigarette (Uptown) in Philadelphia that it thought would have special appeal to African Americans.2 2. The president of Stanford University claimed that big-time college athletics "reeks of exploitation/' because the universities gain a great deal of revenue from the services of the athletes while the athletes (whose graduation rate is much lower than that of nonathletes) gain little from their college experience.3 3. When Iraq invaded Kuwait in 1990, leading to an increase in petroleum prices, New Hampshire governor Judd Gregg said that his administration would "not tolerate exploitation of the New Hampshire consumer^ as a result of the Middle East crisis.4 4. Advocates for people with AIDS protested the (then) annual price of AZT ($8,000), a drug that was thought to slow the progress of HIV, claiming that the Burroughs-Welcome Company, which produces AZT, was exploiting people who were already suffering.5 1

Some actions can, of course, be accurately described as wrongful under more than one rubric. 2 The marketing campaign was abandoned in response to heavy public criticism. See, for example, "A Cigarette Campaign under Fire," New York Times, January 12, 1990, p. Dl. "Critics, like the American Cancer Society, see the Uptown campaign as an alarming escalation in cigarette marketing. They believe the campaign exploits blacks, especially the ghetto poor." Ibid. 3 Donald Kennedy, "So What If College Players Turn Pro Early?" New York Times, January 28, 1990. 4 United Press International, August 5, 1990, Sunday, BC cycle. 5 See, for example, "Wave of Protests Developing on Profits from AIDS Drug," New York Times, September 16, 1989 p. 1. The company subsequently reduced the price of AZT.

CHAPTER ONE

5. A New York Times article described the proliferation of "posh strip clubs/' in which topless young women dance for the customers. It was both claimed and denied that these clubs were loci of exploitation. The article noted that many believe that such bars "exploit women/' At the same time, "many dancers say their work is no more exploitative than most other forms of employment," and the owner of one bar remarked, "If anyone is being exploited it is the men, the guys buying into the fantasy of she really likes me."6 6. USA Today featured an article advocating the legalization of organ sales, whereby a person could be paid cash for, say, a kidney. One reply maintained that such a policy would "open wide the door to exploitation."7 7. When Mary Beth Whitehead refused to surrender her daughter (who was known as "Baby M"), as specified in her $10,000 surrogacy contract with William Stern, the case touched off extended public, legal, and philosophical discussion of surrogate motherhood. One commentator observed that "one of the most serious charges against surrogate motherhood contracts is that they exploit women."8 8. In urging President Clinton to oppose the North American Free Trade Act (NAFTA), Lane Kirkland, then president of the A.F.L.-C.I.O., said that NAFTA was based "solely on the exploitation of Mexican workers."9 9. It is frequently said that psychotherapists who have sexual relations with their patients are engaged in exploitation. 10. It is sometimes claimed that the volunteer army exploits those citizens who lack decent civilian career opportunities: "A society as unjust as ours must draft its military to avoid unfair exploitation. . . ." 10 11. A colleague, who was raised in West Virginia and remains intensely 6

New York Times, April 15, 1992, p. C12. "They [the clubs] are demeaning and degrading to women and treat them like sex objects/' said Caia M. Mockaltitis, a spokeswoman for Concerned Women of America. Ibid. Some dancers observed that by dancing in a G-string they could earn in one night's tips as much as they could in a month in many other jobs. "I can make $1,000 a night or more if I really work at it," remarked one young woman who had deferred her plans to attend New York University Law School. Ibid. 7 USA Today, September 24, 1991. 8 Martha Field, Surrogate Motherhood (Cambridge: Harvard University Press, 1989), p. 25. As another article put it, "The prohibition on payments may be understood as protecting . . . women—especially poor, single women—from being exploited . . . paid 'breeding stock/" Alexander Capron and Margaret J. Radin, "Choosing Family Law over Contract Law as a Paradigm for Surrogate Motherhood/7 in Surrogate Motherhood, ed. Larry Gostin (Bloomington: Indiana University Press, 1990), p. 62. 9 New York Times, February 18, 1993, p. B10. 10 David Wasserman, "Enlistment and Exploitation," Report from the Institute for Philosophy and Public Polio/, Spring 1991, p. 7.

OVERVIEW

5

loyal to the place of his roots, bemoaned the fact that some West Virginians would cruise the highways during snowstorms and exploit stranded motorists by offering to assist them for a high fee. 11

Although we frequently claim that some act, practice, or transaction is exploitative, the concept of exploitation is typically invoked without much analysis or argument. I do not mean that such claims are generally false, although I believe that some such claims are false. I do mean that the concept of exploitation is often utilized in such claims as if its meaning and moral force were self-evident. They are not. If some or even all of the sorts of claims I have noted are true, when and why are they true? And if they are true, what follows? Those are the questions that motivate this project. More precisely, the purpose of this book is to answer two sets of questions: (1) what are the truth conditions of an exploitation claim, and (2) what is the moral weight and moral force of an exploitation claim? Let me explain. For present purposes, "an exploitation claim" refers to statements that A's interaction with B or some more general phenomenon is (or is not) wrongfully exploitative or to statements that presuppose an account of wrongful exploitation. I say "wrongfully exploitative" because "exploitation" can be employed in a nonmoral sense as well.12 To say that colleges exploit student athletes is to make an exploitation claim, as it is to assert that the Burroughs-Welcome Company was exploiting AIDS patients, or to assert that strip clubs do not exploit women. Susan Moller Okin makes an exploitation claim when she says that our family system constitutes "the pivot of a societal system of gender that renders women vulnerable to dependency, exploitation, and abuse/' for we must know what exploitation involves to determine whether this claim is valid.13 Elizabeth Anderson makes an exploitation claim when she says that the exploitation of surrogate mothers could not be avoided by "properly screening surrogates [and] setting low fixed fees," for we must know what constitutes exploitation to assess the argument that certain reforms would not remove it.14 Assuming that exploitation claims are meant to have some definable content, that they are not merely a rhetorical placeholder for expressing disapproval, the first task of a theory of exploitation is to 11

Conversation with Professor Pat Neal, October 1993. See, for example, Richard Arneson's entry on exploitation in Encyclopedia of Ethics, ed. Lawrence C. Becker (New York: Garland, 1992), p. 350. In the nonpejorative sense, "to exploit" is simply "to make use of." Thus, the first (nonobsolete) definition of "exploit" in Webster's Unabridged is "to turn (a natural resource) to economic account." 13 Susan Moller Okin, Justice, Gender and the Family (New York: Basic Books, 1989), pp. 135-36. 14 "Is Women's Labor a Commodity?" 19 Philosophy & Public Affairs 71 (1990), at 88. 12

6

CHAPTER ONE

provide criteria for a valid exploitation claim. On the view that I shall defend, exploitation is a moralized concept. At least one criterion for a valid exploitation claim will turn out to be a moral criterion: a transaction is exploitative only if it is unfair. Because a valid exploitation claim presupposes that the transaction is—at least in that way—already wrongful, it might be thought that all the moral work is done once it is established that someone is wrongfully exploiting another.15 That thought would be in error. The general point is this. Although the sort of exploitation in which we are interested is, by definition, wrong because unfair, the (moral) "fact" of exploitation settles less than meets the eye. To see this, I shall distinguish between the moral weight of exploitation, that is, its degree of wrongness, and the moral force of exploitation, that is, the various moral upshots that exploitation might involve for parties to the transaction or for society: Can people have a right to exploit? Is it wrong to allow oneself to be exploited? Should society prohibit actions if they are exploitative? Should we refuse to enforce exploitative agreements? And the wrongness of exploitation does not dictate the way in which these moral questions should be answered. Because the moral weight and moral force of exploitation are related, I shall have to address both issues. Still, I shall be especially concerned with considering the moral force of exploitation. These are the questions I want to answer. Unfortunately, contemporary work in political and moral philosophy provides less help than one might suppose. Despite the frequency and ease with which we make exploitation claims in ordinary moral and political discourse, I think it fair to say that with the major (and I do mean major) exception of the Marxist tradition, exploitation has not been a central concern for contemporary political and moral philosophy.16 John Rawls's A Theory of Justice has virtually nothing to say about exploitation, as 15

It might also be objected that if exploitation is moralized, then one cannot claim that a practice (such as commercial surrogacy) is wrongful because it is exploitative, because the very invocation of "exploitation" (or its cognates) presupposes that the practice is wrongful because it is unfair. See Allen W. Wood, "Exploitation, "12 Social Philosophy and Policy 136 (1995). I believe that this objection is mistaken, or at least overdone. After all, to claim that surrogacy is wrongful because it is exploitative would point to a particular reason—the unfairness of the transaction—for regarding surrogacy as wrong as contrasted with other reasons that might be advanced for thinking that surrogacy is wrong. 16 Because exploitation is often tied to the now almost universally rejected labor theory of value, even some Marxists argue that those who share their political sympathies should be less concerned with exploitation than has traditionally been their wont. See, for example, John Roemer, "Should Marxists Be Interested in Exploitation?," 14 Philosophy & Public Affairs 30 (1985).

OVERVIEW

7

17

such. Robert Nozick discusses the Marxian account of exploitation in Anarchy, State, and Utopia, but, not surprisingly, only to reject it as a basis for interfering with (most) transactions.18 Although Michael Walzer's Spheres of Justice attempts to advance a democratic socialist vision of justice, its index contains no reference to exploitation.19 Nor does exploitation appear in the indexes to Brian Barry's Theories of

Justice20 and Justice as Impartiality,21 Susan Moller Okin's Justice, Gender, and the Family,22 and Joseph Raz's The Morality of Freedom23, to

mention several other prominent works (and I assure the reader that this is not a particularly biased sample). Bruce Ackerman's Social Justice in the Liberal State contains a chapter entitled "Exploitation," but the text provides no account of what exploitation is.24 Although the American Society for Political and Legal Philosophy's annual NOMOS volumes (which now number thirty-eight) have covered many of the important concepts in political philosophy, there is no volume on exploitation.25 I do not want to overstate the point. There is both a non-Marxist and a Marxist literature on exploitation, and much of it will prove helpful to our project.26 Moreover, works that analyze practices that are said to be exploitative can do much to illuminate our topic, even when they do not engage in extensive analysis of exploitation itself. In the remaining parts of this introductory chapter, I do the following. I try to locate the territory of exploitation within our moral and political discourse. I briefly survey the definitional landscape that has been 17

Cambridge: Harvard University Press, 1971. is New York: Basic Books, 1974. 19 New York: Basic Books, 1983. 20 Berkeley: University of California Press, 1989. 21 Oxford: Clarendon Press, 1995. 22 Okin, Justice, Gender, and the Family. 23 Oxford: Clarendon Press, 1986. 24 (New Haven: Yale University Press, 1980). Ackerman says that "before exploitation exists, it is not necessary for one group to have an advantage over the other in each and every power domain." Note that this passage—which is indexed as a definition of exploitation—does not provide an account of what it would be for exploitation to exist. See pp. 242-43. 25 Indeed, there is no index reference to exploitation in J. Roland Pennock and John Chapman, eds., NOMOS XXXI: Markets and Justice (New York: New York University Press, 1989. 26 In particular, the fourth volume of Joel Feinberg's treatise on the moral limits of the criminal law, Harmless Wrongdoing (New York: Oxford University Press, 1988), contains an extensive discussion of exploitative relations and makes many helpful distinctions, although it is interesting that his analysis of exploitation occurs in the volume with that title. Robert Goodin's Reasons for Welfare (Princeton: Princeton University Press, 1988) also contains a helpful discussion of exploitation.

8

CHAPTER ONE

marked out and the conceptual quarrels that appear there. I indicate the kind of analysis that I hope to produce and the route that I intend to take. I sketch a very rough and preliminary account of the elements of exploitation. Finally, I make some brief remarks about the moral force of exploitation. THE LOCUS OF EXPLOITATION

As a conjecture about the sociology of knowledge, I believe that the relative lack of attention to exploitation may be tied to several features about the location of exploitation in our moral and political discourse. First, precisely because the concept of exploitation has figured so prominently in Marxist political theory, non-Marxists may have been reluctant to enter the field, lest they be thought guilty by association with a theory that they want to reject. Second, and more important, I believe that most political philosophers prefer to address questions of "ideal" political theory, to define the characteristics of a good society such as justice, equality, liberty, and community. By contrast, exploitation is a wrong. There may be less exploitation in a just society than in an unjust society, but we do not characterize justice as the (mere) avoidance of exploitation. Third, exploitation is often a micro-level wrong to discrete individuals in distinct relationships and transactions. By contrast, much of the best contemporary political philosophy tends to focus on macrolevel questions, such as the just distribution of resources and basic liberties and rights. Indeed, one standard picture of a just liberal democratic society maintains that a just society would provide a set of background institutions and resources within which individuals could act without regard to strong principles of justice for individual transactions, such as prices, wages, agreements—in other words, the stuff of the examples with which we began. As Thomas Nagel puts this view, "It clearly is a desirable feature of a social order that within it, people should not be too constrained in the pursuit of their own lives by constant demands for impartial attention to the welfare of others. . . ." 27 This view does not, of course, maintain that we are 27

Nagel goes on to say: "Ideally the moral division of labor would assign the bulk of agent-neutral values to be realized by background institutions, leaving us relatively free to pursue agent-relative values in our personal lives." Equality and Partiality (New York: Oxford University Press, 1991), p. 95. As John Rawls puts it, a just society should provide background institutions that " . . . insure that the system of cooperation is one of pure procedural justice. . . . It is a mistake to focus attention on the varying relative positions of individuals and to require that every change, considered as a single transaction viewed in isolation, be itself just." A Theory of Justice, pp. 87-88.

OVERVIEW

9

entitled to act unjustly towards others. But it does suggest that we need not worry much about individual transactions or relations if the background institutions provide all persons with a just package of primary goods, opportunities, and resources. Now it is entirely plausible that the best division of philosophical labor would allocate more attention to the best principles for a good society rather than to forms of wrongness and that it would also devote more attention to macro-level questions than to micro-level ones. Still, as Judith Shklar has written, political philosophy should also attend to forms of wrongdoing as independent phenomena in their own right: "Common sense and history surely tell us that these are primary experiences and have an immediate claim on our attention/' 28 And the fact is that while the background conditions shape our existence, the primary experiences occur at the micro level. Exploitation matters to people. People who can accept an unjust set of aggregate resources with considerable equanimity will recoil when they feel exploited in an individual or local transaction.29 It may be objected that while exploitation may manifest itself in micro-level transactions, it has its roots in macro-level injustice, that there would be much less micro-level exploitation in a society that is just at the macro level. Even if this were true, as it probably is, it is important to look closely at the micro level itself. If we are concerned about macro-level injustice in part because it gives rise to micro-level exploitation, we need to know what constitutes the micro-level exploitation so that we can determine what must be done at the macro level. Furthermore, micro-level exploitation is not as closely linked to macro-level injustice as might be thought. Even in a reasonably just society, people will find themselves in situations in which they can strike an agreement that will produce mutual gain, and some of those cases will give rise to allegations of exploitation.30 Moreover, even when micro-level exploitation is rooted in background conditions that are clearly unjust by the principles of ideal theory, it may be thought that the best principle of nonideal morality is to respect the wishes of the parties when an agreement is mutually advantageous, exploit28

"Giving Injustice Its D u e , " 98 Yale Law Journal 1125 (1989), at 1135. As we shall see in more detail in Chapter 7, people will sometimes refuse to allow themselves to be exploited even when it is in their (self-) interest to do so, a form of behavior that poses an important challenge to the standard microeconomic model of human behavior. See, for example, Daniel Kahneman, Jack L. Knetsch, and Richard H. Thaler, "Fairness and the Assumptions of Economics/' 59 Journal of Business S285-S300 (1986). 30 See Brian Barry, Theories of Justice (Berkeley: University of California Press, 1989), p. 305. 29

10

CHAPTER ONE

ative though it may be. For all these reasons, then, the truth conditions and moral force of exploitation demand our attention.31

THE DEFINITIONAL LANDSCAPE We can give a broad—lowest common denominator—definition of exploitation with which virtually everyone will agree.32 At the most general level, A exploits B when A takes unfair advantage of B. (I shall always refer to the alleged exploiter as "A" and to the alleged exploitee as "B"). One problem with such a broad account, as Arneson notes, is that there will "be as many competing conceptions of exploitation as theories of what persons owe to each other by way of fair treatment." 33 The aim of this book is to do better. We can gain a somewhat sharper view of the issues that we must confront if we consider a sampling of the accounts that appear in the literature. 1. "[T]o exploit a person involves the harmful, merely instrumental utilization of him or his capacities, for one's own advantage or for the sake of one's own ends/' 3 4 2. "It is the fact that the [capitalist's] income is derived through forced, unpaid, surplus [wage] labor, the product of which the workers do not control, which makes [wage labor] exploitive."35 3. "Exploitation necessarily involves benefits or gains of some kind to someone. . . . Exploitation resembles a zero-sum game, viz. what the exploiter gains, the exploitee loses; or, minimally, for the exploiter to gain, the exploitee must lose."36 4. "Exploitation [in exchange] demands . . . that there is no reasonably eligible alternative [for the exploitee] and that the consideration or advantage received is incommensurate with the price paid. One is not 31

For the distinction between ideal theory and nonideal theory, see Rawls, A Theory of Justice, pp. 244-48. 32 Not quite everyone. Allen Wood h a s argued that exploitation need n o t involve unfairness, that A exploits B w h e n A uses something about B for A's o w n ends. See Wood, "Exploitation, "p. 147. It is not clear w h e t h e r Wood's account is as nonmoralized (or fairness-free) as h e thinks, for he goes o n to say that A exploits B w h e n A uses B for A's o w n e n d s "by playing o n some weakness or vulnerability in that p e r s o n . " This implies that A does not exploit B if A uses B for A's o w n e n d s w h e n A does not play on a weakness or vulnerability, that is, w h e n the transaction is fair. 33 Encyclopedia of Ethics, p. 350. 34 Allen Buchanan, Ethics, Efficiency, and the Market (Totowa: N.J.: R o w m a n a n d Allanheld, 1985), p. 87. 35 Nancy Holmstrom, "Exploitation," 7 Canadian journal of Philosophy 353 (1977), at 357. 36 Judith Farr Tormey, "Exploitation, Oppression and Self-Sacrifice," 5 Philosophical Forum 206 (1974), at 207-08.

OVERVIEW

11

exploited if one is offered what one desperately needs at a fair and reasonable price/' 3 7 5. "Exploitation of persons consists in . . . wrongful behavior [that vio-

6.

7.

8. 9.

10. 11. 12.

lates] the moral norm of protecting the vulnerable/'38 "There are four conditions, all of which must be present if dependencies are to be exploitable. First, the relationship must be asymmetrical. . . . Second, . . . the subordinate party must need the resource that the superordinate supplies. . . . Third, . . . the subordinate party must depend upon some particular superordinate for the supply of needed resources. . . . Fourth, the superordinate . . . enjoys discretionary control over the resources that the subordinate needs from him. . . ."39 "Common to all exploitation of one person (B) by another (A) . . . is that A makes a profit or gain by turning some characteristic of B to his own advantage . . . exploitation . . . can occur in morally unsavory forms without harming the exploitee's interests and . . . despite the exploitee's fully voluntary consent to the exploitative behavior. . . ."40 "Persons are exploited if (1) others secure a benefit by (2) using them as a tool or resource so as (3) to cause them serious harm."41 "A society is exploitative when its social structure is organized so that unpaid labor is systematically forced out of one class and put at the disposal of another. . . . On the force-inclusive definition of exploitation, any exploitative society is a form of slavery."42 "[A] group is exploited if it has some conditionally feasible alternative under which its members would be better off."43 "[EJxploitation is seen as the failure to pay labour its marginal product. . . ,"44 "An exploitative exchange is . . . an exchange in which the exploited party gets less than the exploiting party, who does better at the exploited party's expense. . . . [T]he exchange must result from social

37 Stanley Benn, A Theory of Freedom (Cambridge: C a m b r i d g e University Press, 1988), p. 138. 38 G o o d i n , Reasons for Welfare, p . 147. 39 Robert E. G o o d i n , " R e a s o n s for Welfare: Economic, Sociological, a n d Political— But Ultimately M o r a l , " in J. D o n a l d M o o n , e d . , Responsibility, Rights, and Welfare (Boulder, Colo.: Westview Press, 1988), p . 37. 40 Feinberg, Harmless Wrongdoing, p p . 176-79. 41 S t e p h e n R. M u n z e r , A Theory of Property (Cambridge: C a m b r i d g e University Press, 1990), p . 171. 42 Jeffrey Reiman, "Exploitation, Force, a n d t h e Moral A s s e s s m e n t of Capitalism: T h o u g h t s o n Roemer a n d C o h e n , " 16 Philosophy & Public Affairs 3 - 4 1 (1987), at 3 - 4 . 43 John E. Roemer, " A n Historical Materialist Alternative to Welfarism," in Jon Elster a n d A a n u n d H y l l a n d , e d s . , Foundations of Social Choice Theory (Cambridge: C a m b r i d g e University Press, 1986), p . 136.

44 John Brewer, "Exploitation in the New Marxism of Collective Action," 35 Sociological Review 84-96 (1987), at 86.

12

CHAPTER ONE

13.

14. 15.

16.

relations of unequal power . . . exploitation can be entered into voluntarily; and can even, in some sense, be advantageous to the exploited party/' 45 "[Capitalist] social relations . . . are exploitative, not only in the specific sense of extracting surplus labour, but in the more general sense of using someone as a means, utilizing her to her detriment as a way of promoting one's own good. . . Z'46 "Workers are exploited if they work longer hours than the number of labor hours employed in the goods they consume/' 47 "[EJxploitation forms part of an exchange of goods and services when 1) the goods and services exchanged are quite obviously not of equivalent value, and 2) one party to the exchange uses a substantial degree of coercion/'48 "[EJxploitation is a psychological, rather than a social or an economic, concept. For an offer to be exploitative, it must serve to create or to take advantage of some recognized psychological vulnerability which, in turn, disturbs the offeree's ability to reason effectively."49

Now all these accounts are compatible with the view that A wrongfully exploits B when A takes unfair advantage of B. But there are some important differences among them. Some accounts (10,14) are technical definitions of exploitation that are specific to a Marxist approach. Although none of the accounts denies that exploitation requires a gain to the exploiter, only some (3, 8) specifically mention that criterion. Some accounts invoke the Kantian notion that one wrongfully exploits when one treats another instrumentally, or merely as a means (1, 8,13). On some accounts, the exploited party must be harmed (1, 2, 3, 8, 9,), whereas other accounts allow that the exploited party may gain from the relationship (4, 7,11,12,15). On some accounts, the exploited party must be coerced (2, 4, 6, 9,15), whereas others require at least a defect in the quality of the consent (12,16), and another maintains that exploitation can be fully voluntary (7), in which case the author also maintains that the exploited party has no grievance. 50 45

A n d r e w Levine, Arguing for Socialism (London: Verso, 1988), pp. 66-67. Will Kymlicka, Liberalism, Community, and Culture (Oxford: Clarendon Press, 1989), p. 114. 47 Karl Marx: A Reader, e d . Jon Elster (Cambridge: Cambridge University Press, 1986), p. 121. 48 Barrington Moore, Reflections on the Causes of Human Misery (Boston: Beacon Press, 1973), p . 53. 49 John Lawrence Hill, "Exploitation," 79 Cornell Law Review 631 (1994), at 637. 50 Feinberg, Harmless Wrongdoing, p . 176. Allen Wood agrees with Feinberg that exploitation can be voluntary, but would probably dissent from Feinberg's claim that the exploitee would therefore have no complaint. See Wood, "Exploitation," p. 148. 46

OVERVIEW

13

D E F I N I T I O N OR THEORY?

Given these sorts of disagreements, how should we proceed? Because we have no basis—at this point—for assessing the issues on which there appears to be controversy, I believe that we should adopt a strategy that meets two criteria: (1) it should be theoretically suggestive; and (2) it should avoid begging crucial questions. To say that an account of exploitation should be theoretically suggestive is to say that it should highlight the key elements of exploitative relations and push us to analyze them in greater detail. We want an account of exploitation that cuts some theoretical ice. Robert Goodin has argued that "to exploit people is to wrong them, however much or little they may lose or you may gain from the act."51 This sort of account offers too little. Our task is to distinguish the wrongmaking characteristics of exploitation from other forms of wrong. And it may turn out that at least some forms of exploitation do depend on the relative gains to A and B. But we do not want to begin with an account of exploitation that begs crucial questions. As we have seen, some writers have argued that exploitation must always be harmful to the exploitee or that a transaction cannot be exploitative unless the exploitee is coerced, is defrauded, or cannot reason effectively. I see no reason to put such constraints on what counts as exploitation, at least at the outset. While some exploitative transactions are harmful to the exploitee, we have seen that we also appeal to the notion of exploitation in cases in which the exploitee seems to gain from the transaction, and I do not think we should rule such cases out of bounds. I would go further. I believe that exploitation would be of much less theoretical interest on a "no harm, no exploitation" rule. We do not need to be moral rocket scientists to know that it is wrong for A to gain from an action that unjustifiably harms or coerces B. Even a libertarian will grant that some harmful exploitation may be legitimately prohibited by the state, if only because it is harmful (or rights-violating) rather than because it is exploitative. By contrast, it is more difficult to explain when and why it might be wrong for A to gain from an action that benefits B and to which B voluntarily consents. And it is certainly more difficult to explain why society 51

Robert E. Goodin, "Exploiting a Situation and Exploiting a Person," in Andrew Reeve, ed., Modern Theories of Exploitation (Beverly Hills, Calif.: Sage Publications, 1987), p. 182. This passage is not fully indicative of Goodin's view. In another passage, he says that "if the putative exploiter fails to realize some perceived benefits from the act, then either it was not a successful act of exploitation or else it was not an act of exploitation at all." Ibid., p. 172.

14

CHAPTER ONE

might be justified in prohibiting such transactions or refusing to enforce some such agreements. For these reasons, I shall distinguish between harmful exploitation and mutually advantageous exploitation. In speaking of mutually advantageous exploitation, I do not imply that both parties are exploited. There may or may not be such cases, but, if so, they are not the ones I have in mind. Rather, in speaking of mutually advantageous exploitation I simply mean that the exploitee gains from the transaction as well as the exploiter. It is the advantageousness of the transaction that is mutual, not the exploitation. To use somewhat different terminology, I say that exploitation is mutually advantageous only when the transaction is Pareto superior, as, for example, in the case of the greedy snowstorm rescuer, described at the beginning of this chapter.52 Similarly, I shall distinguish between nonconsensual exploitation, where the exploited party does not give voluntary (or valid) consent, say, because of coercion or fraud, and consensual exploitation, where it appears that the exploited party has given voluntary and appropriately informed consent to the transaction. Do I beg the question by assuming that exploitation can be mutually advantageous and consensual? I think not. Although I think it plain that—as a fact about ordinary language—we sometimes use the word "exploitation" for cases that are mutually advantageous and consensual, suppose that we were to assume, for the sake of argument, that the word "exploitation" would be best limited to cases in which the exploitee is harmed. Nothing would have changed. We would still have to understand whether there are important distinctions between those cases that are (ex hypothesi) wrongly referred to as mutually advantageous exploitation and those mutually advantageous transactions that are not described in that way. It would remain an open question as to whether some mutually advantageous arrangements are wrongful and why they are wrongful. And it would remain an open question as to whether such transactions could be justifiably prohibited. To press this point a bit further, suppose we classify transactions as harmful to B or advantageous to B and as fair to B or unfair to B. Then we have four possibilities, as depicted in Figure 1.1. I shall ignore cases in cells A and B (assuming that there are transactions, as contrasted with impositions, that are both harmful and fair). This book is primarily although not exclusively an examination of the 52 Strictly speaking, we say that an outcome is Pareto superior if at least one party is better off and no party is worse off (in which case Figure 1.4 describes a Pareto-superior transaction), but for my purposes I shall define an outcome as Pareto superior only if the outcome leaves both parties better off (as in Figure 1.5).

OVERVIEW

15

Mutually Advantageous

Harmful

Fair

A

B

Unfair

C

D

Figure 1.1

cases in cell C—the the mutually advantageous and unfair ones. If someone wants to claim that, by definition, cases in cell C are not exploitative, I think that use of language is rather odd, but so be it. These are the cases in which I am interested, and we have to decide what we want to say about them. We will not solve important theoretical problems by definitional stipulations, for very little, if anything, of moral significance turns on the words or labels that we invoke. Now, if someone wants to claim that a mutually advantageous and consensual transaction cannot be unfair, then that is not a dispute over language. That is a substantive claim to the effect that cell C is empty, and not a linguistic claim that what is in cell C is not properly described as exploitative. I see no reason to think that position is correct. I said that I shall focus on mutually advantageous exploitation as contrasted with harmful exploitation because it poses more interesting theoretical issues. It also avoids what I shall call the problem of occlusion. Consider criminal wrongdoing. When a person is thought to have committed a criminal wrong, such as attempted homicide, the prosecution will often charge the defendant with the most serious applicable crime, but the charge may also involve "lesser included offenses," such as assault and battery.53 Yet when we think and talk about such cases, the "lesser included offenses" are apt to drop out of sight. The more serious offenses occlude the lesser offenses. Similarly, when we offer a moral description of an act, we typically invoke the strongest applicable moral description. And the strongest applicable moral description of some exploitative acts may not refer to their exploitativeness. Arneson notes, for example, that "murdering someone would not be characterized as exploitation even though the mur53

A slightly different version of this problem occurs when the federal government prosecutes someone for violating the "civil rights" of the victim, when the violation involved the murder or beating of the victim. Although we understand that the federal statute was developed because some (southern) states were unwilling to prosecute whites for harms against African Americans (or white civil rights workers), we are apt to think that violating the civil rights of the victim is, shall we say, an incomplete description of the wrong that was involved.

16

CHAPTER ONE

derer profits from the act and thus unfairly uses the victim."54 And Robert Goodin observes that "if a transfer were effected through sheer brute force, we would ordinarily not describe it (not primarily, anyway) as an act of exploitation. . . . The thief who simply seizes his victims' property does not just exploit them—he robs them."55 Although we are unlikely to use the word "exploitation" to describe a profitable murder or a typical mugging, it is not because it would be inaccurate to do so. Rather, the stronger moral description overlays and occludes our vision of the lesser included wrong of exploitation. When exploitation is harmful or coercive, it is often the harm or coercion that captures our attention, not the exploitation—whatever that turns out to be. Thus while some argue that where there is no harm and coercion (or fraud), there can be no exploitation, I think we will make more theoretical progress if we focus on those cases in which exploitation is at least arguably mutually advantageous or consensual, where the alleged exploitation is not occluded by harm or coercion. If, in the final analysis, there is no exploitation when an arrangement is mutually advantageous or consensual, if "there's no there, there," then we also will have learned something important. THE ELEMENTS OF EXPLOITATION

I want to travel light, without more assumptions about exploitation than we need to carry, but we will need some signposts for our journey. Let us start with the claim that A exploits B when A takes unfair advantage of B. Taking unfair advantage could be understood in two ways. First, it may refer to some dimension of the outcome of the exploitative act or relation, and this, it seems, has two elements: (1) the benefit to A, and (2) the effect on B. We may say that the benefit to A is unfair because it is wrong for A to benefit at all from his or her act (e.g., by harming B) or because A's benefit is excessive relative to the benefit to B. Second, to say that A takes unfair advantage of B may imply that A has been able to turn some characteristic of B or some feature of B's situation to his or her advantage.56 We imply that there is some sort of defect in the process by which the unfair outcome has come about or the formation of the agreement between A and B, for example, that A has coerced or defrauded or manipulated B. In the final analysis we may find that these three elements are not all necessary to account for exploitation, but they provide us with a way to begin. Let us consider them in a bit more detail. 54 55 56

Encyclopedia of Ethics, p . 351. G o o d i n , Reasons for Welfare, p p . 133-34. Feinberg, Harmless Wrongdoing, p . 179.

OVERVIEW

17

The Benefit to A

Bracketing for the moment what counts as an advantage to A, and also bracketing cases in which A attempts but fails to gain an advantage from a transaction with B, it seems that A cannot take unfair advantage of B unless A gets some advantage from B.57 We can see the relevance of the "benefit to A" by contrasting exploitation with other forms of wrongdoing, such as abuse, discrimination, and oppression. Let us say that A discriminates against B when A wrongly deprives B of some opportunity or benefit because of some characteristic of B that should not have been relevant to A's action. There was a period in American history in which many women became public school teachers because they were denied the opportunity to enter other professions, such as law and medicine. To the extent that the society benefited (in one way) from the pool of highly qualified public school teachers, the discrimination may have been exploitative, even if unintentionally so.58 On the other hand, if A refuses to sell a house to B solely because of B's race, then it would be odd to say that A exploits B. This sort of discrimination looks like a negative-sum game in which B loses and no one gains. As with discrimination, so with abuse. It has been alleged, for example, that medical students are frequently abused by oral insults and denigration and that this abuse may leave "long-lasting emotional scars."59 By contrast, it is also sometimes claimed that medical interns are exploited, that they work long hours for low pay. The contrast is just right. There is no reason to think that anyone gains— at least in any normal sense—from abuse, but it is at least plausible to think that the hospitals or patients gain from the exploitation of interns.60 Similarly, "a severely handicapped person who means well 57 In C h a p t e r 7, I will suggest that w e can say that A "acts exploitatively" t o w a r d B but d o e s n o t actually exploit B w h e n A unsuccessfully a t t e m p t s to gain at B's e x p e n s e .

58

This claim raises several thorny issues. First, even if society derived a benefit, it is not clear that the discrimination was intended to yield such a benefit. This raises the question of whether exploitation can be unintentional. Second, even if we ignore the effects on the women involved, it is arguable that such discrimination was not beneficial to the rest of society, all things considered, for what society gained in the way of teachers and nurses, it lost in the way of talent that could have been used in more important ways. 59

60

New York Times, J a n u a r y 26, 1989, p . B, 7.

If the hospital and its patients do not gain by the work conditions of interns, say, because the interns make too many mistakes because they are tired, we might want to redescribe it as attempted but failed exploitation. This also raises the question of whether exploitation should be understood in terms of a "current time slice" or a "whole life view." It might be argued that while interns are exploited qua interns, they benefit from the exploitation of other interns at a subsequent point in their careers, and

18

CHAPTER ONE

but cannot accomplish much could be abused, forced to perform drudgery . . . without being technically exploited."61 Let us say that A oppresses B when A deprives B of freedoms or opportunities to which B is entitled.62 If A gains from the oppressive relationship, as when A enslaves B, then A may both oppress and exploit B. But there is no reason to think that A always gains from oppression, and when A does not gain, there is no reason to regard the oppression as exploitative. Parents can be oppressive without being exploitative. Or consider the plight of the unemployed. We might want to say that the unemployed are oppressed, but unless we could specify the ways in which some gain from the fact that the unemployed are unemployed, it would be odd to say that the unemployed are being exploited. Of course Marxists would claim that capitalists can pay exploitative wages to the employed precisely because there is a "reserve army" of the unemployed with whom the employed must compete. Thus the unemployed may be exploited and not merely oppressed. But that merely confirms that exploitation involves a benefit, in this case, a gain to the capitalist class. The Effect on B

As our definitional survey indicated, some commentators maintain that exploitation resembles a zero-sum game, that the exploiter gains what the exploitee loses.63 Others maintain that exploitation is always harmful to the exploitee. The first claim is demonstrably false, for there can be harmful exploitation that is not zero-sum (or is zero-sum only in the loose sense that the exploitee loses while the exploiter gains). Harmful exploitation can occur in a negative-sum game when A's gain is less than B's loss, as in Figure 1.2. Harmful exploitation can also occur in a positive-sum game when A gains by imposing a loss on B that is less than A's gain, as in Figure 1.3. But whether harmful that this cancels the force of the earlier exploitation. Similar claims may be made about the relationship between senior partners in a law firm and junior associates. 61 Richard J. Arneson, "What's Wrong with Exploitation?" 91 Ethics 202 (1981), at 213. 62 Michael Freeman says that A oppresses B w h e n A "constitutes, or places a heavy b u r d e n o n B." See " O p p r e s s i o n " in Encyclopedia of Ethics, vol. 2, p . 921. Iris Marion Young says that "[oppression consists in systematic institutional processes which prevent some people from learning a n d using satisfying a n d expansive skills in socially recognized settings." See Justice and the Politics of Difference (Princeton: Princeton University Press, 1990), p. 38. 63 Tormey, "Exploitation, Oppression a n d Self-Sacrifice," p. 207.

19

OVERVIEW

Gains under No Utilization

Gains from A's Utilization of B

A

0

+5

B

0

-10

Total

0

-5

Figure 1.2 Harmful Exploitation as Negative-Sum Game Gains under No Utilization

Gains from A's Utilization of B

A

0

+ 10

B

0

-5

Total

0

+5

Figure 1.3 Harmful Exploitation as Positive-Sum Game exploitation is zero-sum, negative-sum, or positive-sum is not important in the present context, for it is relatively uncontroversial that exploitation can be harmful to B, as it is in both of the previous cases. There is, however, a further question as to whether A can exploit B if B is not directly affected by A's utilization of B (as in Figure 1.4) or if B gains from the transaction (as in Figure 1.5). Feinberg refers to some cases that fit the payoff structure depicted in Figure 1.4 as cases of harmless parasitism, as when A follows B's taillights in a dense fog. A uses B to A's own advantage, but does not render B worse off (assume that B is not bothered by A's headlights in B's mirror), for A's act does not affect B at all. 64 In other cases of nonharmful exploitation, the transaction appears to benefit both A and B, as when A rescues a stranded motorist for a fee. Now, in asking how A's action affects B's interests, we must be careful to adopt an all-things-considered point of view. There are, after all, negative elements in virtually all uncontroversially beneficial transactions. Paying money for a good that is clearly worth the price is still a negative element in the transaction. It would be better to get it for free. If A and B enter into a cooperative agreement where A gives B $100 for a book that is worth a lot to A (because it completes a collection) but is worth little to B, we do not say that B has been harmed by the transaction just because B has lost the book any more than we say that A has 64

Feinberg, Harmless Wrongdoing, p. 14.

20

CHAPTER ONE Gains under No Utilization

Gains from A's Utilization of B

A

0

+5

B

0

0

Total

0

+5

Figure 1.4 Harmless Parasitism No Cooperation

Cooperation

0

+ 10

B

0

+2

Total

0

+12

A

r

Figure 1.5 Mutually Advantageous Exploitation been harmed because the transaction required A to pay $100. Here is a case in which the heterogeneity of preferences—the fact that A and B value the book and the money differently—work to the advantage of both parties. Similarly, we do not say that a worker is harmed by employment merely because the worker prefers leisure to work. If the benefits to B from employment are greater than the costs to B from being employed, then employment is beneficial to B, all things considered. Thus in deciding whether a case of alleged exploitation should be classified as harmful exploitation or mutually advantageous exploitation, we must look at its net effect on B. With this in mind, let us consider mutually advantageous exploitation in a bit more detail. It should be obvious that I am not arguing that any distribution that looks like Figure 1.5 is a case of mutually advantageous exploitation. To characterize a payoff structure as exploitative requires us to say that the payoffs are unfair, and nothing in this payoff matrix entitles us to make such an assumption. I am simply arguing that some cases of alleged exploitation—those that are mutually advantageous—have this sort of payoff structure. We can characterize this case in somewhat different language. All cases of mutually advantageous exploitation are ones in which a transaction between A and B generates a social surplus as contrasted with the no-transaction baseline. In any potential transaction, each party has a reservation price, that is, the value that the person must receive if he or she is to agree to the transaction. Suppose that A is a home buyer and that B is a home seller. Suppose that A is prepared to pay no more than $175,000 for B's house and that B is not prepared to

OVERVIEW $150,000 $100,000

" B's Reservation Price

21 $175,000 " $200,000 A's Reservation Price

Figure 1.6 Zone of Agreement sell to A for less t h a n $150,000. 6 5 A's reservation price is $175,000 a n d B's reservation price is $150,000. In principle, A a n d B are willing to m a k e a n a g r e e m e n t at a n y price b e t w e e n $150,000 a n d $175,000. T h e space b e t w e e n t h e parties' reservation prices constitutes their zone of agreement, or bargaining range (see Figure 1.6). A n y o u t c o m e w i t h i n their z o n e of a g r e e m e n t g e n e r a t e s a social surplus, w h i c h could be d e fined as t h e difference b e t w e e n t h e b u y e r ' s a n d seller's reservation prices. 6 6 If A a n d B agree o n a price of $160,000, t h e social s u r p l u s is $25,000: A receives t h e h o u s e for $15,000 less t h a n A w a s willing to p a y ($175,000) a n d B receives $10,000 m o r e t h a n B w a s willing to accept ($150,000). Joel Feinberg a r g u e s that if a transaction is mutually a d v a n t a g e o u s , t h e n it is w r o n g to say that o n e party gains at the o t h e r party's exp e n s e . 6 7 If A gains at B's e x p e n s e only w h e n A h a r m s B, t h e n Feinb e r g is right b y definition. But there is a n i m p o r t a n t s e n s e in w h i c h a n y marginal gain to o n e p a r t y w i t h i n t h e z o n e of a g r e e m e n t is ind e e d at t h e o t h e r party's e x p e n s e : while the parties m a y prefer a n y o u t c o m e w i t h i n t h e z o n e of a g r e e m e n t to the n o n a g r e e m e n t solution, t h e y are n o t indifferent to t h e distribution of t h e social s u r p l u s w i t h i n t h e z o n e of a g r e e m e n t . 6 8 Each w o u l d prefer a price t h a t is furthest from his or h e r reservation price. A n d a n y m o v e m e n t a w a y from o n e ' s o w n reservation price is, in that sense, at t h e o t h e r party's ex-

pense, although it is not harmful to the other party as contrasted with the no-transaction baseline. What I have called mutually advantageous exploitation occurs when B gains relative to the noncooperation baseline but the distribu65

They may not know each other's reservation prices and may do their best to conceal them. Indeed, they may not know their own reservation prices until they are put to the test. 66 Robert Frank, Passions within Reason (New York: W.W. N o r t o n , 1988), p . 164. 67 Feinberg, Harmless Wrongdoing, p. 178. 68 Rawls makes a similar point about the role of principles of justice: " [ A l t h o u g h a society is a cooperative venture for mutual advantage, it is typically marked b y a conflict as well as b y a n identity of interests. There is a n identity of interests since social cooperation makes possible a better life for all than any would have if each were to live solely by his o w n efforts. There is a conflict of interests since persons are not indifferent as to h o w the greater benefits produced b y their collaboration are distributed. . . ."A Theory of Justice, p. 4.

22

CHAPTER ONE

tion of the social surplus is unfair to B. Consider a garden-variety case of alleged exploitation. An unexpected blizzard hits an area and people rush to the hardware store to buy a shovel. The hardware store owner sees the opportunity to make an abnormal profit and raises the price of a shovel from $15 to $30. If B agrees to pay $30 for the shovel, because the shovel is worth more than $30 to B under the circumstances, then the transaction is clearly Pareto superior. Both parties gain. But B feels exploited because B gains less (or pays more) than B thinks reasonable.69 A similar structure applies to some of the other cases of alleged exploitation with which we began—profits from the AIDS drug AZT, surrogate motherhood contracts, organ sales. We need not deny that B benefits from these transactions, all things considered. Rather, we say that A exploits B because we believe that B pays too a high price for what he or she gains. I have suggested that a mutually advantageous transaction is (wrongly) exploitative only if the outcome is (in some way) unfair to B. This is not merely definitional. After all, it may be thought that a mutually advantageous transaction is exploitative if A takes advantage of B's vulnerabilities or desperate situation to strike a deal that is, nonetheless, advantageous to B—given B's desperate situation. But that is false. For if A makes a reasonable proposal that B has no alternative but to accept (A charges B $15 for the shovel) given B's desperate situation, it would be wrong to claim that A exploits B. And we cannot say that A exploits B when A gets a "better deal in negotiation than [B] would have obtained for the same good . . . had [B] lacked the difficulty," because it is often the case that there would be no transaction between A and B were it not for B's difficulty.70 There is, of course, a normal price for snow shovels that are purchased under nonemergency conditions. But there is no normal price for AZT for people who do not desperately need AZT. And there is no normal price for rescuing stranded motorists who are not in difficulty, because to be a stranded motorist is to be in difficulty. It might be said that what I have called mutually advantageous exploitation can and should be understood as a form of harmful exploitation. If we evaluate a transaction by reference to a fairness baseline as contrasted with a no-transaction baseline, then we can say that B is harmed by paying $30 for a shovel by comparison with the fairness 69

I say "feels exploited" because we should not assume, without further argument, that it is unfair for A to charge $30. 70 Scott Altman, "A Patchwork Theory of Blackmail/' 141 University of Pennsylvania Law Review 1639 (1993), at 1644.

OVERVIEW

23

baseline (say, where B pays $15 for a shovel) even if B gains by comparison with the no-transaction baseline or B's reservation price.71 But, once again, such relabeling would not change anything. Although it might permit us to say that all exploitation is harmful to the exploitee, we would still have to distinguish between those cases in which B is harmed relative to both the fairness baseline and the notransaction baseline and those cases where B is harmed only by refer ence to the fairness baseline but not by reference to the no-transaction baseline. The distinction I have drawn would simply reappear in other terms. Still, it may be argued that the proposed distinction between harmful exploitation and mutually advantageous exploitation ignores a deeper—Kantian—way in which what I have called mutually advantageous exploitation is actually harmful to B, namely, that A treats B merely as a means to be utilized to his own advantage rather than as an end in herself. And so treating a person is to harm her. Along these lines, Allen Buchanan argues that exploitation occurs "whenever persons are harmfully utilized as mere instruments for private gain," and adds that this could apply to business transactions between two affluent bankers.72 Although both bankers have property in the means of production and there are no class differences between them, "[e]ach harmfully utilizes the other as a mere means to his own advantage. Each views the needs and desires of the other not as needs and desires, but rather as levers to be manipulated, as weaknesses to be preyed upon." 73 It is not clear what to make of this view. First, and on at least one plausible reading of the Kantian maxim, one treats another as a mere means only when one treats "him in a way to which he could not possibly consent," as in cases of coercion and fraud.74 On this view, A violates the Kantian maxim only when A seeks to undermine B's capacity as an autonomous decisionmaker within B's objective situation. A does not violate the Kantian maxim when seeking to profit from a situation in which there is a mutuality of interests with B. And there is no reason to think that each banker could not possibly consent to be so treated by the other banker if each banker gains from the transaction as well. 71

Does B exploit A if B purchases a snow shovel in March for less than A's cost because A bought too many snow shovels in anticipation of a snowy winter? 72 Marx and Justice (Totowa, N.J.: R o w m a n a n d Allanheld, 1984), p . 44. 73 Ibid., p . 39. 74 Christine Korsgaard, ' T h e Reasons We C a n S h a r e / 7 10 Social Philosophy and Policy 24 (1993), at 40.

24

CHAPTER ONE

Bracketing that difficulty, to say that A exploits B when A "harmfully utilizes" B as a "mere means" is equivocal as to whether "harmfully" is a reinforcing or modifying adverb. On one view, "harmfully" is a reinforcing adverb, because the utilization of B merely as a means constitutes an independent harm to B. On another view, "harmfully" is a modifying adverb, because we can contrast the cases in which A harmfully utilizes B as a mere means with cases in which A nonharmfully utilizes B as a mere means.75 On this view, the bankers may utilize each other as means, but absent an independent form of harm, there is no reason to think that they are harmed by their utilization as a means itself. Suppose we accept the first interpretation, on which being utilized as a means constitutes a harm to B. We would still have to distinguish between what I have called harmful exploitation and mutually advantageous exploitation, although we might prefer different terminology. First, if we regard the "mere use" of B as a distinct form of (moral) harm to B, it is possible that the "mere use" dimension of harm can be outweighed by other benefits to B, in which case B may or may not be harmed, all things considered. Still, we might reject that possibility, say, because the "mere use," or Kantian, dimension of harm always outweighs or trumps any other benefits that B might receive from being so used, so that all instances in which A uses B as a mere means to A's own advantage are harmful to B, all things considered. Yet even if this view were correct, and that seems quite unlikely, we might still want to distinguish between those cases in which B is harmed apart from being treated merely as a means from those in which B is not harmed apart from the harm that derives from being treated merely as a means. And so it seems better simply to grant that some allegedly exploitative transactions are mutually advantageous and go on to ask two questions: (1) what distinguishes mutually advantageous transactions that are exploitative from those that are not, and (2) what is the moral force of that distinction? The answer to the first question will, of course, require a principle of fairness for the distribution of the social surplus. The answer to the second question will, among other things, require us to determine whether society is justified in prohibiting mutually advantageous exploitative transactions, particularly when B voluntarily agrees to the transaction. And that brings us to the third dimension of exploitation. 75

See Allen Buchanan, Ethics, Efficiency, and the Market, p. 88.

OVERVIEW

25

Process

As we have seen, it seems plausible to argue that A does not exploit B simply because there is unfairness in the distribution of rewards, that exploitation also depends on the way in which that distribution came about. If B voluntarily agrees to what might otherwise be a maldistribution of advantages, as when B voluntarily decides to make a gift of goods or labor to A, then it seems wrong to say that A has exploited B. It would, for example, be odd (although perhaps not impossible) to claim that a hospital exploits its volunteer workers just because the workers are volunteers rather than paid employees. Interestingly, both Marxists and libertarians seem to accept the view that voluntary transactions cannot be exploitative. Marxists tend to adopt a "force-inclusive definition" of exploitation.76 Marxists do not say that capitalists exploit their workers in spite of the fact that the workers voluntarily agree to their employment status. No, they argue that workers are exploited because they do not voluntarily agree to their employment status. Marx and Marxists concede that the proletariat is not enslaved, because the workers are not tied to any particular employer, but they maintain that, appearances to the contrary notwithstanding, they are coerced. They transfer their labor to the capitalist under the "dull compulsion of economic relations," but compulsion it is, its dullness notwithstanding.77 Libertarians can be understood as accepting this "force-inclusive" definition of exploitation, but come to the opposite conclusion. They could maintain that since market transactions are not coerced, the workers are therefore not exploited. I shall suggest that we do not need to accept these alternatives, that A can exploit B even if B is not coerced (or defrauded). The question arises, however, as to whether what I have (it may be argued misleadingly) referred to as consensual exploitation involves some other sort of defect in consent that does not amount to coercion or fraud. Let us press this issue a bit further. There are some instances of alleged exploitation in which the issue of consent does not seem to arise at all. Although this project focuses on exploitative interactions and transactions, on situations in which A gets B to do something, there are cases in which the exploitee may be entirely passive. A may sell photographs of B without B's knowledge, or rob a purse from a 76

See Reiman, "Exploitation, Force, a n d the Moral Assessment of Capitalism/' p. 3. See Jon Elster, "Exploitation, Freedom, a n d Justice/' in J. Roland Pennock a n d John C h a p m a n , eds., Nomos XXVI: Marxism (New York: N e w York University Press, 1983), p p . 277-78. 77

26

CHAPTER ONE

sleeping B or follow B's taillights in a dense fog.78 In these cases, B's will is not involved at all. I shall refer to this as nonvolitional exploitation. If nonvolitional exploitation operates without the engagement of B's will, then nonconsensual exploitation operates against B's will or without B's appropriately informed will. Coercion and fraud are two principal ways in which A compromises the voluntariness or validity of B's consent.79 Coercion compromises the voluntariness of consent; by distorting the information on which B gives consent, fraud compromises the rationality or advisability of B's consent. In general, A coerces B to do X only if A proposes (threatens) to make B worse off with reference to some baseline condition if B chooses not do X, although, and as we shall see, specifying the appropriate baseline against which to measure the proposal can be a complicated question.80 If A gets B to pay A $100 per week by threatening to bomb B's store if B does not pay up, then A coerces B into paying $100 a week. By contrast, if A gets B to pay A $100 per week by proposing to clean B's store each night, then A has made a noncoercive (or inducive) offer to B. A does not propose to worsen B's situation if B rejects A's proposal. Fraud also undermines the voluntariness or, perhaps more accurately, the validity of B's consent. Suppose that A offers to sell B a car for $10,000. A tells B that the car has been driven only 50,000 miles, but has set back the odometer from 90,000 to 50,000. We would not say that B is coerced into buying A's car, but we would say that B has been deceived, and that B has not given valid consent, because valid consent must be informed (or not misinformed) as well as free. Both coercion and fraud produce a situation in which B chooses to do that which a rational B would not or might not have done in the "preinteraction" situation (or normatively required situation). And they 78

Indeed, the exploitee may be dead. Robert Goodin suggests that Norman Mailer's book The Executioner's Song (Boston: Little Brown, 1979), exploited the plight of Gary Gilmore, a convicted murderer, who asked to be executed. See Goodin, "Exploiting a Situation and Exploiting a Person," pp. 172ff. 79 I shall not put much pressure on the distinction between these two sorts of claims: (1) B gives voluntary consent, but it is not valid because it is not suitably informed; (2) B does not give suitably voluntary consent because it is not suitably informed. John Lawrence Hill argues that A may exploit B by disturbing B's capacity to reason effectively. See Hill, "Exploitation." 80 I say "only if" because A's proposal may not get B to d o X, in which case A has m a d e a coercive proposal to B but has not coerced B. On this and the question of how to set the baseline, see m y Coercion (Princeton: Princeton University Press, 1987).

OVERVIEW

27

both typically produce results that are harmful to B, all things considered.81 By contrast, there are at least some cases of alleged exploitation in which B's consent is not defective in either of these ways. In many cases of alleged exploitation, A takes advantage of B's circumstances to get B to agree to a mutually advantageous transaction to which B would not have agreed under better or perhaps more just background conditions, where A has played no direct causal role in creating those circumstances, where A has no special obligation to repair those conditions, and where B is fully informed as to the consequences of various choices. Although B might prefer to have a different range of options available, B can make a perfectly rational decision as to the advisability of the various options. Consider, for example, the decision of a low-income person to enlist in the military. As David Wasserman has put it, It seems hard to deny that the actual choices made by poor recruits are often fully voluntary, and even quite reasonable under the circumstances in which they make them. The unemployed young father who joins the service in order to provide decent housing and schooling for his family may be valuing his life appropriately.82 It may be argued that the recruit's poor background circumstances "force" the recruit to join the military—given the range of options, the recruit has no other acceptable choice. Still, there would be a distinction between this sense of "force," where no one threatens to make the recruit worse off if the recruit does not enlist in the military, and the sort of coercion involved in extortion. And there is a distinction between a case in which background conditions make it rational for B to enlist and a case in which B enlists because the recruiter has fraudulently described the benefits of military service. I do not want to beg the question as to whether this sort of decision involves some other sort of defect in the quality of B's consent, even if there is no coercion or fraud. It might be argued, for example, that choices made under conditions of desperation or from an inequality of bargaining power are not appropriately consensual. An analysis of this line of argument will have to wait. But once again, labels do not an argument make. Even if we label as nonconsensual those cases in 81

It is possible for A to coerce or defraud B into doing something that is not harmful to B, as when, say, A sets back the odometer because A knows that B is unduly concerned about mileage but A's proposed price is otherwise eminently fair. A could also paternalistically coerce B into an agreement that is beneficial to B. 82 Wasserman, "Enlistment and Exploitation/' p.6.

28

CHAPTER ONE

which B enters into a mutually advantageous transaction with A because of poor and unjust background conditions, we would still have to contrast the cases that are nonconsensual because of coercion or fraud and typically harmful to B and those that are nonconsensual because of poor and unjust background conditions and advantageous to B, given those background conditions. And we will still have to ask what the moral force of such exploitation amounts to: Should we prohibit A from making such proposals? Should we refuse to enforce agreements made under such conditions? And that brings us to the moral force of exploitation. MORAL WEIGHT AND MORAL FORCE

I have suggested that the notion of exploitation provides a moral description of a transaction, but that its moral force is less clear.83 To recall, I distinguish between the moral weight and moral force of exploitation. The moral weight of exploitation refers to the intensity of its wrongness—however that is measured. The moral force of exploitation refers to the various moral upshots or reasons for action that exploitation might or might not involve for parties to the transaction or for society: Can people have a right to exploit? Is it wrong to allow oneself to be exploited? Should society prohibit actions if they are exploitative? Should we refuse to enforce exploitative agreements? The moral force of harmful and nonconsensual exploitation is relatively unproblematic. Whatever the added moral importance of the gain to A from the harm to B, it is certainly at least prima facie wrong for A to harm B, and it seems that the state is at least prima facie justified in prohibiting or refusing to enforce such transactions.84 Still, there are interesting questions about the moral evaluation or moral weight of harmful exploitation. Is harmful exploitation worse than nonexploitative harm? Is it worse, for example, if A enslaves a talented B, whose efforts produce a large gain to A, than if A enslaves an untalented B, whose efforts produce no gain for A? 83 We might think of exploitation as a "thick [concept] . . . such as treachery a n d promise a n d brutality and courage, which seem to express a union of fact a n d v a l u e / ' Bernard Williams, Ethics and the Limits of Philosophy (London: Fontana, 1985), p p . 12930. 84 Michael Gorr has argued, for example, that w e cannot sensibly speak of the "justified exploitation of another p e r s o n . " See Coercion, Freedom and Exploitation (New York: Peter Lang, 1989), p . 147. That, I think, seems too strong, for it is possible that A can exploit B in ways that are justifiable, all things considered, at least if w e think that good consequences can override prima facie wrongness.

OVERVIEW

29

Mutually advantageous transactions present a more difficult set of problems. Even if a transaction between A and B is unfair, it might be thought that there can be nothing seriously wrong about a Paretosuperior agreement, particularly if A has no obligation to enter into any transaction with B. And even if there is something seriously wrong about such transactions, it might be argued that it could not be a wrong that would justify state intervention. Consider the case in which a hardware store (A) charges a customer (B) an abnormally high price for a snow shovel during a blizzard.85 Some might maintain that there could be nothing wrong in charging an abnormally high price if the transaction leaves B better off.86 And, it might be argued, even if A acts wrongly or fails to act virtuously in such cases, A does not harm anyone or violate anyone's rights, and only harm or rights violations justify state intervention.87 Perhaps this view is correct. Bracketing arguments based on externalities, it seems perfectly plausible to maintain that the state is justified in interfering with transactions only if one party is violating the other's rights. Indeed, it might be thought completely irrational to prohibit A and B from entering into a Pareto-superior transaction just because the transaction is unfair to B or arises out of unfair background conditions. That said, it is important to note that those who invoke the concept of exploitation frequently maintain that such exploitation provides a reason for state intervention, even when the transaction seems to be mutually advantageous. When it is claimed that commercial surrogacy exploits the birth mothers, the critics are not only offering a moral description of the practice. They typically claim that surrogacy contracts should be unenforceable or entirely prohibited. When it is said that the volunteer army is exploitative, it is also said that our society should therefore opt for some form of conscription.88 85 In one survey, 82 percent of the respondents said that it was unfair for A to take advantage of the short-run increase in demand. See Daniel Kahneman, jack L. Knetsch, and Richard Thaler, "Fairness as a Constraint on Profit Seeking: Entitlements in the Market/' 76 American Economic Review 729 (1986). 86 It might b e a r g u e d that a high price performs a socially useful rationing function, by e n s u r i n g that t h e shovels are allocated to those w h o place t h e highest value u p o n t h e m , as revealed b y their willingness to pay. 87 A s James Buchanan argues: "There is . . . n o legitimate basis for prohibiting or restricting market exchange in a n y commodity or service so long a s individual preferences are accepted a s controlling a n d so long as c o n s u m p t i o n externalities are a b s e n t . " "Political Equality a n d Private Property: T h e Distributive Paradox," in Gerald Dworkin, G o r d o n Bermant, a n d Peter Brown, e d s . , Markets and Morals (Washington, D.C.: Hemis p h e r e Press, 1977), p. 72. 88 Wasserman, "Enlistment and Exploitation."

30

CHAPTER ONE

On what moral principles might this view rest? Joel Feinberg has maintained that because mutually advantageous exploitation is not harmful, such exploitation would constitute a "free-floating evil," a wrong that is bad for no one; "In these cases there is no wrongful loss for the exploitee, who can himself have no grievance."89 He goes on to maintain that since mutually advantageous exploitation is a freefloating evil, an argument for the prohibition of such exploitation would have to appeal to a principle of legal moralism or perfectionism, which he is quite reluctant to endorse as a basis for limiting liberty. There are, I suppose, three questions here: Is mutually advantageous exploitation a free-floating evil? Does the prohibition of mutually advantageous exploitation rest on legal moralism? If so, could such legal moralism be defended? Is mutually advantageous exploitation a free-floating evil? I think not. Suppose that B and C both need blood transfusions, and that the only available blood is compatible with B's blood type, but not with C's blood type. There are only two possible worlds: (1) no transfusion, where neither B nor C gets a transfusion, and (2) transfusion, where B gets the transfusion and C does not (see Figure 1.7). To say that it would be wrong—in any way—to give the transfusion to B would seem to involve a "free-floating evil." Giving the transfusion to B is good for B and bad for no one, including C.90 If we give the transfusion to B, it is not just that C has suffered no loss. There is no feasible alternative world in which C could have gained. No Transfusion

Transfusion

B '

0

5

C

0

0

Figure 1.7

But the world of mutually advantageous exploitation is not like this. Recall the "snow shovel" example. Here there are, let us say, three feasible alternative worlds: (1) no transaction; (2) Transaction 1, where A sells B a shovel for an exorbitant $30; and (3) Transaction 2, where A sells B a shovel for the normal price of $15 (see Figure 1.8).. Is Transac89

Feinberg, Harmless Wrongdoing, p. 176. I assume that there is no risk to the donor. Larry Temkin might claim that there is at least one sense in which the world in which the transfusion is performed is worse: it creates an inequality between B and C. Temkin notes that it would not follow that it is worse when all things are considered, for "equality is not all that matters." See Inequality (New York: Oxford University Press, 1993), p. 248. 90

OVERVIEW

31

No Transaction

Transaction 1

Transaction 2

0

10

8

0

5

10

Figure 1.8 tion 1 better for A and worse for no one? Yes and no. Yes, when compared to the no-transaction baseline. No, when compared with Transaction 2. Transaction 1 is Pareto-superior to the no-transaction world, but it is not Pareto-superior to the Transaction 2 world. By comparison with Transaction 2, the "wrong" in Transaction 1 is not free floating. B is not harmed in Transaction 1, but B's interests are clearly negatively affected by A's choice to engage in Transaction 1 as contrasted with Transaction 2. This point also holds with respect to what Feinberg calls harmless parasitism. Recall Feinberg's example in which A follows B's taillights in a dense fog.91 Assume once again that A's following B has no negative effects on B, that B is not bothered by A's headlights in the rearview mirror. Is it plausible to say that A is exploiting B, and if so, would it be a free-floating evil? Yes, and not necessarily. Suppose that A, having followed B for ten miles, could pass B so that B could then follow A for a while, thereby reducing the burden on B. We then can compare three worlds: (1) no following; (2) A follows B; and (3) A and B alternate (see Figure 1.9). If (1) and (2) were the only possible worlds, say, because A cannot (safely) pass B, then to think that the harmless parasitism in (2) is wrong would appeal to a free-floating evil. But if (3) is a feasible world, then B is less well off in (2) than in (3). And so to condemn (2) by comparison with (3) is not to appeal to a free-floating wrong, as I understand that concept. To say that the wrong involved in mutually advantageous exploitation is not free floating does not establish its moral weight or moral force. Still, if the wrong involved in mutually advantageous exploitation is not free floating, the prohibition of mutually advantageous exploitation would not have to rest on moralism, at least not moralism of an ordinary sort. There might be important, albeit not necessarily overriding or dispositive, moral reasons for preferring a legal regime that does not intervene with respect to mutually advantageous exploitation, wrongful though it is. At the same time, I shall suggest that there may be reasons that override the presumption against legal in91

Feinberg, Harmless Wrongdoing.

32

CHAPTER ONE

(2) (3) (1) A Follows A and B No B Alternate Following A

0

10

5

B

0

0

5

Figure 1.9 tervention with mutually advantageous exploitation. All this remains to be seen. Of course, even if there are good reasons not to interfere with most cases of mutually advantageous exploitation, it does not follow that the moral force of that exploitation is trivial. The disposition not to take unfair advantage of others may be among the more important moral virtues and a necessary condition of civilized life, even if there are also good reasons for not penalizing the failure to display that virtue.92 But what about the exploitee? Assuming that it is in the interest of the exploitee to allow the exploitation (as contrasted with the no-transaction alternative), does the exploitee have any moral reason not to allow the exploitation? That is less clear, but I shall suggest that the answer may sometimes be yes. I shall say a good deal more about the moral force of exploitation. Still, it is important to note that even if exploitation is seriously wrong, it may not be the worst form of injustice or inequality. Although this is not a project in comparative immorality, it is important to put exploitation in some perspective. Suppose that social justice requires a relatively egalitarian distribution of resources. If an inequality between A and B is exploitative only if there is some causal relationship between A's and B's positions, then egalitarians might be well advised to "completely forget the idea, still popular in certain quarters, that the root of social injustice is exploitation—in the sense of a failure to reward people in accordance with their productive contribution or the true value of their labor/' 93 Indeed, given that an egalitarian society will be particularly concerned to provide for those who are unable to contribute much if anything, a just society will be concerned to ensure that some people "receive much more of the social product than they contribute."94 In effect, the less advantaged will be 92 "Justice is the disposition not to take advantage of one's fellows, not to seek free goods or to impose uncompensated costs, provided that one supposes others similarly disposed." David Gauthier, Morals by Agreement (Oxford: Clarendon Press, 1986), p. 113. 93 Nagel, Equality and Partiality, pp. 99-100. 94 Ibid. Moreover, if "the unemployed . . . are not exploited . . . since they do not produce any surplus value for the capitalist to appropriate," we may well conclude that being excluded and unemployed may be much worse than being exploited. Will

OVERVIEW

33

taking advantage of others, although perhaps not unfair advantage of others. In addition, if the degree to which one is exploited is a function of the gap between what one receives and what one contributes, and if one's contributions are, in turn, a function of morally irrelevant factors (such as social background and native talents), then the degree of exploitation may be of less moral relevance than first appears.95 Still, even if exploitative inequality is not always worse than nonexploitative inequality, particularly when the nonexploitative inequalities are greater than the exploitative inequalities, it is an interesting and important question whether the inequalities and suffering that arise from exploitation have a special call on our moral attention. PLAN OF THE BOOK

Those are my aims. This is my plan. Rather than immediately try to continue the analysis at the most abstract level, my strategy in Part I is to consider these issues in several specific contexts of alleged exploitation, to let the analysis of specific contexts of alleged exploitation shape our theory of exploitation (I say "alleged" exploitation because we may determine that some exploitation claims are difficult to defend). Although the three elements of exploitation (benefit to A; effect on B; process) and the moral force of exploitation will arise in each context, we will find that some contexts put more pressure on some elements than others. The analysis of unconscionable contracts (Chapter 2) will help us to understand the notion of a fair bargain, inequality of bargaining power, and the rationale for various legal remedies. The analysis of the exploitation of student athletes (Chapter 3) will stress the importance of the gain to the exploiter—the disKymlicka, Contemporary Political Philosophy (New York: Oxford University Press, 1990), p. 176. As Erik Olin Wright has put it, if exploitation in labor implies the appropriation of "at least part of the social surplus by the oppressor," then the unemployed may be oppressed but they are not exploited. "A General Framework for the Analysis of Class Structure," 13 Politics and Society 383 (1984), at 387. 95 A similar point applies to inequalities among nations. Marxists have long maintained that the inequalities among nations are due to a form of harmful exploitation. On this view, the poor nations of the world are poor because they have been impoverished by the affluent nations. If the impoverishment view were correct, then we would expect that there would be a positive correlation between the level of contact with— exploitation by—the affluent nations and the relative poverty of the poorer nations. Instead, the poorest nations are those that have had the least economic contact with the highly industrialized nations. P. T. Bauer, Equality, The Third World and Economic Delusion (Cambridge: Harvard University Press, 1981), p. 67. Many newly industrializing nations have suffered not so much because imperialism has made them worse off, but because the affluent nations have "found too little there to be exploited." Robert Gilpin, U.S. Power and the Multinational Corporation (New York: Basic Books, 1975), p. 289.

34

CHAPTER ONE

tinction between being exploited and being treated badly—and will raise some serious questions about the moral weight of exploitation. The analysis of commercial surrogacy (Chapter 4) will ask, among other things, whether "commodification" objections to a practice can be linked to the claim that it is exploitative and will put particular pressure on questions about the justification of state intervention. The analysis of the "unconstitutional conditions" doctrine (Chapter 5) will allow us to understand more fully the justification for restrictions on Pareto-superior transactions. Although that doctrine is not typically couched in terms of exploitation, it maintains that some mutually advantageous and arguably consensual agreements between governments and individuals (or other governments) should not be permitted, and so is of direct relevance to our project. The analysis of sexual exploitation of patients in psychotherapy (Chapter 6) will raise important questions about consent: What counts as genuine consent? Can one be exploited even if one consents to the alleged exploitation? There are, of course, other important contexts in which claims of exploitation figure prominently, some of which I noted at the outset of this chapter. For the most part, I shall leave them aside. The book is long enough already. More important, although many of the topics (for example, the sale of body organs) are inherently interesting and important, I do not think that an analysis of those topics would add much to our theoretical understanding of exploitation beyond what is produced by the contexts that I do consider. Diminishing marginal utility applies here as elsewhere. In Part II, I return to the theoretical issues with which we began. In Chapter 7, I develop an account of unfair bargains. In Chapter 8, I consider the issue of consent. In Chapter 9, I try to say something more systematic about the moral weight and moral force of exploitation. To anticipate a small but important set of my conclusions, I will argue that mutually advantageous agreements can be unfair and therefore exploitative. I will argue that a defect in consent is not a necessary condition of exploitation, but that the moral weight and moral force of nonconsensual exploitation may be different from the moral weight and moral force of consensual exploitation. I will argue that inequalities of bargaining power do not constitute a defect in consent, nor do desperate or unjust background conditions. With respect to moral force, I will maintain that there are good but not conclusive reasons to allow and enforce mutually advantageous and voluntary exploitative agreements. I will also argue that while the weight of moralistic or perfectionist reasons for interfering with such agreements is highly problematic, the prohibition and nonenforcement of such agreements can sometimes be justified without appeal to such arguments.

Chapter Two

UNCONSCIONABLE CONTRACTS

MCNAMARA ADVERTISED a color

television set on a "rent to own" plan that required "no deposit," "no credit," and "no long-term obligation."1 After seeing the advertisement, Carolyn Murphy, a welfare recipient, leased a 25-inch Philco color television set from McNamara. She agreed to pay a $20 delivery charge and seventy-eight weekly payments of $16, after which she would own the set, having paid $1,268 for a set that retailed for $499. After paying $436 over a sixmonth period, Murphy saw a newspaper article criticizing the plan and stopped making payments. McNamara sought to repossess the set, threatening to file criminal charges if Murphy failed to return it. Murphy filed for an injunction against repossession, claiming that the agreement was unconscionable and therefore unenforceable. The court granted the injunction on the ground that "an agreement for the sale of consumer goods entered into with a consumer having unequal bargaining power, which agreement calls for an unconscionable purchase price, constitutes an unfair trade practice."2 Many legal principles serve to prohibit, deter, or compensate for harmful and nonconsensual exploitation. In the present case, however, Murphy sought out McNamara, albeit in response to the latter's advertisement. There is no evidence of coercion, duress, or fraud. And it is not clear that she was harmed by the deal. In nullifying Murphy's contract, the court appears to be saying that it will not allow one party to take unjust advantage of another party even if the parties agree to the transaction, even if the relation is (arguably) mutually advantageous, and even when the transaction does not involve a necessity—at least when the transaction is so unfair as to be "unconscionable." In effect, the law seems to regard A's exploitation of B, per se, as a reason for legal intervention. But what makes a contract unconscionable? And if a contract can be both unconscionable and mutually advantageous, should such a contract be unenforceable? In this chapter, I explore the doctrine of unconscionability in the hope that 1 Murphy v. McNamara 416 A.2d 170, 36 Conn. Supp. 183 (1979). The injunction was granted, but McNamara was permitted to file suit for the difference between the amount Murphy had already paid and the value of the set. 2 Ibid., at 177.

38

CHAPTER TWO

we can discover principles and arguments that will help us understand the characteristics and moral force of exploitation. CONTRACT LAW

If unconscionability is a reason for invalidating a contract (or not requiring performance), this is likely to be so because unconscionable contracts are not consistent with the purposes of contract law. So let us start there. At the most general level, contract law can be justified in two ways. 3 First, it may be argued that contract law facilitates individual autonomy, or freedom. 4 To be autonomous is to be able to plan and control one's life, and a robust conception of autonomy includes the ability to form binding relationships with others. If A and B want to trade A's X for B's Y, where, say, A must perform before B, the trade may not occur unless B can be held to the terms of the agreement. By enforcing such agreements, the law facilitates both A's and B's autonomy. Second, contract law can also be justified in consequentialist or utilitarian terms. From this perspective, contract law promotes utility by allowing individuals to put their resources to more valued uses. 5 If A values B's Y more than his X and B values A's X more than her Y, then an exchange enhances value for both parties. Nonetheless, the parties may refuse to make a value-enhancing agreement if each fears that the other will enjoy the benefits of the agreement and then fail to do his part. By providing sanctions for breach, contract law reduces the magnitude of the compliance problem, thereby facilitating Pareto-superior transactions. 6 The contrast between these background justifications for contract law is (roughly) replicated in the way in which we evaluate the morality of contracts. We can, it seems, evaluate an agreement or contract in terms of both the process of contract formation and the contract's substance. From the perspective of autonomy, we might say that the process of contract formation is morally legitimate if the parties come to 3 Some may argue that there is, in the final analysis, no distinction between these justifications. For the purpose of the present argument, I do not need to take a position on that argument. 4 See Charles Fried, Contract as Promise (Cambridge: Harvard University Press, 1981). Also see Michael J. Trebilcock, The Limits of Freedom of Contract (Cambridge: Harvard University Press, 1993). 5 The classic defense of this view is Richard Posner, Economic Analysis of Law (Boston: Little Brown, 1977). 6 See Jules Coleman, Douglas D. Heckathorn, and Steven Maser, "A Bargaining Theory Approach to Default Provisions and Disclosure Rules in Contract Law/' in R. G. Frey and Christopher Morris, eds., Liability and Responsibility (New York: Cambridge University Press, 1991).

UNCONSCIONABLE CONTRACTS

39

their agreement freely and with (relatively) full information. From a consequentialist perspective, we might say that the substance of the contract is morally legitimate if (minimally) it is mutually advantageous and (less minimally) if it is fair to the parties, or at least not too unfair. The principal common law defenses to a contract—duress, fraud, misrepresentation, incapacity, mistake—refer to defects in the process of contract formation rather than in the terms of the contract itself. If B is to invalidate a contract with A, B typically claims that there was some cognitive or volitional defect in the formation of the contract—that there was not actual and voluntary assent, notwithstanding outward manifestations of assent, such as a valid signature. For just as freedom of contract requires that persons be able to assume obligations if they do so freely, it requires that contractual obligations not be imposed unless they were assumed freely.7 Interestingly, the utilitarian view of contract law emphasizes the same procedural dimensions. Gifts and deliberate self-sacrifice aside, people will not freely enter into an agreement unless they expect it to be advantageous to them. On this view, a contract procured through coercion or fraud should be nullified not because it is nonconsensual, but because coercion and fraud negate the presumption that the contract is beneficial to B.8 The importance of procedural as contrasted with substantive defects can be seen in another way. On the standard contemporary view, a valid contract requires (1) parties with capacity, (2) manifested assent, and (3) consideration.9 It might be thought that whereas "capacity" and "assent" refer to the sorts of procedural criteria discussed above, "consideration" has to do with the substance or fairness of the 7

As Duncan Kennedy has remarked, "[T]he exceptions to the enforcement of agreements made for cases of fraud, duress, and incapacity are constitutive of the model of free contract." "Distributive and Paternalist Motives in Contract and Tort Law, with Special Reference to Compulsory Terms and Unequal Bargaining Power," 41 Maryland Law Review 563 (1982), at 577. 8 Indeed, Posner claims that contracts made in response to coercive threats are made quite voluntarily. Economic Analysis of Law, p. 49. Also see Daniel T. Ostas, "Predicting Unconscionability Decisions: An Economic Model and an Empirical Test," 29 American Business Law Journal 535. 9 Arthur Leff, "Unconscionability and the Code—The Emperor's New Clause," 115 University of Pennsylvania Law Review 485 (1967), 486. These are necessary but not sufficient requirements of a valid contract. An agreement that meets these criteria may be unenforceable on paternalistic grounds, as when a tenant is not permitted to waive a warranty of habitability, because it threatens to harm third parties, or because society has decided that certain goods should not be "commodified"—for example, votes, drugs, or sex. More generally, an agreement that is against "public policy" is not enforceable, even though it may meet the tests of capacity, assent, and consideration.

40

CHAPTER TWO

agreement. But that would be a mistake. The standard view is that consideration must flow between the parties to establish that an agreement or exchange has occurred (rather than a one-sided promise), but the adequacy or amount of the consideration is generally thought to be irrelevant. In the famous words of a nineteenth-century decision, "A cent or a peppercorn, in legal estimation, would constitute consideration." The terms of a contract may be harsh, but the court "mends no man's bargain."10 Of course even if current legal doctrine refuses to look at relative value, there is no doubt that we can and do ask whether a contract represents a fair exchange of value in evaluating the moral quality of an agreement. As Michael Sandel has observed, we can always ask about a voluntary agreement, "But is it fair, what they have [freely] agreed to?"11 Moreover, it has been maintained that when we examine what judges do as distinguished from what they sometimes say, we will find that "relief from contractual obligations is frequently given on the ground of unfairness."12 For example, in a famous maritime case, The Port Caledonia and the Anna, a vessel in difficulty asked for assistance from a nearby tug. The tugmaster/s terms were £1,000 or no rope. The master of the vessel agreed to pay the £1,000. In refusing to uphold the agreement (the court awarded the tugmaster £200 for his efforts) the court asked "whether the bargain that was made was so inequitable, so unjust, and so unreasonable that [the court] cannot allow it to stand."13 Although we shall consider the notion of substantive fairness in more detail, here I want to raise two preliminary concerns. First, that it seems meaningful to ask, "But is it fair, what they have [freely] agreed to?" does not show that fairness is an independent moral criterion. What is sometimes referred to as "the open question test" may 10 Lord Nottingham, in Maynard v. Moseley, 3 Swans, 651, at 655, 36 E.R. 1009 (H.L. 1676), cited in S. M. Waddams, The Law of Contracts, 2d ed. (Toronto: Canada Law Books, 1984), p. 326. 11 Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982) p. 106. 12 Waddams, Law of Contracts, p. 326. 13 The Port Caledonia and the Anna (1903), quoted in S. M. Waddams "Unconscionability in Contracts," 39 Modern Law Review 369 (1976), at 385. Also see Lord Goff of Chieveley and Gareth Jones, The Law of Restitution, 3d ed., (London: Sweet and Maxwell, 1986), p. 267. Or, as Justice Frankfurter put it, "[T]here is no principle more familiar or more firmly embedded in the history of Anglo-American law than the basic doctrine that the courts will not permit themselves to be used as instruments of inequity and injustice." U.S. v. Bethlehem Steel, 315 U.S. 289 (1938). The government had made its principal argument on grounds of duress. The Court held that the government's will had not been "overborn" and that it was implausible to believe that a corporation could overpower the government of the United States.

UNCONSCIONABLE CONTRACTS

41

show less than meets the eye. As Jeffrey Reiman has observed, it seems meaningful to ask, "What is the highest number?" but that does not show that there is a highest number or that the question is, in fact, meaningful.14 Similarly, although we can ask whether the results of a fair lottery are just, the result of a fair lottery must be just if a lottery is an appropriate decision-making device in a particular context. A lottery may exemplify what Rawls calls "pure procedural justice," a situation in which a fair process guarantees a just result but in which there is no independent criterion of a just result.15 It is possible that contracts are also cases of pure procedural justice, that whatever the parties freely agree to is just. Second, there is a distinction between the moral weight and the moral force of unfairness in contracts. If we assume, for the sake of argument, that it is possible to evaluate contracts in terms of fairness, whether an unfair but consensual agreement should be enforced is another question. Although the autonomy view of contract may often be associated with a pure procedural justice view of contract, that view need not assume that the terms of a contract are fair just in case they are consensual. The autonomy view could maintain that one who freely enters into an unfair contract should be held to its terms, its unfairness notwithstanding. So if, as I believe, a contract can be evaluated in terms of substantive fairness, we must still show why and when that unfairness should be grounds for nonenforcement. EARLY COMMON LAW

In his classic work on contract law, Patrick Atiyah maintains that early English law was relatively unconcerned with process or voluntariness.16 Prior to the nineteenth century, it is said, the governing principle of contracts was some conception of fairness or reciprocity, as exemplified by the Roman law principle laesio ultra dinidium vel enormis, under which a contract could be avoided if a party received less than one-half the normal market value of what was exchanged. On one version of its history, unconscionability has its roots in an earlier era of the law of contracts, and, in particular, in the "medieval notion" of a "just price." It is often supposed that the modern conception of value, in which value is determined by markets, can 14

Justice and Modern Moral Philosophy (New Haven: Yale University Press, 1990), p. 202. 15 John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971), pp. 85ff. 16 See P. S. Atiyah, The Rise and Fall of Freedom of Contract (Oxford: Oxford University Press, 1979), p. 435.

42

CHAPTER TWO

be contrasted with an earlier regime in which value was thought to be objective. On closer examination, however, it turns out that the "medieval" notion of a just price is less objective and much less hostile to market forces than is often supposed.17 Aquinas himself allows that a fair price could reflect differences in costs of production, transportation, and risk.18 It seems that the notion of a just price did not correspond to a market independent conception of value. Rather, the just price corresponded to the price that was set by a "fairly conducted m a r k e t . . . in the absence of fraud, monopoly, or coercion" and allowed public authorities to impose that price in cases of collusion or emergency.19 On this view, the laesio principle accepts the normal market (bazaar)defined price as a baseline, and reflects the not unreasonable view that if a party receives less than one-half the normal market value, it is likely that there was some procedural defect in the transaction. In any case, despite frequent references to an earlier era in which considerations of fairness were supposed to be more important, there is no evidence that courts frequently invalidated agreements simply on the ground that they were substantively unfair. Equity courts were prepared to apply a doctrine of unconscionability, but the traditional equity case involved a claim that one party took advantage of some special weakness of the other party as contrasted with a more impersonal imbalance of economic power.20 In the words of Chesterfield v. Johnson, an eighteenth-century English decision, an unconscionable contract is one that "no man in his senses and not under delusion would make on the one hand, and as no honest and fair man would accept on the other."21 17

See A.W.B. Simpson, "The Horwitz Thesis and the History of Contracts/' 46 University of Chicago Law Review 533 (1979). 18 "The price of an article is changed according to difference in location, time, or risk to which one is exposed in carrying it from one place to another or in causing it to be carried. Neither purchase nor sale according to this principle is unjust." Summa Theologica, 2a, 2ae, question 77. Cited in Bernard W. Dempsey, "Just Price in a Functional Economy," 25 American Economic Review 471 (1935). 19 Raymond de Roover, "The Concept of the Just Price: Theory and Economic Policy," 18 Journal of Economic History 418 (1958), at 420. 20 In addition, equity cases typically involved the transfer of land rather than a fungible good, such as a television set and also involved a claim of specific performance. Unlike many modern cases of unconscionability, equity cases did not generally involve a dispute about the price of a transaction nor an effort to have the transaction go through on more favorable terms. Rather, the promisee demanded transfer of the land and the promisor sought to overturn the transaction in its entirety. See Leff, "Unconscionability," pp. 533ff. 21 28 Eng. Rep 82 (P.C. 1750), cited in Richard J. Hunter, Jr. "Unconscionability Revisited: A Comparative Approach," 68 North Dakota Law Review 145 (1992), at 147.

UNCONSCIONABLE CONTRACTS

43

Chesterfield refers to two different criteria of an unconscionable contract—irrationality (on B's part) and unscrupulousness (on A's part). An unconscionable contract is one that only a person (B) under a "delusion" would make and that "no honest and fair man" (A) would insist upon. But what if these two criteria diverge? It is possible, after all, that A finds B in very difficult straits, that B is under no delusion about what is in his interests, and that A, who is neither honest nor fair, takes advantage of B's situation by getting B to agree to a bargain which is "inequitable" but advantageous to B. If we were to say that a contract is valid if B's judgment is not distorted, then unconscionability would not apply to the inequitable but Paretosuperior contracts we found in rescue cases such as The Port Caledonia and the Anna. Finally, it seems that unconscionability in equity always involved wrongdoing by A in the bargaining process. The bargaining processes typically involved either an abuse of fiduciary relationships or "individual cases of overreaching [which] while not quite adding up to fraud or duress, formed the picture of bargaining processes which the chancellors declared 'unconscionable.'" 22 But bargaining there was. By contrast, contemporary courts often find unconscionability in arm's-length transactions or "standard form" contracts, which are allegedly unconscionable not because there is an abuse of bargaining, but because there is no bargaining at all. THE UNIFORM COMMERCIAL CODE

The unconscionability doctrine was incorporated into statutory law in Section 2-302 of the Uniform Commercial Code, "Unconscionable Contract or Clause": (1) If the court as a matter of law finds the contract or any clause of the contract to have been unconscionable at the time it was made the court may refuse to enforce the contract, or it may enforce the remainder of the contract without the unconscionable clause, or it may so limit the application of any unconscionable clause as to avoid any unconscionable result. (2) When it is claimed or appears to the court that the contract or any clause thereof may be unconscionable the parties shall be afforded a reasonable opportunity to present evidence as to its commercial setting, purpose and effect to aid the court in making the determination.23 22

Leff, "Unconscionability/' p. 537. Uniform Commercial Code Commentary and Law Digest (Boston: T h o m a s M. Q u i n n , Warren, Gorham and Lamont, 1978), pp. 2-94. 23

44

CHAPTER TWO

By adopting the U.C.C., state legislatures gave courts a specific statutory mandate to invalidate unconscionable contracts in commercial contexts. It is, however, much less clear what that meant.24 The text of Section 2-302 strongly suggests that unconscionability is substantive rather than procedural. Paragraph (1) refers to the contract itself, not its procedural history. It also says that a particular clause of the contract might be excised or limited so as to avoid an "unconscionable result." Or consider the language in Paragraph (2), to the effect that a contract that appears to be unconscionable may not in fact be unconscionable, once all the background economic factors are taken into account. It is plausible to suppose that a contractual term that appears to assign an undue share of risk to one party could appear more reasonable once the commercial setting is better understood. It is less plausible to suppose that a procedural defect that compromises consent could be made to appear more reasonable in the light of evidence about the "commercial setting." The language of the provision notwithstanding, the official commentary on Section 2-302 says that the doctrine of unconscionability is "designed to prevent oppression and unfair surprise . . . and not the disturbance of risks because of superior bargaining power." I confess that I do not understand how the authors understand "oppression" in this context, but the wording implies that if the stronger party has not used its strength to secure a contract that the weaker party does not fully understand, then the doctrine of unconscionability does not apply merely because the terms of a contract are "unfair." CONTEMPORARY CASES

Let us put doctrinal statements to the side and consider the way in which unconscionability has been understood in some (relatively) contemporary American and British cases that have, I should say, been chosen not because they are representative of the class of unconscionability cases but because they contain situations or arguments that provide grist for our theoretical mill. Henningsen v. Bloomfield Motors25

Mr. Henningsen purchased a Plymouth, manufactured by Chrysler Corporation, from Bloomfield Motors. Ten days later, Mrs. Henningsen was injured while driving the car. According to an insurance 24 It seems that unconscionability is defined as "a matter of law" rather than a matter of fact in order to make it easier to appeal a lower court's decision. 25 32 N J . 358, 161 A.2d 69 (1960).

UNCONSCIONABLE CONTRACTS

45

adjuster, something went "wrong from the steering wheel down to the front wheels." Henningsen sued for damages on grounds of negligence and breach of express and implied warranties. The trial court dismissed the negligence counts, but an award was granted on the ground of implied warranty of merchantability. Bloomfield Motors and Chrysler appealed. They argued that by signing the purchase order, which contained, in fine print, the uniform warranty of the Automobile Manufacturers Association, Henningsen had absolved Chrysler and Bloomfield Motors of any responsibility for the injury.26 In finding for Henningsen, the New Jersey Supreme Court advanced two different reasons for refusing to apply the principles of freedom of contract "on a strict, doctrinal basis."27 First, because the automobile has the potential for severe harm, the public has an interest in the nature of purchase agreements. On this argument, considerations of public policy override freedom of contract. The court's second line of argument can be understood as an attempt to interpret the principle of freedom of contract rather than as a reason to override it. And here the court focused on the fact that the warranty disclaimer was contained in a standard form. According to the court, "freedom of contract" is most at home when a contract is the result of "free bargaining of parties . . . who meet each other on a footing of approximate economic equality" rather than the "standardized mass contract. . . used primarily by enterprises with strong bargaining power and position."28 There were two factors about the use of "standard forms" that gave pause to the court, and it is important to distinguish between them. First, Henningsen was likely not to have read or understood the relevant parts of the purchase agreement, which appeared in very fine print on the back of the form. Second, the standard form was pre26

That section of the standard form reads, in part: "It is expressly agreed that there are no warranties, express or implied, made by either the dealer or the manufacturer on the motor vehicle. . . . The manufacturer warrants each new motor vehicle . . . to be free from defects in material or workmanship under normal use and service. Its obligation under this warranty being limited to making good . . . any part or parts . . . this warranty being expressly in lieu of all other warranties expressed or implied, and all other obligations or liabilities on its part. . ." Quoted in Henningson, 161 A.2d at 74. I am ignoring other issues that were raised in the case. For example, Chrysler claimed that it had dealt only with Bloomfield Motors and that Henningsen had no claim against the manufacturer. 27 "In the absence of fraud, one w h o does not choose to read a contract before signing it, cannot later relieve himself of its b u r d e n s . " Ibid., at 84. 28 Ibid., at 86. The court approvingly invoked the phrase "contracts of adhesion" because the provisions are not subject to discussion and are "adhered to" by the "adherer." The locus classicus is Friedrich Kessler, "Contracts of Adhesion—Some Thoughts about Freedom of Contract," 43 Columbia Law Review 629 (1943).

46

CHAPTER TWO

sented to Henningsen on a take-it-or-leave-it basis, and the same standard warranty was used by all domestic manufacturers, at a time where the "Big Three" (General Motors, Ford, and Chrysler) accounted for 93.5 percent of the cars sold in the United States. The court commented: The gross inequality of bargaining position occupied by the consumer in the automobile industry is thus apparent. There is no competition among the car makers in the area of the express warranty. Where can the buyer go to negotiate for better protection? Such control and limitation of his remedies are inimical to the public welfare and, at the very least, call for great care by the courts to avoid injustice through application of strict common-law principles of freedom of contract.29 On this view Henningsen's contract might have been invalid even if he had read and understood the relevant clause because it was not the result of a bargain between (relatively) equal parties. Williams v. Walker-Thomas Furniture Co.30

Ms. Williams, who was a welfare recipient, purchased a number of items from Walker-Thomas, an appliance and furniture dealer, under an installment agreement that contained an "add-on" provision giving Walker-Thomas the right to repossess all items purchased from Walker-Thomas if all items were not paid off.31 In 1962, Williams bought a stereo set with a stated value of $514.95. She owed $164 on prior purchases. When she defaulted, Walker-Thomas sought to repossess all the items purchased since 1957. Williams claimed that the contract was unconscionable. The district court of appeals condemned "the sharp practice and irresponsible business dealings" that resulted in such "exploitative contracts," but believed that it had no legal basis for voiding the contract, because the U.C.C. had not been adopted in the District of Columbia.32 On appeal, the U.S. Court of Appeals for the District of Columbia Circuit held that it could apply a U.C.C.-type rule and that "where the element of unconscionability is present at the time a contract is made, the contract should not be enforced."33 An unconscionable contract, said the court, includes "an 29

Ibid., at 87. 198 A.2d 914 (D.C. 1964), rev'd, 350 F.2d 445 (D.C. Cir. 1965) 31 The provision stated: "All payments now and hereafter made by [purchaser] shall be credited pro rata on all outstanding leases, bills and accounts due the Company by [purchaser] at the time each such payment is made/7 Quoted in Williams, 350 F.2d at 447. 32 Williams, 198 A.2d at 914, 916 (1964). 33 Williams, 350 R 2 d 445, at 449. 30

UNCONSCIONABLE CONTRACTS

47

absence of meaningful choice on the part of one of the parties together with contract terms which are unreasonably favorable to the other party/' 34 Toker v. Perl35 An agent of People's Foods of New Jersey visited Perl and described a plan under which Perl would purchase eighteen weeks' worth of food at a time. When Perl mentioned that he had no freezer to store such a quantity, the agent replied that the freezer was included in the food plan. The agent presented three forms for Perl to sign, placing one on top of the other. The top page was the food plan contract, and only the signature line was visible on the lower two forms—a contract to purchase a freezer and an installment contract on the freezer. The freezer was worth about $300. The installment contract provided for a purchase price of $799.95. When interest and other charges were added in, the total came to $1,092.96. The next day, when Perl discovered that he had signed a financing application and an installment contract for a freezer, he tried to cancel the contract, but the seller would not agree. In the ensuing lawsuit, the court nullified the contract on grounds of fraud, but indicated that it would have been prepared to do so on grounds of price alone:"[T]he exorbitant price of the freezer makes this contract unconscionable and therefore unenforceable." Gianni v. Gantos36

In June 1980, Gantos, a clothing retailer, submitted to Gianni, a clothing manufacturer, an order for women's holiday clothing, to be delivered on October 10, 1980. The back of the purchase order contained a cancellation clause under which Gantos could cancel theorder at any time prior to shipping.37 In late September 1980, and before any clothing had been shipped, Gantos canceled the order. Faced with the prospect of holding inventory that it could not sell, 34 Ibid., at 450, q u o t i n g 1 Corbin, Contracts § 128 (1963). A l t h o u g h the court did n o t claim that the contract was unconscionable in light of these criteria, it concluded that the district court had a legal basis for considering the question. Judge Danaher, dissenting, maintained that Williams knew just where she stood, that the pricing and credit policies may have been reasonably consistent with the risk of default. Ibid., at 450. 35 103 N J . Super. 500, 274 A.2d 78 (1970). 36 151 Mich. A p p . 598, 391 N.W.2d 760 (Mich. Ct. A p p . 1986). 37 "Buyer reserves t h e right to terminate b y notice to Seller all or a n y part of this Purchase Order with respect to Goods that have not actually been shipped by Seller or as to Goods which are not timely delivered for any reason whatsoever."

48

CHAPTER TWO

Gianni subsequently agreed to a 50 percent price reduction if Gantos would accept the goods. The agreement was held invalid on the grounds that the cancellation clause was substantively unconscionable and that it resulted from an inequality in bargaining power.38 The cancellation clause was held to be unreasonable because Gianni had made the goods especially for Gantos and, given the fast-changing character of the fashion business, Gianni could not put the goods on the shelf to await another order. The court rejected Gantos's claim that there was no inequality of bargaining power since both parties were experienced businessmen and therefore not unequal. It stressed that the "holiday order" comprised about 20 percent of Gianni's annual business and Gantos's sales were twenty times those of Gianni and the testimony of a buyer for Gantos, who said that such clauses were standard in the business because "the buyer in our industry is in the driver's seat."39 Lloyds Bank v. Bundy40

Herbert Bundy, an elderly farmer, and his only son, Michael, had been customers of a branch of Lloyds Bank for many years. Michael had formed a company, which banked at the same branch; Michael's company was not doing well, and on several occasions Herbert Bundy gave the bank additional security for his son's overdrafts. In May 1969, Michael went to see Herbert, accompanied by the bank's assistant manager, who suggested that Herbert sign a further guarantee for £5,000. Herbert Bundy consulted his solicitor, who advised him that he should put up no more than £5,000 in collateral for his son's business (about half of his assets—the house in which he lived). He followed his solicitor's advice. Michael's business continued to falter. In December 1969, the son visited the father, accompanied by Mr. Head, a new assistant manager of the bank. Mr. Head told Herbert that the bank would continue to support his son only if he, Herbert, secured the son's overdrafts with all of his remaining assets—his farm—but that the bank would not offer any additional line of credit. Without seeking further advice, he signed the papers. When the son's business failed, the bank sought possession of the farm. The Court of Appeal ruled that the contract was unenforceable. 38

The court approvingly cited Allen v. Michigan Bell Telephone Co., 18 Mich. A p p . 632, 637, 171 N.W.2d 689 (Mich. Ct. A p p . 1969). Allen lost business because t h e p h o n e company neglected to publish its advertisement in the yellow pages. It contested a clause in the contract that immunized the phone company from liability for damages arising from the lost business. 39 391 N.W.2d 760, at 762. 40 3 W.L.R. 501 (C.A. 1974).

UNCONSCIONABLE CONTRACTS

49

Speaking for the court, Lord Denning maintained that, in most cases, a customer who signs a guarantee cannot get out of it: "No bargain will be upset which is the result of the ordinary interplay of forces/'41 Indeed, he stated, this rule applies to many "hard cases," as when a homeless person agrees to pay a high rent to a landlord. Yet, he continued, there are a variety of situations in which courts have traditionally refused to uphold agreements. The jurisprudential challenge was to "find a principle to unite them," a principle that he found in the notion of the inequality of bargaining power. Denning argued that if we look below the surface of cases that involve claims such as "duress of goods," "unconscionability," "undue influence," "undue pressure," and "salvage agreements," we will find that courts have set aside contracts "when the parties have not met on equal terms—when the one is so strong in bargaining power and the other so weak—that, as a matter of common fairness, it is not right that the strong should be allowed to push the weak to the wall."42 Applying this principle to the present case, Lord Denning maintained that the terms were unfair because neither the father nor the son received adequate consideration—the son's company was not given any additional line of credit or even a promise to continue the overdraft. In addition, the formation of the contract was defective because the father's judgment was impaired by his love for his son, because Herbert did not and was not urged to consult an independent advisor, and because there was a conflict of interest between the bank and the father that the bank failed to acknowledge given that the relationship between the bank and the father was one of trust and confidence. Macaulay v. Schroeder Music Publishing Co. 43

Macaulay, an unknown young songwriter, entered into a standard form agreement with Schroeder, whereby Schroeder acquired the right to publish all songs written by Macaulay for a five-year period and the 41

Ibid., at 506. Ibid. "By virtue of [this principle], the English law gives relief to one w h o , without independent advice, enters into a contract upon terms which are very unfair or transfers property for a consideration which is grossly inadequate, when his bargaining power is grievously impaired by reason of his own needs or desires, or by his own ignorance or infirmity, coupled with undue influences or pressures brought to bear on him by or for the benefit of the other." Ibid., at 509. Lord Denning maintained that "undue" influence does not require proof of intentional wrongdoing; the party may be moved solely by self-interest and/or unaware of the effects on the other party. Moreover, it is by no means required that the will of the weaker party be "overcome" by the stronger party. The contract may meet the traditional standards of voluntariness, yet still be the product of undue influence. 43 1 W.L.R. 1308 (H.L. 1974). 42

50

CHAPTER TWO

agreement would be automatically extended for another five years if Macaulay earned royalties exceeding £5,000 during the five-year period. Schroeder was under no obligation to publish any of Macaulay's works and could terminate the contract at any time. When Macaulay earned the requisite loyalties and Schroeder exercised its option, Macaulay sought to have the contract nullified. Acting as a court of appeals, the House of Lords ruled that the contract was void as an unreasonable restraint of trade. The court's decision rested primarily on two factors: (1) the terms of the contract were unreasonably asymmetrical; and (2) the terms of the contract were not reached through negotiation or bargaining, but were contained in a "standard form."44 Lord Diplock (concurring) argued that his colleagues had, in effect, invoked the principle of "inequality of bargaining power," although they had not framed their reasons in these terms: It is, in my view, salutary to acknowledge that in refusing to enforce provisions of a contract whereby one party agrees for the benefit of the other party to exploit or to refrain from exploiting his own earning power, the public policy which the court is implementing is not some 19th-century economic theory about the benefit to the general public of freedom of trade, but the protection of those whose bargaining power is weak against being forced by those whose bargaining power is stronger to enter into bargains that are unconscionable.45 Diplock was not attacking all standard form contracts. He observed that standard form contracts were perfectly benign when the standard clauses were the evolutionary outgrowth of years of negotiation and "have been widely adopted because experience has shown that they facilitate the conduct of trade . . . by parties whose bargaining power is fairly matched."46 By contrast, when standard forms are "the result of the concentration of particular kinds of business in relatively few hands" so that the stronger party can say. "These are my terms—take it or leave it," they should be treated with suspicion.47 PROCESS AND/OR RESULT?

So much for background. Let us consider whether unconscionability is best understood as a defect in process or result or both. Joel Feinberg says both: 44 Actually, the court noted that the agreement had been signed "with a few alterations''—implying that at least some negotiation had, in fact, occurred. Ibid., at 1309. 45 46 47

Ibid., at 1315. Ibid., at 1316. Ibid.

UNCONSCIONABLE CONTRACTS

51

It seems clear from the legal doctrine as so far developed that a contract is "unconscionable," hence invalid, when and only when: 1. it is either coercive or deceptive . . . and 2. the terms are harsh to the weaker party because of provisions that "shock the conscience" or "wreak with oppression" and "unfair surprise" . . . and 3. the terms are unequal involving disproportionate benefit, or "excessive profit" to the stronger party.48 On Feinberg's view, three conditions are independently necessary and jointly sufficient to establish unconscionability. The point of condition 1 is to insist that an unconscionable contract must involve "procedural unconscionability." Conditions 2 and 3 both involve what might be called "substantive unconscionability." The terms of the agreement must be harsh to B and disproportionately beneficial to A. There is both textual and conceptual support for the centrality of substantive unconscionability, although it is not clear whether both harshness and disproportionality are both necessary to substantive unconscionability, or whether they can actually be distinguished. If we consider the relevant statute, we note that U.C.C. Section 2-302 refers to the contract or a clause of the contract as unconscionable and makes no explicit mention of the process of contract formation. If we consider the cases described above (and they were not selected to support this point!), we note that all make reference to a problem in the terms of the contract: an "unconscionable purchase price";49 a liability waiver that is "inimical to the public welfare";50 "contract terms which are unreasonably favorable to the other party";51 "the exorbitant price of the freezer";52 the reasonableness of a termination clause ("If a termination clause appears reasonable . . . disparity in bargaining power . . . will not make the clause unenforceable");53 "terms which are very unfair";54 "bargains that are unconscionable."55 Cases aside, there are good theoretical reasons for regarding a defect in result as necessary to unconscionability. After all, we can well imagine a relevant defect in process, such as duress or fraud, that would not entail unconscionability. Suppose, for example, that B 48 Joel Feinberg, Harm to Self (New York: Oxford University Press, 1986), p p . 251-52 (emphasis in original). Feinberg adds that on a "narrower usage/' harshness in result is the only criterion for unconscionability. 49 Murphy v. McNamara, supra note 1, at 177. 50 Henningsen v. Bloomfield Motors, supra note 25, at 87. 51 Williams v. Walker-Thomas, supra note 30, at 450. 52 Joker v. Perl, supra note 35, at 80. 53 Gianni v. Gantos, supra note 36, at 762. 54 Lloyds Bank v. Bundy, supra note 40, at 509. 55 Macaulay v. Schroeder Music Publishing Co., supra note 43 at 1315.

52

CHAPTER TWO

owes A money but has refused to pay. A drafts a contract in which B agrees to pay what B owes A and says to B, "Sign the contract or I'll shoot you." The contract would be invalid on grounds of duress, but we would not likely regard it as an unconscionable contract.56 PROCEDURAL UNCONSCIONABILITY

If a defect in result is necessary for unconscionability, is it sufficient? One thing, I think, is clear. If unconscionability involves a defect in process, it is not a straightforward defect in consent, such as duress or fraud. There are three reasons to adopt this view. First, although the standard procedural defenses to contracts constitute reasons to think that B's agreement was not in B's interests, at least some contracts that have been held unconscionable appear to be Pareto superior. Second, if unconscionable contracts could be invalidated on standard procedural grounds, such as duress or fraud, there would be no need to develop a principle of unconscionability, although it might serve as a proxy for the standard defenses in cases in which it was difficult to produce direct evidence rather than as an independent nonevidentiary defense to a contract. Third, there is no reason to think that consent is, in fact, compromised in many cases of unconscionability. Let us first consider duress. Joel Feinberg argues that Henningsen did not voluntarily agree to the terms of the contract because he signed the contract under duress. And it is a case of duress because "the weaker party has no reasonable alternative to the terms offered by the stronger party and is thus forced to choose what to him appears the lesser evil/' 57 Norman Bowie argues that the contract was "coercive in nature" because "all auto manufacturers had virtually identical warranties . . . Henningsen . . . had no alternative but to take a warranty of that type." 58 I disagree. I do not deny that Henningsen may have had no reasonable alternative but to accept these terms. I do want to suggest that "having no reasonable alternative" has never been definitive of coercion or duress in the law. We do not say, for example, that a wage contract is nonconsensual just because B has to work and has no better alternatives. We do not say that B's purchase of food from the supermarket is nonconsensual because B needs to eat and there are 56 As Joel Feinberg puts it, "Even w h e r e there is both coercive intent in A a n d coercive effect o n B, the offer will not properly be called "exploitative" unless its terms are either harsh in their costs to B, or u n e v e n or disproportionate in their gains for A, or b o t h . " Harm to Self, p . 249. 57 Ibid., p p . 251-52. 58 "Fair Markets," 7 Journal of Business Ethics 89-98 (1988), at 96.

UNCONSCIONABLE CONTRACTS

53

no better prices available. And we do not say that a patient's consent to allow a surgeon to perform an operation is invalid because the patient wants to live and the only alternative is death. It is true that when the robber brandishes a gun and says "Sign this contract or your life," the victim may have no reasonable alternative to signing the contract. And that may have misled some into thinking that having no reasonable alternative is definitive of duress. But that is not why the contract with the robber is made under duress. It is made under duress because A proposes to violate B's rights or to harm B if B refuses to sign, and that feature is absent in the cases in which B also has no reasonable alternative but which are not cases of duress.59 Consider Murphy, Gianni, and Macaulay. Even if Murphy could not have purchased a television set on more favorable terms, say, because she was a poor credit risk, we can hardly say that she signed under duress or that McNamara coerced her into signing the contract, and she did not claim that he had. McNamara may have enticed Murphy into signing the contract. He may even have misled her. But I do not think that he coerced her, for he did not propose to harm Murphy or violate her rights if she did not accept his terms.60 In Gianni, a manufacturer agreed to allegedly harsh terms when it might have preferred softer terms. Was it coerced? No, because Gantos was not proposing to harm Gianni if Gianni did not accept the contract on Gantos's terms. And the same was true in Macaulay. To say "These are my terms—take it or leave it" may sometimes be objectionable, but it hardly seems coercive in the absence of other factors. Contrast these cases with The Port Caledonia. We might say that the vessel agreed to pay £1,000 for the rescue under duress, not simply because it had no reasonable alternative, but because the tugmaster had an obligation to rescue the vessel on better terms, that it was not morally or legally free to sail away. In effect, and like the robber, the tug threatened to do something wrong if its terms were not accepted. But in seeing why this case might constitute coercion, we see why the other cases do not. For it seems preposterous to say that McNamara, Gantos, and Schroeder were not free not to enter into deals with their respective offerees. If unconscionability need not involve straightforward coercion or duress, it is also unclear to what extent cognitive defects have played a role. In Joker, the court found the contract invalid for reasons of fraud, but it also stated that it would have found the contract uncon59

See m y Coercion (Princeton: Princeton University Press, 1987), ch. 2. Moreover, even if we were to accept the broader "no reasonable alternative" account of duress, Murphy had a perfectly reasonable alternative: not to buy a television set. 60

54

CHAPTER TWO

scionable on grounds of the purchase price alone. It is not clear whether Williams understood the effect of the "add-on" provisions in her contract with Walker-Thomas, but the decision suggests that this would have been irrelevant in any case. Certainly, there is no reason to think that Gianni did not know about or understand the force of the "terminate-at-any-time" clause or that Macaulay did not understand that Schroeder had the rights to his songs for ten years if he proved successful. In Henningsen, the court noted that the liability limitations were contained in "fine print" and that Henningsen may not have read or understood them. Nonetheless, this does not seem to have been crucial to the case, for the court's decision turned less on the element of "surprise" than on the harmful consequences of this provision and the fact that no better agreements were available in the market. If I am right in arguing that unconscionability does not require a straightforward or standard defect in consent, it does not follow that unconscionability turns on result alone.61 First, unconscionability might involve a nonstandard defect in consent. In an important study of unconscionability decisions, Daniel Ostas found that courts are more likely to find that a contract is unconscionable when they believe that A has a fiduciary responsibility to act in B's interests than when a contract constitutes a commercial arm's-length transaction.62 From this perspective, some retail sales transactions may be held to be unconscionable because courts believe that sellers have a "quasifiduciary" responsibility to buyers.63 And so it may be held unconscionable for A to take advantage of B's trust, not because B does not consent, but because we prefer a legal regime in which A has a paternalistic responsibility with respect to contractual terms over a legal regime based on the principle of caveat emptor. There are two additional ways in which unconscionability might be at least partially procedural. First, when the terms of a contract are clearly unjust or substantially deviate from a normal market price, there may be reason to think that there was a procedural defect even though there is no direct evidence to that effect. On this view, substantive unconscionability is not meant to serve as a distinct defense to a contract, but is principally a proxy for procedural uncons61 [D]espite indications of increasing freedom for courts to manipulate contract terms, the concept of unconscionability still directs judicial inquiry to the bargaining behavior of the parties."Lewis Kornhauser, "Unconscionability in Standard Forms," 64 California Law Review 1151 (1976), at 1162. 62 Ostas, "'Predicting Unconscionability Decisions," p. 549. 63 "Typically, consumers do not closely inspect long and complicated sales contracts. They simply trust that the contracts contain customary terms." Ibid.

UNCONSCIONABLE CONTRACTS

55

cionability in the face of the epistemological problems. Second, it is possible that procedural unconscionability might involve nonstandard defects in consent, such as B's inability to "reason effectively," or bargaining inequality, or taking advantage of B's weakness of will.64 Let us consider both substantive and nonstandard procedural defects in a bit more detail. SUBSTANTIVE UNCONSCIONABILITY

Let us assume that a contract is minimally advantageous to B as compared with the precontractual baseline. As we have seen, it seems that there are two ways in which we might regard the terms of a mutually advantageous contract as unconscionable: (1) harshness— the terms may be particularly harsh for B; and (2) disproportionality —A's benefit from the agreement may be exorbitant or disproportionate to B's gain.65 Harshness

When courts find that a contract is unconscionably harsh, they tend to focus on the nonmonetary terms of the contract. There are exceptions. In The Port Caledonia and the Anna, the court found that £1,000 for a rope was "inequitable and "unreasonable," and in Toker v. Perl, the court found the "exorbitant price" of Perl's freezer "unconscionable." Nonetheless, it seems that most unconscionability cases, or at least the most interesting ones, turn on nonprice contractual terms: limitations on liability, add-on clauses, cancellation clauses, exclusive rights.66 In principle, of course, there can be no rigid distinction between monetary and nonmonetary terms. A contract represents a package of price, quality, financial provisions, and risks. Harsh terms may be compensated for by a relatively generous price. Suppose, for example, that a couple prefers to hire a live-in nanny who will agree not to entertain boyfriends at home.67 If the couple is prepared to pay a 64

John Lawrence Hill, "Exploitation," 79 Cornell Law Review 631 (1994), at 637. Also see Christopher M c M a h o n , Authority and Democracy (Princeton: Princeton University Press, 1994), p. 199. 65 I say "seems" because it may turn out that, on further analysis, absent some defect in process, a result cannot be oppressive to B if A is not gaining significantly from the agreement. 66 But see Frank P. Darr, "Unconscionability a n d Price Fairness," 30 Houston Law Review 1819 (1994). 67 I borrow this example from Brian Barry, "Lady Chatterley's Lover a n d Doctor Fischer's Bomb Party: Liberalism, Pareto Optimality a n d the Problem of Objectionable

56

CHAPTER TWO

wage premium in order to hire on its preferred terms, then there is hardly anything unconscionable with the total package, even if the terms would otherwise be viewed as excessively demanding. Why do unconscionability cases tend to focus on terms rather than (monetary) price?68 First, price is better understood than nonmonetary terms, and so it is difficult to claim that there are cognitive defects. Second, because price is a continuous variable whereas nonmonetary terms have a binary quality, it may be easier to make judgments about terms. It is difficult to say at what point on the continuum a price becomes exorbitant. It is less difficult to say that a cancellation clause is oppressive. Third, although a price may be exorbitant, it does not typically establish an oppressive relation that endures over time. By contrast, Williams was always vulnerable to repossession of all the furniture bought from Walker-Thomas, and Macaulay was tied to Schroeder for ten years. There is another reason why courts may have focused on terms rather than price: there is often more explicit bargaining over price than over terms. Recall Henningsen. Despite the (then) oligopolistic character of the automobile industry among manufacturers, there is no reason to think that dealers did not engage in price competition or that they did not bargain over price, even if they did not bargain over warranty terms. Similarly, even if appliance companies did compete over price, it is possible that they commonly sold their goods to highcredit risk customers with the add-on clause that was found unconscionable in Walker-Thomas, and that there was no bargaining over such terms. And so we may have the appearance of a harsh result. I say we may have the "appearance" of a harsh result because, as we say, appearances can be deceiving. Contrast Henningsen's situation with that of our hypothetical nanny. If the nanny were to complain that the no-boyfriend rule was unconscionably harsh, the couple could point to the wage premium, demonstrating that she had been compensated for agreeing to the rule. No such comparison is possible when all competitors use the same terms. Nonetheless, it is at least possible that automobile buyers have, in effect, been compensated in advance (without their asking to be compensated in this way) for the limited warranty, that Chrysler would have been willing to allow Henningsen to purchase a more inclusive warranty for an added premium. In effect, Henningsen received a price discount for the limited warranty. Preferences," in his collection, Democracy, Power and Justice (Oxford: Clarendon Press, 1989), p. 374. 68 I shall use the term "price" to refer to monetary price in what follows although it is obvious that the terms are part of the overall price of a contract.

UNCONSCIONABLE CONTRACTS

57

Two additional and related points need to be made. First, when we evaluate the terms of a contract, we should focus on the ex ante value of an agreement rather than the ex post value. Suppose that A buys one lottery ticket apiece from each of ten sellers, Bj-B^, for $1,000 per ticket where each ticket gives its holder a .10 chance to win $15,000 (e.g., each ticket allows A to play a game in which A wins if A picks a digit [0-9] that matches a randomly produced digit). Assume that A is prepared to take the risk of losing ten consecutive times (about 30%) and that each seller prefers a certain $1,000 to playing the game. If B/s ticket should prove to be a winner, B1 can hardly claim that the deal was unfair because he or she received only $1,000 for a ticket whose ex post value was $15,000. Consider Macaulay in this light. Macaulay signed on with Schroeder—putting himself at risk for a ten-year commitment— when he was an unknown songwriter. It is entirely possible that a music publisher will make a profit on, say, only 10 percent of its unknown songwriters. In effect, Schroeder buys a lottery ticket from each unknown songwriter. Schroeder is willing to lose money on 90 percent of such clients because it makes a significant profit on the other 10 percent.69 The successful songwriter may feel "ripped off " ex post because the publisher is making a large profit on his songs, but as the lottery example shows, that would be a mistake. The ex ante value of the contract to Schroeder was not particularly high. And this gives rise to the second point, namely, that what appear to be harsh terms may sometimes reflect the risk of dealing with an offeree. Deals with high-risk persons—unknown songwriters, budding restaurateurs, low-income debtors—may appear to be harsher than they really are. (Re)consider the add-on provision of Williams's agreement with Walker-Thomas.70 Given the high probability that Williams would default, given that her purchases were likely to depreciate quickly and that repossession of the sole item in case of default would entail a loss for Walker-Thomas, it is possible that the only way for her to have received more favorable terms would have been for her to pay a higher down payment on the goods involved.711 say it is possible. It is also possible that the terms were harsh even given the risks involved. Whether the contracts fairly reflect the risks of dealing with high-risk offerees is, in principle, an empirical question that can be 69

For a n interesting analysis of this case, see M . J. Trebilcock, "The Doctrine of Inequality of Bargaining Power in t h e H o u s e of Lords," 26 University of Toronto Law journal 359 (1976). 70 See Richard Epstein, "Unconscionability: A Critical Reappraisal," 18 journal of Law and Economics 293 (1975). 71 Ibid., at 307.

58

CHAPTER TWO

resolved, in part, by examining the profit margins in various industries. But if A is not gaining an exorbitant or disproportionate profit from such contracts, it seems wrong to refer to them as particularly harsh, or perhaps more accurately, it seems wrong to claim that A is acting harshly.72 And that brings us to the next criterion. Disproportionality

As we have seen, Feinberg says that a Pareto-superior agreement is exploitative or unconscionable only when A receives an excessive profit or a disproportionate amount of value from the exchange.73 What would constitute proportionate value? Equality is an obvious candidate. It might be thought that a fair transaction is "one in which the surplus is divided (approximately) equally," where each party gains equally from the exchange.74 Several difficulties might be raised with respect to the notion of equality in exchange. One objection can quickly be put aside. Randy Barnett has argued that the notion of equality in exchange is incoherent because both parties to an exchange necessarily perceive the value of the goods to be exchanged as unequal—they would not be willing to trade unless each regarded the other's good or service as more valuable than the good or service being traded for it.75 True as this is, it does no damage to the coherence of equality in exchange, for the question is not whether A and B regard each other's goods as equal to what they have to offer, but whether each gains equally from the exchange in terms of their own preferences or some other criterion. But that still leaves the question of how we should conceptualize equal or disproportionate gain. An old chestnut used by game theorists asks how a rich person and a poor one should agree to share $200: "The rich man could argue for a $150-$50 split in his favor because it would grieve the poor man more to lose $50 than the rich man to lose $150. "76 If we measure a party's gain in terms of utility or preference satisfaction, and if money is subject to the principle of diminishing marginal utility, then the rich man's argument is extremely 72

We may say that the terms are harsh, but that A is not acting harshly if we shift the blame for the harsh terms to society's responsibility for the background conditions. 73 Feinberg, Harm to Self, p . 252. 74 Robert Frank, Passions within Reason (New York: W.W. Norton, 1988), p. 164. 75 Barnett, " A C o n s e n t Theory of Contract," 86 Columbia Law Review 269 (1986), at 284. 76 H o w a r d Raiffa, The Art and Science of Negotiation (Cambridge: Harvard University Press, 1982), p. 52. Raiffa also observes that an arbitrator might suggest the reverse split because the poor person needs the money more and adds that the rich person might also argue for an even split on the ground that it is wrong to mix business with charity.

UNCONSCIONABLE CONTRACTS

59

persuasive. Reconsider The Port Caledonia in this light. There is every reason to think that the vessel gained more utility from being rescued than the tugmaster received from £1,000. If this case exemplifies disproportionate gains, then A's gain must be exorbitant by some independent normative criterion even if A gains less utility than B. We might argue that A's gains are disproportionate not by comparison with B's but by comparison with A's normal profits, costs, or price. It is entirely possible that there are many transactions in which B typically gets more utility than A. The beer drinker may get more utility from a bottle of beer than the retailer gets from the profit. The motorist in need of a boost for a dead battery may get more utility from the aid than the service station receives from the fee. For this reason, we may want to say that A's gain is excessive by comparison with A's normal price for such transactions rather than by comparison with B's gain. Consider, in this light, Henningsen, Gianni, and Macaulay. Despite concerns about the terms of the contracts and inequalities of bargaining power, the courts did not actually inquire into the gains of the allegedly stronger party, and there is simply no evidence that the automobile business or the retail clothes business or the music publishing business generated supracompetitive or exorbitant profits. But what of the cases in which the offeree is poor, such as in McNamara's "rent to own" plan or Williams's contract with Walker-Thomas? I do not know whether these businesses are highly profitable. A priori, there is no reason to think that these contracts generated exorbitant profits. Indeed, if these businesses were highly profitable, we would have to explain why potential competitors allow the McNamaras of the world to garner monopoly or oligopoly profits rather than enter the business themselves and, ex hypothesi, drive the price down. Still, it is possible that these businesses do generate comparatively large profits, if only because potential competitors would prefer not to deal with high-risk customers. But suppose, for the sake of argument, that these businesses do not generate supracompetitive profits. There would then be two possibilities: (1) unconscionability requires supracompetitive profits, and therefore many allegedly unconscionable contracts are not unconscionable; or (2) these contracts are unconscionable, but unconscionability does not require exorbitant or disproportionate profits. STANDARD FORMS

We have seen that allegations of unconscionability have often referred to provisions contained in standard forms. Standard form provisions are thought to be suspicious for several reasons. First, the terms are

60

CHAPTER TWO

not the result of the sort of explicit bargaining (dickering) that is thought by some to be at the core of genuine freedom of contract.77 Second, at least some of the crucial terms may be contained in "fine print," and the offeree may not have read or understood the relevant provisions. Third, there may be no better terms available in the market; all offerors of the relevant product or service may use a similar or identical form. Fourth, the terms themselves are one-sided or oppressive. And fifth, these features of standard form contracts arise from a gross inequality of bargaining power between the offeror and the offeree.78 These objections to standard form contracts are often run together, but they are actually quite distinct. Let us begin by distinguishing between the first and third objections. A particular offeror's terms may be contained in a standard form even though similar offerors use different forms. My university has a standard employment contract for professors. It outlines dates of service, responsibilities, fringe benefits, and so on. To the best of my knowledge, these terms are nonnegotiable. Yet other institutions offer different terms on their standard contracts. That an offeree has no choice with respect to the terms of a particular offeror does not entail that there is no choice in the market. Let us assume, however, that a standard form is industry wide, that the relevant terms are common among all comparable offerors. We then need to consider another distinction. As Lord Diplock observed in Macaulay, there is no necessary relation between standard forms and "inequality of bargaining power."79 Standard form terms may assign a certain risk to one party because it reduces total costs to so assign the risk while compensating the risk bearer through other contractual terms.80 The relevant terms may have evolved in a commercial setting among highly rational and well-informed parties whose economic situation is roughly comparable or in industries that are highly competitive. These standard form contracts must be contrasted with those that are "the result of the concentration of particu77

See Friedrich Kessler, "Contracts of Adhesion—Some Thoughts about Freedom of Contract," 43 Columbia Law Review 629 (1943). 78 Jules Coleman defines adhesive contracts as those that are "negotiated against a background of unequal bargaining power," but that seems wrong. As I shall show, some "standard form" contracts are used when inequality is not an issue, and the stronger party in a situation of great inequality of bargaining power may not make use of standard forms. See Jules Coleman, Markets, Morals and the Law (New York: Cambridge University Press, 1988), p. 208. 79 Supra n o t e 43, at 1316. 80 For example, automobile rental companies do not reimburse for gas, but they do reimburse for oil because the absence of oil will hurt the car.

UNCONSCIONABLE CONTRACTS

61

lar kinds of business in relatively few hands" or (and this is a different point) that pit poorly informed offerees against highly knowledgeable offerors. Given this distinction, what precisely was the problem in Macaulay or Henningsen or Williams? In Henningsen and Williams, the offeror

surely knew more about the terms of the contract than did the offeree. Here we have a "fine print" problem. To the extent that standard forms are defective because they give rise to cognitive defects, they are easily remediable, perhaps by requiring offerees to show specific consent to controversial provisions or by printing such provisions in bold type and stating them in non-technical language. But reducing cognitive defects will not eliminate the structural inequality of bargaining power that is said to give rise to the use of "take it or leave it" standard forms in the first place. If inequality of bargaining power is the problem, A could insist that B accept the terms even if B understood precisely what they involve. But is inequality of bargaining power the problem in standard form contracts? Certainly not always. It is simply not true that standard forms are typically used only in oligopolistic industries. Standard forms are often used in highly competitive industries, for example, by dry cleaners, hotels, automobile rental agencies, automobile dealers, music publishers, and furniture and appliance stores.81 Moreover, to say that these are but several counterexamples to the general truth about "take it or leave it" terms is to miss the point, for in a competitive market, virtually all transactions occur on a "take it or leave it" basis. In a truly competitive market, no one gets to negotiate anything.82 Everyone is a "price taker." The seller must offer the equilibrium price, and the buyer must either pay the price or do without. Thus while it may be true that the offeree in a "standard form" situation has no place to go for better terms, there is no reason to think that this is due to a lack of competition, any more than there is to think that all of my local supermarkets charge (roughly) the same "take it or leave it" price because there is no competition.83 81 One study noted that there were about sixty appliance and furniture stores around Third Avenue and 125th Street in East Harlem and that, in addition, there were numerous agents (for stores) and independent peddlers selling door to door. D. Caplowitz, The Poor Pay More (New York: Free Press, 1965), p. 12. 82 Kennedy, "Distributive and Paternalist Motives," p. 616. 83 Somewhat ironically, given Henningsen, the purchase of an automobile is the one consumer transaction in which explicit bargaining is commonplace. Yet when a recent study showed that African Americans and women tend to pay more for new cars than do whites and men, respectively, the author recommended strategies for reducing price dispersion among individual buyers, thereby putting all sales on a "take it or leave it"

62

CHAPTER TWO

How, then, can we explain the use of standard forms in competitive markets? Without claiming that it is the whole story, we can at least partially explain the use of standard forms as a mechanism for reducing transaction costs. To require or even to encourage bargaining in individual transactions would be expensive (and would also, somewhat ironically, give rise to charges of unfair price discrimination that favors better bargainers over less effective bargainers). In a competitive market in which offerors bargain over terms, the offerees would end up paying some if not most of those bargaining costs (they would be included in the price of the good) or they would be driven out of the market by the increased price. If this is so, how can we also explain the apparently onerous terms to which the courts have objected in Henningsen, Macaulay, Williams, and the like? True, onerous terms often appear in fine print legalese, and the offeree may be quite unaware of their content or effect. But this does not help to explain the persistence of such terms in a competitive market where some consumers (or their advocates) are likely to notice such terms and raise concerns about them. And if consumers such as Mr. Henningsen would have preferred a better warranty, albeit perhaps at a higher price, why was it not offered by Bloomfield Motors or another dealer? If the add-on provision in Williams's contract with Walker-Thomas Furniture was, in fact, a bad deal, why was not some competitor exposing this fact and offering a better deal—not out of kindness, but in order to make a buck?84 One possibility is that these markets are not competitive. This may, for example, be part of the explanation in Henningsen. The automobile industry was a mildly competitive oligopoly in the 1950s, and it is not clear that a manufacturer would have reason to "break ranks" and offer a more inclusive warranty for a higher price.85 Second, important market imperfections can remain even in a reasonably competitive market. A perfect market assumes not only competition, but rationality and relatively complete information. And there is no reason to think that either of the latter conditions always hold. Offerees may, for example, be unduly sensitive to price as contrasted with nonprice terms. Even if it would have been better for Williams to reject the addon provisions and pay a higher price, she might have chosen the addbasis. Ian Ayres, "Fair Driving: Gender and Race Discrimination in Retail Car Negotiations," 104 Harvard Law Review 817 (1991). 84 As Duncan Kennedy puts it, "If there is competition among sellers, and good information about buyer preferences, sellers will offer whatever terms they think buyers will pay for." "Distributive and Paternalist Motives," p. 616. 85 See J. G r e g o r y D e e s , " U n c o n s c i o n a b i l i t y a n d F a i r n e s s , " 2 Business Ethics Quarterly 497(1992), at 501.

UNCONSCIONABLE CONTRACTS

63

on/low-price contract. And even if some consumers would prefer soft terms/high-price contracts, there may be too few such consumers to motivate sellers to offer such contracts, given the extra costs generated by offering a variety of terms and price packages. Those who desire products with unique qualities are often out of luck in a mass market. If few consumers want chartreuse cars, manufacturers will not bother to produce them. So, too, with contractual terms. There is, of course, a third possibility, namely, that when all is said and done, the allegedly onerous terms provide the maximum benefit to the offerees, given the risks inherent in contracting with them.86 Consider Macaulay. Suppose that Schroeder offered two options to Macaulay: (1) the standard contract under which Schroeder has the right to Macaulay's songs for ten years if he proves successful, or (2) a contract giving Schroeder the rights to Macaulay's songs only for one year, but requiring Macaulay to pay Schroeder £2,000 up front to defray some of Schroeder's costs should Macaulay's songs prove unsuccessful. The second option reduces Macaulay's obligation to Schroeder, but it requires him to pay for that reduction. If the courts were to prohibit songwriters from choosing the first option on the ground that the terms are too onerous, then they would, in effect, be requiring songwriters to pay more out of pocket or excluding some from entering the business. Consider Williams on the assumption that the price and terms constituted the equilibrium price in a competitive market. Given that the seller of a stereo can reasonably expect to have negative equity in the stereo that it sells to a risky purchaser (its resale value upon repossession is worth less than the outstanding debt), it has four options: (1) it can raise the price of the stereo set; (2) it can require a larger down payment to reduce its risk; (3) it can refuse to sell to risky purchasers; (4) it can expand its equity base under the add-on provision. If the courts claim that the add-on provision in (4) is unconscionable, then they are effectively limiting Williams to (1), (2), and (3)—all of which may be less satisfactory.87 Of course, if there is reason to think that add-on arrangements are not well understood and would be rejected if understood, then there may be reason to hold such terms unconscionable. Indeed, there is some reason to think that it is because such terms were not understood by consumers—not the substance of the terms—that best explains why they have been held to be unconscionable.88 86

Richard Posner, Economic Analysis of Law, p . 85. Richard Epstein, "Unconscionability: A Critical Reappraisal/' 18 Journal of Law and Economics 293 (1975), at 307. 88 See Ostas, "Predicting Unconscionability Decisions," p . 581. Also see Terry Pin87

64

CHAPTER TWO

In any case, our discussion of standard forms highlights a more general point. If the principal aim of the doctrine of unconscionability is to benefit the allegedly exploited party, then we may not "know it when we see it." We need to engage in careful analysis to know if the contractual terms reflect some imperfection in the market that can be corrected or if they are not so harsh given the risks inherent in dealing with a class of offerees. INEQUALITY OF BARGAINING POWER

As we have seen, unconscionable contractual terms are often said to arise from an "inequality of bargaining power." Yet despite the frequency with which this phrase is used, the concept has received precious little analysis as to its defining characteristics or moral significance. What gives A greater bargaining power than B? As Thomas Schelling reminds us, we must be very cautious, because "strength" is not always an asset in bargaining.89 Bracketing such worries, we can begin by distinguishing between what I shall call bargaining ability, which is a function of one's internal characteristics (e.g., information, toughness, patience, perceptiveness, and the like), and what I shall call bargaining potential, which is primarily a function of one's external resources or circumstances. Put colloquially, bargaining ability concerns how well one plays one's cards whereas bargaining potential is a function of the cards themselves. Both bargaining ability and bargaining potential will contribute to a bargaining outcome. Suppose that A is selling a house and B is a potential buyer. A's reservation price is $140,000, whereas B's reservation price is $160,000 (B cannot afford more). If A is a particularly good bargainer, say, particularly good at misrepresenting his or her reservation price, A might get B to pay $159,000. In this case, the sale price is a function of their respective bargaining abilities. Suppose, on kard's discussion of the case in Democratic Liberalism and Social Union (Philadelphia: Temple University Press, 1987). "It is the lack of informed consent that is really what is objectionable in the Williams case." Ibid., p. 120. 89 " 'Bargaining power/ 'bargaining strength/ 'bargaining skill' suggest that the advantage goes to the powerful, the strong, or the skillful. It does, of course, if those qualities are defined to mean only that negotiations are won by those who win. But if the terms imply that it is an advantage to be more intelligent . . . or to have more financial resources, more physical strength . . .or more ability to withstand losses, then the term does a disservice. These qualities are by no means universal advantages in bargaining situations; they often have a contrary value." The Strategy of Conflict (New York: Oxford University Press, 1963), p. 22.

UNCONSCIONABLE CONTRACTS

65

the other hand, that B needs to buy a house quickly, that this is known to both A and B, and each knows that the other knows. Then we might say that A has B "over a barrel," and we would expect a price close to $160,000.90 In this case, the sale price is a function of their respective bargaining potential. Indeed, if A and B and were to agree on $150,000 (halfway between their reservation prices) because B was able to conceal his or her external circumstances or successfully appeal to A's sense of fairness, then we might say that A's superior bargaining potential was offset by B's superior bargaining ability. This suggests that the assessment of "bargaining ability" is dependent on "bargaining potential." We cannot evaluate how well A and B do within the objective situation unless we have some independent standard for how they could be expected to do if their bargaining ability were roughly equal. To introduce some additional stipulative language, let us say that inequalities of bargaining ability can take two forms: bargaining incapacity and bargaining weakness. B has a bargaining incapacity when defects in B's information or judgment lead B to "consent" to contracts which are not, ex ante, in his or her interests. A bargaining incapacity need not be a permanent condition, and it may be a response to A's action rather than an "internal" trait of B. B may be unable to act rationally precisely because A wrongly puts B under stresses that cause B to make an irrational decision. To say that B has a bargaining weakness with respect to A is to say that B will not agree to disadvantageous contracts but may nevertheless do less well than A—given their respective bargaining potential. Some unconscionability cases appear to involve a genuine bargaining incapacity. Herbert Bundy's agreement was set aside because he made an agreement for which he incurred risks but for which he received virtually nothing in return. Under the circumstances, he either lacked the capacity to act in his interests or reasonably believed that he did not have to protect his interests because Lloyds had a fiduciary obligation to do so. To the extent that "inequality of bargaining power" refers to this sort of bargaining incapacity, it is a relatively unproblematic reason for voiding an agreement. It represents the sort of defect in consent that will lead to non-Pareto-superior agreements.91 On the other hand, to the extent that "inequality of bargaining power" represents a problem of bargaining weakness, it is less 90

This is true, of course, precisely because this is not a perfectly competitive market. In a perfect market, with numerous A's selling similar houses at similar prices, there would be no bargaining, and B would not be "over a barrel." 91 I say relatively unproblematic, because it is not clear whether a contract should be nullified if A neither caused nor knowingly took advantage of B's incapacity.

66

CHAPTER TWO

clear what should follow, although we may want to consider policies that work to reduce the effects of those inequalities. But what of those "one-sided" contracts that result from an inequality of bargaining potential? How can we explain Gianni and Macaulay—both of which allegedly involved an inequality of bargaining power? There is no reason to think that they involved inequalities of bargaining ability. In these cases, the courts can be understood as arguing that there were gross asymmetries between the parties with respect to bargaining potential and that these asymmetries led to an unjust result. What constitutes inequality in bargaining potential? Given comparable ability to press one's position within the framework of one's resources, what gives A more bargaining potential than B? I believe that two possible candidates—size and necessity—can be set aside. Although the image of the lonely individual facing the large corporation seems to have figured in several decisions, this model is quite misleading. I do not think that the contemporary (1996) automobile buyer is at a disadvantage in bargaining (through the dealer) with General Motors, nor is it true that almost anyone buying gasoline from any of the major energy companies will find themselves in an enormously weaker bargaining position.92 Size is irrelevant when the "larger" party needs the "smaller" party as much as the other way around. Inequalities of necessity are similarly irrelevant. That B needs food in order to live does not give A (a supermarket) any special power over B if there is a competitive market among sellers of food.93 If B's car battery is dead, but there are numerous sellers of automobile batteries close by, there is no reason to think that any seller has any special power over B. A third candidate looks more promising. It might be thought that the more resources one brings to a bargaining situation, the greater one's bargaining power. But this, too, is incorrect. Consider the following example. The Investment Case. B needs $500 to invest in a venture that will yield $1,000. B is prepared to contribute $400, which is all B has, but B needs $100 from A, who is quite wealthy. 92

Pinkard, Democratic Liberalism and Social Union, p . 119. As Duncan Kennedy remarks, "If there are many sellers of a necessity, none of them will be able to charge more than the going package of price and terms without losing all his buyers." "Distributive and Paternalist Motives," p. 618-19. "Any consumer facing a perfectly competitive market for some necessity . . . has no real choice but to pay the market price; just as the producers have no real choice but to accept that price." Fried, Contract as Promise, p. 104. 93

UNCONSCIONABLE CONTRACTS

67

Suppose that A proposes that the $1,000 be divided thus: A gets $500, B gets $500. This gives A a 400 percent ($400/$100) return on investment, while B gets a 25 percent return ($100/$400) on investment. This may look unfair. It may be unfair. But A can make a credible threat to walk if B does not accept the proposal precisely because a proportional split, where each gets a 100% return on investment (A receives $200 and B receives $800), does not, in fact, do much for A— given A's current position. Although B is contributing more resources, A has more bargaining potential. I suggest that bargaining potential is not a function of size or necessity or resource contribution. Rather, bargaining potential seems to be better understood as a function of one's "threat advantage," where one's threat advantage is one's relative willingness not to contract if one's proposal is not accepted.94 Although the story is no doubt more complicated, we might say that a party's bargaining potential is a function of a party's utility gain from the precontractual baseline. Ceteris paribus, A has a threat advantage over B when A stands to lose less if agreement is not reached. Put slightly differently, although the stronger party may get the greater share of "objective resources" from a proposed bargain (relative to that party's contribution), it is not true that he or she gets more utility from the bargain. To the contrary, it is precisely because A gets less utility from a proposed bargain than B that A is able to get a greater share of objective resources. It is important to note that we will mistakenly evaluate the degree of inequality of bargaining potential if we wrongly assume that a bargaining context between A and B is a one-shot situation. On the one hand, we may overestimate A's bargaining potential if A would jeopardize future gains with B or other potential transactors by using a threat advantage with B. Recall Gianni.95 The court ruled that the cancellation clause was onerous because it enabled Gantos to drive an unfair bargain to buy clothing with Gianni. But it is, after all, quite unlikely that Gantos could frequently use the cancellation clause to get better deals, for manufacturers would either refuse to deal with Gantos or would want to be compensated for the risk of cancellation in the price of the goods. After all, if retailers were consistently able to change the terms of a contract by threatening cancellation, we would need to explain why anyone would set up shop as a clothing manufacturer. Thus there is probably less inequality of bargaining potential here than meets the eye. 94 "[I]n bargaining the less willing always has the advantage." H e n r y Sidgwick, The Methods of Ethics, 7th ed. (1874; 7th ed., 1907; reprint, Chicago: University of Chicago Press, 1962), p. 288. 95 Supra note 36.

68

CHAPTER TWO

On the other hand, we may underestimate A's threat advantage if we focus on the loss to A from not contracting in a single transaction. After all, A has an incentive not to hold out in a one-shot game even if A would lose less than B if they do not reach agreement. But if A expects to be in similar situations in the future, then it is not irrational for A to hold out even if A forgoes a profit in this transaction by doing so. By demonstrating to B (and others) willingness to pass up a profitable deal rather than accept (what he or she regards as) too little, A establishes the credibility of his or her threats for future transactions. Suppose that there is an inequality of bargaining power between A and B. What follows—normatively speaking? Here I want briefly to make four points, although we shall return to many of these issues in Part II. First, we need a moral principle as to what constitutes a just distribution of the social surplus. Unless we can say something about how the social surplus ought to be divided, we cannot say that an agreement is morally one-sided—however unequal the distribution appears. We cannot, for example, simply assume that it is unjust for A to receive a 400 percent return in the Investment Case while B receives a 25 percent return. As I have noted, if equal utility gain is the best principle, then there may be nothing unjust in A's proposal at all. Second, although there are several plausible candidates for a solution to our normative question, none is nonproblematic. At first glance, equal utility gain was an attractive moral principle. But, as we have seen, A has a threat advantage over B precisely because A does not gain as much utility from a unit of resources as does B. Thus equal utility may well prescribe exactly the sorts of distributions that motivated the concern with inequality of bargaining potential in the first place. That does not show that equal utility gain is the wrong principle. Perhaps we should worry less about inequalities in bargaining power. But I do not see how we can adopt equal utility gain as the best principle for distributing the social surplus without setting aside our worry about inequalities of bargaining power. We might consider equal division of objective resources, say where the $500 surplus in the Investment Case is divided equally. There are at least two difficulties here. First, the notion of an equal division of objective resources seems impossible to apply in some cases. We can say that A and B are dividing a social surplus of $500 in the Investment Case because we have an easily measurable and objective resource. But what should we say when they are trading goods, services, and risks? If we say that they are trading gains in utility, then we have abandoned the notion of objective resources in favor of the principle that we have already found problematic. Second, it may be

UNCONSCIONABLE CONTRACTS

69

wrong to distribute resources equally when the parties are making unequal contributions to the transaction. Perhaps B should get more in the Investment Case because B is contributing more. We may think that the fairness of a payoff structure cannot be set by an "end-state" or structural principle, but crucially depends upon the parties' (historical) contributions to the social surplus.96 And so we might look to the proportionality principle—"one that distributes the cooperative surplus in proportion to the contributions of individuals."97 Bracketing the question as to whether this principle is morally attractive, it, too, raises thorny questions about measurement. Should each party's contribution be evaluated in "objective" terms or in terms of its utility to the contributor, in which case each $100 from B is worth more than $100 from A? And should a party's contribution be understood as its precooperative or postcooperative (marginal) value? Is A's contribution of $100 a contribution of $100 (as compared with B's contribution of $400), or is it a contribution that made it possible to realize a $500 gain? Is the Anna contributing a rope to the Port Caledonia, which, ex hypothesi, is worth little prior to the transaction, or is it contributing a vessel-saving device, which, ex hypothesi, is worth a lot? And what is the vessel contributing? Its need for a rescue? Although I have raised questions about various plausible principles of fair division, I do not want to deny that there are solution(s) to the problem. Indeed, I shall suggest some possible solutions in Chapter 7. I do want to claim that absent some principle of fair division, it is not clear that the alleged injustices that arise from inequalities of bargaining power are in fact unjust. Moreover, even if the results of a bargain are unjust, it is not clear that the cause or remedy for the injustice should be laid at the feet of a party to the transaction. And that gives rise to the third point. The moral quality of the outcome of a bargaining situation will depend upon two factors: (1) whether the bargaining solution corresponds to reasonable principles of fair division, given the background conditions; and (2) whether the background conditions are themselves reasonably just. Both (1) and (2) are important. Principles of fair division are important because the parties may not abide by the correct principle for dividing the social surplus, even if the parties' background endowments are just or not unjust. But the background endowments are also important, for even 96

On the distinction between "historical" and "end-state" or "patterned" principles, see Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), p. 155. 97 Jody S. Kraus a n d Jules L. Coleman, "Morality a n d t h e Theory of Rational Choice," in Peter Vallentyne, ed., Contractarianism and Rational Choice (New York: Cambridge University Press, 1991), p . 267.

70

CHAPTER TWO

if the parties abide by reasonable principles of fair division, the bargain may result in unjust holdings if the background conditions that they bring to the table are themselves unjust. Recall the Investment Case. Suppose that unbeknownst to B, A has stolen $100 from B. Otherwise B would not have needed A's $100 and B would have used his or her own $500 to generate a $500 gain. Suppose as well that having stolen the $100, A decides not to press the threat advantage. A agrees to a proportional ($100/$400) split of the surplus. Even if the bargain was consistent with principles of fair division, the result of the bargain was eminently unfair because it passed through—nay, amplified—the original injustice.98 This story is, of course, a bit too neat. If the background distribution of holdings that parties bring to a transaction is unjust by some general principle of social justice, the injustice will often not be due to any historical injustice between A and B. Whatever we think of the bargain between Murphy and McNamara, for example, we have no reason to think that McNamara was responsible for Murphy's poverty. This suggests that we may need to distinguish between the claim that the results of a bargain between A and B are unjust and the claim that A acted unjustly toward B. We need to distinguish between the claim that A takes unfair advantage of B and the claim that A takes advantage of an unfairness to B. I have suggested that a bargain may result in an unjust distribution if it follows the wrong principle for dividing the social surplus or if it follows the right principle but originates in an unjust distribution of holdings. The fourth point concerns the moral force of the previous point. That a bargain results in an unjust distribution does not entail that we should seek to prevent such bargains from occurring. If it is desirable to allow B to improve his or her situation, and if we assume that the background injustices notwithstanding, A is under no special moral requirement to repair that injustice, then it is at least arguable that contract law should uphold bargains that result from inequalities of bargaining potential. It is one thing to disallow contracts in which the strong "push the weak to the wall," as in Bundy, and quite another to disallow contracts in which the strong find the weak "at the wall," and where the contract will provide the weak with a little distance. 98

"Even if the bargained distribution of the surplus is as moral as can be, the assignment of rights that results is still mainly determined by the respective holdings that the parties brought to the bargaining table. And there is nothing [necessarily] moral about that/' Jeremy Waldron, "Criticizing the Economic Analysis of Law/' 99 Yale Law Journal 1441 (1990), at 1470. This article is a review of Jules Coleman, Markets, Morals and the Law (Cambridge: Cambridge University Press, 1988).

UNCONSCIONABLE CONTRACTS

71

CONCLUSIONS

Without restating the arguments of this chapter, we can take stock and press a few points a bit further. The first point is this. Without having settled on the best principle of fair division, I think we can say that while some allegedly unconscionable contracts are defective on procedural or substantive grounds, other allegedly unconscionable contracts may be fair after all. Without knowing more, I see no reason to think that Schroeder took unfair advantage of Macaulay or to doubt that the ex ante value of the contract fairly reflected the genuine risks of dealing with unknown songwriters. Similarly, and without knowing more, it is entirely possible that the cancellation clause in Gantos's contract with Gianni is best understood as an assignment of risk for which Gianni was compensated in the price it received from Gantos. But other cases are more troublesome. Let us focus on those contracts that we find morally troublesome, acknowledging that we may not all find the same cases morally troublesome. We can classify putatively unconscionable contracts on two criteria: (1) whether the contract is (ex ante) harmful or beneficial to B as contrasted with the precontractual baseline; and (2) whether the contract is or is not highly profitable to A. Potentially, we can identify four types of unconscionable contracts: (a) harmful to B, highly profitable to A; (b) harmful to B but not highly profitable to A; (c) highly profitable to A but at least somewhat beneficial to B as well; and (d) at least somewhat beneficial to B and not highly profitable to A (see Figure 2.1). As we have seen, contracts of type (a) and (b) will occur when B is placed under such stress that B cannot resist what looks suspiciously like a nonadvantageous transaction (hence rules providing for a "cooling-off" period for contracts made with door-to-door salespersons) or when B does not understand or lacks the capacity to understand the terms of the contract or the value of the goods that have been purchased. This was the primary force of unconscionability in equity, and it may capture the intent of U.C.C. Section 2-302 in seeking to protect the buyer from "unfair surprise." In these cases, the principle of unconscionability can be understood as a special type of legal paternalism. It is a special type because B's self-harming action is High Profit

Low Profit

Harmful

(a)

(b)

Beneficial

(c)

(d)

Figure 2.1

72

CHAPTER TWO

a direct response to A's arguably wrongful behavior, whereas there is no "A" in most cases of legal paternalism, such as seat belt laws. But it is nevertheless a form of paternalism because in contrast to most cases in which A harms B, the harm to B cannot occur without B's consent. In any case, because we have reason to think that B's consent is not appropriately voluntary or valid in these cases, this invocation of unconscionability is consistent with respect for both freedom of contract and efficiency, and therefore relatively easy to justify. It is important to note in this connection that there is no reason to assume that contracts that are harmful to B typically generate high profits for A, that they are type (a) rather than type (b). Consider Murphy. Assume for the sake of argument that Murphy would not have agreed to the "rent to own" plan if she had fully understood the plan and the alternatives. It is, nonetheless, distinctly possible that "rent to own" plans are not particularly profitable for the dealers, perhaps because many television sets are quickly returned, having already depreciated greatly in value. We may say that such contracts are unconscionable not because they generate high profits to the sellers but because it is unconscionable for sellers to make any profits on deals that are likely to prove harmful to the buyer, and that this is particularly true when we have reason to doubt the validity of the buyer's consent—"there will be times when it would be better, morally for a firm to go out of existence than to do what is necessary to survive."99 Let us now consider type (c) cases, where A makes a high profit, yet the agreement is beneficial to B. This will typically occur when there is a market imperfection, where A has a monopoly on the good or service or B has incomplete information. Such agreements may be greatly beneficial to B, but we may think that they should have been even more beneficial (or less costly) to B. Consider The Port Caledonia. Given the choice between contracting with the tug on "extortionate" terms and not contracting at all, the vessel will prefer the "freedom to choose." But the vessel might prefer a regime in which it is prevented from entering into an "extortionate" Pareto-superior agreement with the tug if the tug would then rescue the vessel on better terms. Here, the argument for unconscionability is not paternalistic but strategic.100 It is in the vessel's interest to have its bargaining range limited, not because the vessel does not know what is in its interest, but because limiting its options puts it in a stronger bargaining position.101 99 100

M c M a h o n , Authority and Democracy, p . 10. S e e A . J o h n S i m m o n s , On the Edge of Anarchy

(Princeton: P r i n c e t o n U n i v e r s i t y

Press, 1993), pp. 237-238. 101 "What we have here is a situation in which someone may be better off doing a

UNCONSCIONABLE CONTRACTS

73

Here the doctrine of unconscionability effectively seeks to replicate the results of a more competitive market in rescues, or if there is no room for a competitive market in rescues, it seeks to replicate the profit margins that businesses are likely to receive in a competitive market. In effect, the doctrine of unconscionability tries to supply the price that would have motivated the rescuer to go into the business of rescuing.102 The most interesting and difficult cases are the type (d) ones, beneficial/low-profit contracts, where arguably unconscionable but minimally beneficial terms exist in a competitive environment. Assume for the sake of argument that Henningsen and Williams both exemplify this situation, but that there is the following difference between them. In one type of case, perhaps illustrated by Henningsen, there is an information and collective action problem that can be solved on Pareto-superior terms. In the second type, perhaps illustrated by Williams, no Pareto-superior moves are available. Suppose the following: (1) Most car buyers would be better off with a more inclusive warranty, even if it required them to pay a higher price. (2) Automobile manufacturers are quite prepared to offer such warranties if doing so does not put them at a competitive disadvantage. (3) It is too risky for any manufacturer to break ranks and sell the soft terms contracts on its own. If society declares that the low price/harsh terms contracts are unconscionable, then most buyers will be better off and the manufacturers will not be worse off. Declaring the contracts unconscionable serves to create a larger social surplus, where the buyers get virtually all of the increase. In this case, softer contractual terms are analogous to safety features. We require automobile manufacturers to install safety features, in part because it solves a collective action problem (it is cheaper for those who want those devices if all cars contain them) and in part for paternalistic reasons (some buyers would opt for a low price/unsafe car over a high price/safe car because they systematically underestimate the certain deal than not doing it if he is permitted to do it, but would be better off still if he were not permitted to do it." Brian Barry, "Lady Chatterley's Lover and Doctor Fischer's Bomb Party: Liberalism, Pareto Optimality, and the Problem of Objectionable Preferences," in Jon Elster and Aanund Hylland, eds., Foundations of Social Choice Theory (Cambridge: Cambridge University Press, 1986), p. 21. 102 The courts act "as a kind of price-setting commission, much like the state commissions formed to regulate public utilities and other natural monopolies." Richard Craswell, "Property Rules and Liability Rules in Unconscionability and Related Doctrines," 60 University of Chicago Law Review 1 (1993), at 20. Moreover, by restricting the range of possible agreements, the doctrine of unconscionability reduces transaction costs between the parties.

74

CHAPTER TWO

risks of unsafe cars). This may be an eminently sensible reason to interfere with the market, but it should be recognized that this invocation of unconscionability has little to do with inequalities of bargaining power. Another beneficial/low profit case deserves to be mentioned in this context. Suppose that automobile buyers are heterogeneous with respect to bargaining ability and that dealers encounter a comparable mix of good bargainers and bad bargainers. Assume as well that the average profit margin to the dealer is not exorbitant, but that it consists of low marginal profits with the good bargainers and high marginal profits with the bad bargainers.103 Under these conditions, dealers must "take advantage" of the bad bargainers if they are to stay in business. If we were to prevent dealers from exploiting the bad bargainers by imposing measures that reduce price dispersion (ironically, by putting all deals on the "take it or leave it" basis that is itself often cited as a form of exploitation), the bad bargainers would gain, but the good bargainers would lose. This suggests that we can misidentify the locus of exploitation. In effect, the bad bargainers are "exploited" by the good bargainers and not the dealers, although it is not clear that we would want to say that the good bargainers are wrongfully exploiting anyone if, given the system, it is not wrong for them to bargain aggressively even though the bad bargainers suffer as a result. The most difficult cases arise when contractual terms are minimally beneficial to B and generate only competitive profits to A and no Pareto-superior contracts are available. Suppose (what probably is true) that Williams would not have been (so) poor in a just society and (what is probably not true) that Williams's contract with WalkerThomas accurately and fairly reflected the background risks of dealing with very poor customers, that there were no collective action problems, and that Williams understood the implications of the addon provision. In other words, I want to suppose that the harsh (addon) terms are the best that the market can provide, given the background endowments of the parties. Here we encounter a question of moral force. Should we refuse to enforce this contract on grounds of unconscionability? At first glance, it would seem that it would not be justifiable to refuse to enforce such contracts, for at least two reasons. First, because we have reason to think that Williams is acting voluntarily, we should honor the contract 103 Say, for example, that good and bad bargainers each constitute 50 percent of the purchasers and that the dealers make a $400 gross profit per car with the good bargainers and a $1,600 gross profit per car with the bad bargainers, for an average gross profit of $1,000 per car.

UNCONSCIONABLE CONTRACTS

75

out of respect for her autonomy. Second, because the contract is beneficial to Williams, it would show a lack of regard for her welfare if we refused to honor such agreements. Thus if our primary concern is to promote the welfare of the weaker party in the specific transaction, it seems that these sorts of contracts should not be held unconscionable. Several arguments however, might be advanced for refusing to enforce these sorts of unconscionable contracts, which I will only mention here, because they will receive more extensive discussion in Chapter 9. The first line of argument maintains that unconscionable contracts should not be enforced because they generate psychic disutility for other members of the society, who do not like seeing people such as Williams enter into such arrangements. This argument does not seem particularly compelling. If the society is not prepared to rectify the unjust background conditions, it would seem self-indulgent for the society to prevent people such as Williams from improving on their admittedly unjust situation just because such contracts make other members of the society uncomfortable. A second line of argument for prohibiting unconscionable contracts focuses not on psychic disutility but on principles of justice. The fact is that people are not insensitive to the justice of their agreements. The fact is that people can and sometimes do accept less (or refuse to pay more) than they would receive (or pay) if they acted as rational maximizers. From that perspective, the doctrine of unconscionability can be understood as one way that a society signals its commitment to principles of fair division. Here the law performs an expressive or educative function.104 It prevents some individuals from entering into beneficial arrangements to increase the probability that other contracts will be more fair. The arguments that we have considered justify refusing to enforce unconscionable agreements by reference to the effects on individual welfare—whether it be the welfare of the weaker party whose strategic position is enhanced or the welfare of other members of the society. By contrast with such welfarist arguments, we might note two perfectionist arguments for refusing to enforce such agreements. First, it could be argued that a person is morally harmed by entering into an unjust relation, even if the relation does not set back his or her nonmoral interests. On this view, unconscionable contracts may injure a person's integrity or character, "even though all his [nonmoral] interests flourish/'105 A second perfectionist argument would hold 104

And a society can be committed to principles of fair division even though it also maintains a system of rights that allows parties to enter (somewhat) unfair agreements. 105 Joel Feinberg, Harm to Others ( N e w York: Oxford University Press, 1984), p . 67.

76

CHAPTER TWO

that an unconscionable contract is evil even though it is not bad for anyone, and that it is legitimate for the state to refuse to endorse evil agreements. And that brings me to a final thought about the relevance of the doctrine of unconscionability to our broader concern with exploitation. I have been examining the legal doctrine of unconscionability precisely because it seems especially concerned with exploitation, and I believe that the analysis will be helpful to the project. Yet legal principles may not fully capture our thinking about exploitation. The wrong of mutually advantageous exploitation may lie partly in the realm of decency, fairness, and virtue, whereas the law operates principally in the realm of obligation and rights. There may be good reasons—good moral reasons—for preferring that the law not attempt to replicate or enforce the whole of morality. It would be better if Scrooge were not so stingy, but it would not necessarily be better if the legal system were designed so as to prevent or punish his indecent behavior. If this is so, there is no reason to think that a legal doctrine such as unconscionability will provide a complete account of the wrongness of exploitative transactions. It may illuminate important features of that landscape, but it will not illuminate them all.

Chapter Three THE EXPLOITATION OF STUDENT ATHLETES

Item: In a critique of university athletic programs, a sports reporter states that "the nation's universities are shamelessly exploiting the players."1 Item: A well-known African American sociologist states that the "black athlete on the white-dominated college campus . . . is typically exploited, abused, dehumanized."2 Item: A former president of Stanford University said that "the [university athletic] system, to be brief and blunt about it, reeks of exploitation."3 Item: Eric Ramsey, a football player at Auburn University, secretly taped a conversation with his coach, Pat Dye, that indicated Dye's willingness to provide Ramsey with financial help that was contrary to the rules of the National Collegiate Athletic Association (NCAA). Commenting on the case, a sports writer observed: "Presidents holier than thou refuse to be fair to the athletes because it would cost them money. The scandal... is not that someone paid [Ramsey] $100 here and there. He earned more and deserved more."4 Item: Gale Sayers, a former star football player for the University of Kansas and the Chicago Bears, said that he was "sick and tired of reading about how the black athletes are being exploited by colleges and universities. . . . Today's black athletes should and do know what is expected of them when they arrive on a college campus." 5 1 Rick Telander, ''Something Must Be Done," Sports Illustrated, October 2, 1989, p. 92. 2 Harry Edwards, The Revolt of the Black Athlete (New York: Free Press, 1969), p. 20. 3 Donald Kennedy, "So What If College Players Turn Pro Early?" New York Times, January 28, 1990, sec. 4, p. 21. 4 David Kindred, The Sporting News, December 7, 1972, p. 5. 5 USA Today, October 13, 1992, p. 14C.

78

CHAPTER THREE

Item: Representative Cardiss Collins (D. 111.) stated that "colleges [must] not exploit our youngsters for victories and profits. Instead, colleges must provide an opportunity for a real education/' 6 Item: Responding to the charge that the University of Southern California exploited its student athletes, Mike Garrett, winner of the 1965 Heisman Trophy (awarded to the outstanding college football player) said: "Exploited? Hey, exploit me again like that. For a kid to get a scholarship to USC is a great, great opportunity. To some degree, it has to be up to the kid to take advantage of it. It was absolutely the best thing that happened to me." 7 Item: Terry Nelson, a basketball player at the University of Cincinnati, was incensed when the university started selling trading cards of the team's basketball players for $6 a pack: "It's not right for the school to make a profit off of us and we can't make any money."8 As these items illustrate, the claim that American universities exploit student athletes is frequently made and sometimes denied. In capsule form, the exploitation story goes something like this. "Big-time" college athletic programs—a description that is typically limited to football and men's basketball at a limited number of institutions—provide significant financial benefits to the institution. The athletic programs generate direct revenue from ticket sales, television contracts, trading cards, and a growing business in royalties for licensing the university's seal, logo, and colors.9 Athletic programs may also generate indirect revenue in the form of contributions from alumni/ae, whose commitment to the institution is said to be tied to the success of the athletic programs. In addition, a state's willingness to appropriate funds for public institutions may reflect the importance of the institution's athletic program. In exchange for these benefits to the institution, the universities give several benefits to the student athletes. The athletes receive the opportunity to obtain a college education and a college degree, and this at institutions to which they may not have been admitted on the basis of their academic record. They are provided with full scholar6

USA Today, August 17, 1992, p. 10A. Ibid. At the time, Garrett was Associate director of athletics at USC. 8 New York Times, March 26, 1993, p. B7. According to NCAA rules, the players cannot receive any of the profits. 9 Some college football stadiums hold approximately 100,000 persons and are filled for every game. During the 1993-1994 academic year, the University of Michigan received nearly $5.8 million from such licenses. New York Times, September 11, 1994, sec. 3, p. 5. 7

EXPLOITATION OF STUDENT ATHLETES

79

ships, covering tuition, room, and board. And because professional football and basketball recruit virtually all their players from the college ranks, some student athletes receive the opportunity to play in professional sports.10 In exchange for those benefits, student athletes must practice for as much as thirty hours per week during the athletic season, keep in condition during the off season, maintain their academic eligibility, and, of course, perform. Why is this arrangement thought to be exploitative? There are two versions of the exploitation story. One version maintains that the institution receives substantial benefits from the performance of these student athletes, while the student athletes fail to receive a genuine college education and a college degree. True, some universities (e.g., Penn State, Michigan, Notre Dame, Duke) have reputations as the "good guys," as schools that have reasonably high admissions standards for student athletes and from which most student athletes graduate.11 These institutions seem to make reasonably high academic demands on their athletes and provide the support necessary to enable most student athletes to meet them. By contrast, other institutions (e.g., North Carolina State, Nevada-Las Vegas, Oklahoma) have reputations for low admission standards, little concern for the academic progress of their athletes, and low graduation rates for the athletes.12 This version of the exploitation story claims that only the "bad" institutions are exploitative because the university profits while the student athletes do not receive "anything lasting in return." 13 The second version of the exploitation story claims that universities exploit student athletes because the gains to the institution and its student athletes are grossly asymmetrical: even if the student athletes receive a high-quality education and a college degree, they do not 10

Major league baseball continues to recruit many of its players from its minor league system. 11 On the "good guys," see "Jocks with Books/' Newsweek, January 9, 1989, p. 60. Frederick Schauer pointed out to me that there are several different criteria for being a "good guy": (1) scrupulous adherence to NCAA rules; (2) high admission standards; (3) high graduation rates, which may be a function of safe-harbor programs, demanding coaches, and compliant faculty; (4) genuine opportunity for the typical athlete to get a real education; and (5) no academic breaks for athletes. These criteria do not always correlate with one another. 12 Nand Hart-Nibbing and Clement Cottingham, The Political Economy of College Sports (Lexington, Mass.: Lexington Books, 1986), p. 88. Indeed, many student athletes fail to gain the most basic educational skills. It has been estimated that 20 to 25 percent of African American athletes at four-year colleges are functionally illiterate. See Derek Quinn Johnson, "Educating Misguided Student Athletes: An Application of Contract Theory," 85 Columbia Law Review 96 (1985). 13 Kennedy, "So What If College Players," sec. 4, p. 21.

80

CHAPTER THREE

receive an appropriate return on the financial surplus that they generate. This version of the exploitation story applies to both "good" and "bad" institutions. Are student athletes exploited? If so, which version of the exploitation story is correct? THE UNIVERSITY'S GAIN

I have suggested that A exploits B only if A gains from the transaction. Let us first consider the gains to the university. We can categorize university athletic programs on two dimensions: scope and surplus. With respect to scope, I distinguish between big-time and small-time athletic programs. A big-time athletic program is one that competes at the highest level, seeks national prominence, and provides almost all student athletes (in the major sports) with full scholarships. I consider virtually all the institutions in the major conferences (e.g., Big Ten, Big Eight, Southeast, Atlantic Coast, Pacific Ten, etc.) and some independent institutions (e.g., Notre Dame) to be bigtime programs. By contrast, small-time programs do not aspire to national rankings, and few student athletes receive financial aid unrelated to financial need, although many would not have been admitted to the institution on the basis of their academic record alone. These institutions include almost all small private colleges, most small public universities, and some major private universities (e.g., Harvard, Yale, Chicago) that do not emphasize their athletic programs (although there may be intense athletic rivalries among them). With respect to surplus, I distinguish between those athletic programs that actually produce surplus revenue (direct and indirect) for the institution and those that do not. As we shall see, whether athletic programs produce surplus revenue will depend upon the breadth of our conception of surplus revenue. On a broad conception, more athletic programs produce surplus revenue than a financial analysis would suggest. But let us start with surplus financial revenue, as it is ordinarily understood. Given the two distinctions, four types of institutions are theoretically possible: (a) big-time, with surplus revenue; (b) small-time, with surplus revenue; (c) big-time, without surplus revenue; and (d) small-time, without surplus revenue (see Figure 3.1). I place a question mark in cell (b) because it may be empty. I do not know that any small-time programs produce significant surplus financial revenue. More important for present purposes, there may be many type (c) programs, big-time programs that do not produce a significant financial surplus. Big-time athletic programs are very expensive. An unsuccessful team is not likely to generate significant revenues from television and postseason (or tournament) play. In ad-

EXPLOITATION OF STUDENT ATHLETES

Big-Time

Small-Time

Surplus

(a)

(b)?

No Surplus

(c)

(d)

81

Figure 3.1 dition, the ability of athletic programs to generate alumni/ae donations (particularly donations from which academic programs benefit) appears to be vastly overstated.14 If we concentrate on big-time programs, it should be noted that there is no necessary correlation between the capacity of the program to generate a surplus and the institution's commitment to the academic progress of its student athletes. There are "good" and "bad" institutions in both the surplus and nonsurplus categories. But the distinction between surplus and nonsurplus institutions raises a central issue with respect to exploitation. I have argued that one person can exploit another only if the alleged exploiter gains from the relationship. If this is so, then nonsurplus institutions do not exploit student athletes because their student athletes do not produce a net surplus for the institution. To say that nonsurplus institutions do not exploit their student athletes is not to let them off the moral hook. They may be callously indifferent to the interests of their student athletes. We might say that these institutions are attempting to exploit student athletes or are acting exploitatively because they are trying to generate a surplus from the efforts of the athletes. And we may say that attempted (or unsuccessful) exploitation is seriously wrong, just as we regard attempted (unsuccessful) crimes as both wrong and punishable.15 Note, however, that while the victim of a battery may have a civil claim for compensation against the offender, the intended victim of an attempted battery (where A throws a punch at B but misses) has no such civil claim. Similarly, even though nonsurplus institutions may be acting wrongly, their student athletes could not complain that they are being 14

See Roger G. Noll, "The Economics of Intercollegiate Sports," in Judith Andre and David N. James, eds., Rethinking College Athletics (Philadelphia: Temple University Press, 1990). At most institutions, only football and basketball generate surplus income, and that surplus is used to fund other athletic programs. For a survey of research on the relationship between successful athletic programs and alumni/ae giving, see James H. Frey, 'The Winning-Team Myth," Currents, January 1985, pp. 32-35. 15 See Lawrence Becker, "Criminal Attempt and the Theory of the Law of Crimes," 3 Philosophy & Public Affairs 262 (1974). Becker argues that attempted crimes may be as wrong—qua crimes—as successful crimes and that there are good reasons for punishing attempted crimes as severely as successful crimes.

82

CHAPTER THREE

denied a fair share of the profits they generated for their institution, for no profits were generated. It might be objected that I have adopted an excessively narrow view of exploitation. As I observed in Chapter 1, it may be thought that A exploits B when A "uses" B merely as a means for A's own ends and without proper regard for B's interests, even if A does not gain or does not gain unfairly from A's use of B. I believe that this objection fails. First, as I argued in Chapter 1, it is not clear that A violates the Kantian maxim if A respects B's capacity to consent to a transaction, even if A uses B's situation as a means to A's own ends. Second, so long as we can say that it is wrong for A to act exploitatively toward B, I see no reason for the Kantian to resist the view that successful exploitation requires an actual advantage to A. Third, acting "exploitatively" in that sense does not always seem wrong. We typically treat others primarily as a means to our own ends in ordinary market transactions. Even if we would prefer a world in which buyers and sellers treated each other as ends in themselves rather than merely as means to their own ends (although, and depending upon what this all means, I'm not sure that we would prefer such a world), the supermarket does not take unfair advantage of the customer (or vice versa) if A does not gain an advantage from B, whatever the spirit in which the parties undertake the transaction. It is worth noting that the argument I am advancing here is consistent with a Marxian account of exploitation. On a Marxist view, "exploitation is the appropriation by a class of nonworkers of the surplus product of a class of workers."16 On this view, if workers generate no surplus product beyond what is necessary for their subsistence, there can be no exploitation. And this, says Richard Arneson, is a disturbing implication of that account: This implies there is no exploitation in a slave society in extremis in which conditions of production have declined so that slaves do not produce more than is needed for their subsistence but are continually starved so that their masters may live sumptuously. In the ordinary sense of the term there could hardly be exploitation more brutal than this.17 It is not clear that Arneson has reason to be concerned about the implications of his example. First, although there is no surplus product, when surplus is defined technically in terms of what is required for the slaves' subsistence, the masters are extracting a surplus from the slaves because the slaves are receiving less than subsistence requires. 16 17

Richard Arneson, "What's Wrong with Exploitation?" 91 Ethics 202 (1981), at 203. Ibid., pp. 203-4.

EXPLOITATION OF STUDENT ATHLETES

83

It is true that such a practice might be self-defeating from the master's perspective, but that is a different matter. But suppose that the masters continue to provide for the slaves' subsistence during the decline in productivity, say, so that they can have a group of productive slaves when better times return. During this period, the masters continue to live well off the surplus accrued from previous periods, but they extract no surplus from the current group of slaves. I see no reason to resist the view that the slaves are not exploited during this unusual period. We have other terms available to describe the wrongful treatment of the slaves. Although I am comfortable with the view that exploitation requires a surplus, the previous discussion raises the question as to how to define, measure, and locate that surplus. Suppose that a big-time but nonsurplus institution correctly claims that its football program generates no financial surplus, given the high salaries that are paid to a large coaching staff, but that there would be a substantial surplus if the revenues were not used to pay exorbitantly high salaries to an excessive number of persons.18 If the surplus is effectively "hidden" in the cost structure of the enterprise, there still exists precisely the sort of financial surplus that makes exploitation possible, although it might be more accurate to say that it is the recipients of the salaries who are exploiting the student athletes and not the "institution" as such.19 This gives rise to a more general point. To say that the "university" is reaping the surplus is to speak metaphorically. Does the surplus go to the library? To faculty salaries? The Classics Department? To scholarships for needy students? Does the surplus pay for gender equity by subsidizing women's athletic programs that run a financial deficit?20 It is, of course, extremely difficult to identify the recipients of a surplus in an organization with a complex budget, although we do not need to know who receives the benefits in order to know that the 18

Less than one-third of Division I-A football programs break even. John Underwood, "Reading, Writing and Remuneration," New York Times, September 11, 1994, sec. 5, p. 13. 19 Although there is no competitive market among players, there may be a highly competitive market among coaches. Bobby Bowden, coach of Florida State University's 1994 national championship football team, receives a salary of $700,000. 20 Because Title IX of the Education Amendments of 1973 requires that "all aspects of a university education, including athletics, should be free of gender discrimination," federal law might be characterized as requiring that universities use financial surpluses generated by male athletic programs to subsidize women's athletic programs rather than for other purposes. Jill K. Johnson, "Title IX and Intercollegiate Athletics: Current Interpretation of the Standards of Compliance," 74 Boston University Law Review 553 (1994), at 553. I thank Fred Schauer for raising this issue with me.

84

CHAPTER THREE

student athletes do not. This also suggests that we need to distinguish between those who make the decisions that generate and allocate a surplus and the recipients of the surplus themselves. The recipients may be entirely unaware that they are benefiting from a surplus derived from the efforts of others. In addition to issues connected to financial surpluses, it might be claimed that the student athletes are generating nonfinancial benefits for which they are not compensated. They provide entertainment for students, faculty, alumni/ae, and the general public, some of whom receive the benefit on television or radio. They may assist in recruitment of students (and dare I say faculty?). A state's major public institutions may be an important source of "state pride." In these cases, it is not that the universities are hiding financial revenues. Rather, the student athletes create important benefits that are not converted into financial revenues. If nonfinancial benefits to the universities should be regarded as the sort of surplus gain that makes exploitation possible, then many institutions that appear to be of type (c) are actually type (a), and perhaps some type (d) institutions are actually type (b) institutions. In addition, if the non-revenue-generating benefits to others (entertainment, state pride) should be regarded as the sort of surplus gain that makes exploitation possible, then it is possible that other persons are receiving benefits through the institution's activities apart from any benefits to the university itself. In any case, this suggests that moral criteria are required not only to evaluate the distribution of a social surplus but to establish the presence and magnitude of a social surplus. THE STUDENT ATHLETE

Suppose that the student athlete, B, is a football player enrolled at University A, a type (a) institution that generates a significant surplus from its football program. Let us also assume that B contributes to the competitive and (therefore) financial success of the athletic program.21 Is B exploited? And, if so, is it a case of harmful exploitation or mutually advantageous exploitation? Before we can consider these questions, I think that we need to confront directly the character of the educational and athletic benefits that B receives from A. Let us consider the educational opportunities first. As we saw in Chapter 2, we should understand such benefits in terms of their ex ante value rather than their ex post value. In one 21

It is an interesting question whether B might validly claim to be exploited even if B did not contribute to A's surplus (say, because B is injured) if, given the ex ante value of B's contribution to A, B receives less than he should.

EXPLOITATION OF STUDENT ATHLETES

85

sense, educational opportunities are like a lottery ticket. There is no guarantee that anyone admitted to an institution will succeed, athlete or nonathlete. If A were to give B a lottery ticket that has a .50 chance of winning a $100 prize, it would be wrong for B to claim, ex post, that he or she had received nothing of value just because the ticket didn't win. Similarly, it would be a mistake to claim that University A gives nothing of value to Student B merely because Student B does not graduate or receive an education. And that, of course, brings out the sense in which a college education is unlike a lottery ticket. For whether B actually receives an education or a degree depends in large part although not altogether upon whether B makes the appropriate effort. The ex post value of the educational opportunity is at least somewhat subject to B's control.22 Thus the question is not whether B actually receives an education or graduates, but specifying the ex ante value of the educational resources or opportunities that are provided to B. The value of educational opportunities to B is, of course, not only a function of B's efforts; it is also a function of B's capacity to use those educational opportunities. Some students, whom I shall call "lowability'' students, are not capable of succeeding in an academic environment. Some such students have considerable natural ability, but are woefully underprepared to succeed in an academic environment. Others may lack the natural ability to succeed. In any case, and whatever the source of the deficiency in capacity to succeed, it is distinctly possible that some students lack the capacity to profit from college courses and do not have a reasonable chance of obtaining a college degree, even with good support. Are low-ability students vulnerable to exploitation?23 On the assumption that low-ability students can somehow maintain their academic eligibility and thus continue to engage in intercollegiate athletics, it is arguably unfair for A to receive a benefit from B's services when B cannot receive a compensating benefit from A. It might be replied that A does not exploit B if B voluntarily assumes the risk of not benefiting from the educational opportunities provided by A. But this is wrong. First, as I shall argue at more length in Chapter 8, I believe that A can exploit B even if B consents to the exploitation. 22

A s M i k e G a r r e t t p u t it, ' T o r a k i d t o g e t a s c h o l a r s h i p . . . is a g r e a t , g r e a t o p p o r tunity. To some degree, it has to be up to the kid to take advantage of it." Quoted in Phil Taylor and Shelley Smith, "The Black Athlete Revisited," Sports Illustrated, August 12, 1991, p. 46. 23 We might want to say that colleges should not admit low-ability students because it violates what should be the standards of an academic institution, but that is a different issue.

86

CHAPTER THREE

Second, it is not clear that this exploitation is genuinely consensual. If it is not B's intention to make a gift of his or her services to A, if we assume that B expects to gain from the educational opportunities offered by A, then we may also have reason to think that B does not have the information or judgment necessary to voluntarily assume such risks. There is, of course, another package of benefits that the university provides to student athletes, and those are related to athletic participation itself. First, student athletes may value playing college sports in its own right. Second, big-time athletic programs also provide the opportunity for student athletes to become professional athletes. They provide training facilities, coaching, publicity, and exposure. It is true that only a small proportion of student athletes ever become professional athletes, so the value of the athletic opportunity may be relatively low. But it is not nothing. Indeed, for some students, the value of the athletic opportunity, low as it may be, will be greater than the value of the educational opportunities. Suppose that the value of the athletic opportunities to student athletes is reasonably high. It might nonetheless be argued that an adequate moral accounting of a relationship requires a principle for allocating credit for the values that are received in the context of that relationship. Thus it might be argued that we must distinguish between the athletic opportunities that are provided by A and the athletic opportunities that B receives from having been at A, and that when we make this distinction, we will find that the benefits provided by A are less than they may seem. Alumni/ae may believe that they should acknowledge and reciprocate the value of the education that they received from their universities, but do not think that they need to acknowledge and reciprocate the university because it provided the context in which they made lifelong friends or met their spouses. Similarly, it is arguable that the university should not receive (all) the credit for the exposure that student athletes receive, even if it provides the context in which they receive that exposure. Let us put the previous worries aside. Let us assume that we are able to evaluate the ex ante value of the educational and athletic opportunities received by B at surplus generating institutions (however that surplus is measured). Let us assume that the value of those opportunities at what I shall call a "good" institution, AG, is equivalent to X and that the value of the opportunities provided at a "bad" institution, AB, is much less, say, .2 X. Are student athletes exploited at AB or both AB and AG? If so, is it a case of harmful exploitation or mutually advantageous exploitation? Table 3.1 lists the plausible possi-

EXPLOITATION OF STUDENT ATHLETES

87

TABLE 3.1

1 2 3 4 5 6

HE NE MAE MAE NE NE

HE HE HE MAE MAE NE

bilities (where HE = harmful exploitation, MAE = mutually advantageous exploitation, and NE = no exploitation). For case 1 or 2 or 3 to be the case, it must be true that student athletes are typically harmed by their relationship with their institutions at both AG and AB (case 1) or at AG (cases 2 and 3). What would that mean? On the most natural reading, to say that student athletes are harmed by their experience is to say that they are typically less well off as student athletes than they otherwise would have been. To verify that claim, we would require some counterfactual information. We need to know what would B's life would have been like if B had not gone to college.24 This is important. For despite all of the commentaries on the myriad ways in which student athletes are mistreated, I know of no studies that attempt to compare the overall life prospects of student athletes with others who are similar to them in terms of relevant variables (e.g., race, class, academic ability, and athletic ability) but who do not attend college or attend as nonathletes. Suppose that things go badly for Student B at AB or AG. B rarely attends classes and learns very little, does not graduate, and does not receive an offer to play professional sports. It is still possible that B's life goes better if B becomes a student athlete than if B does not.25 If, for example, B would otherwise be unemployed or employed in a 24

Perhaps more accurately, w e would need to compare an ex ante probability distribution of B's life prospects in the world in which B attends the college A G or A B a n d the ex ante probability distribution of B's life prospects in the world in which B does not go to college. 25 Fred Brown, who played basketball at Georgetown University, put it this way: "If you come into a school, you may not be on an academic par with the general population of the school, but if you as an individual can sit there and learn something and better yourself, that's an education. . . . I always ask my mother, 'If I hadn't played basketball, what would have happened?' Ninety percent of the people I grew up with are dead or in jail, and I would have been the same way." Quoted in Ted Gulp, "Foul," Time, April 3, 1989, p. 60. It should be noted that Georgetown University has a reputation as one of the "good" athletic programs.

88

CHAPTER THREE

low-skill job, it is arguable that four or five years of free room and board and some degree of local fame is not a bad deal. And even if the student does not graduate, having some additional educational experience may prove to be a long-term benefit. And this suggests still another consideration. In attempting to determine whether the typical student athlete is harmed by the college experience, we must take a view that aggregates the college years and the postcollege years. It is possible that the effects on B are similar in both periods, that B's life goes better (or worse) during both the college and postcollege years than it otherwise would. But it is also possible that these two periods have different effects on B's life. It is possible that B's life in college goes better than it would have gone had B not attended college, but that B's subsequent life prospects are worse, say, because B gives up the opportunity to learn a trade or join the military. On the other hand, it is possible that B's life in college goes worse than it otherwise would have gone, but that B's postcollege life goes better because of the knowledge and contacts that B acquires. Bracketing such aggregation problems, we can say that if B's life goes worse if he or she attends AG or AB than if he or she had not attended college at all, then we may have a case of harmful exploitation. Otherwise, not. Suppose that the typical student athlete is not harmed by the college experience. Does it follow that B is not exploited? One writer has maintained that a student athlete is exploited only if "he would have otherwise done something else more constructive with his time."26 On this view, only cases 1, 2, and 6 are conceptually possible—no harm, no exploitation. But this view assumes what has to be shown and what I have denied, namely, that mutually advantageous exploitation is not possible. Moreover, most critics do not claim that student athletes are worse off as a result of their college experience than they otherwise would have been. They do not claim that the students are harmed. Rather, they say that the student athletes are not as well off as they should be, given that they attend college, particularly given the benefits that these students provide to their universities. The problem is to determine just what that might mean. After all, even if we were to say that student athletes are not exploited because they are not harmed, we must still determine whether their package of benefits is adequate. Figure 3.2 represents the value that B receives from AG or AB along a continuum. X represents the value that B receives from College AG. 26

Elbert B. Smith, a professor of history at the University of Maryland, in a letter to the editor, Washington Post, November 17, 1986, p. A12.

EXPLOITATION OF STUDENT ATHLETES

-HE

H > <
MAE'F

89 F

Figure 3.2

AG provides its student athletes with appropriate educational opportunities. Its graduation record is high. Its students take rigorous academic programs. Its coaches demand that students attend classes, and so on. By contrast, AB provides .2 X to its student athletes. Its graduation record is low. Its students take nonrigorous courses. Its coaches take a laissez-faire attitude with respect to students' academic progress, and so on. H represents the point below which B is harmed by the college experience. F represents what B would receive if B were to receive a "fair share" of A's surplus on whatever theory of fairness should turn out to be most acceptable. On one view (what I label MAEO), B is exploited only when B receives less than X, an "objective" or "independent" standard as to what a student should receive from an institution. On another view (what I label MAEF), B is exploited if B receives less than F, if B does not receive a fair share of the surplus that his or her performance has generated. If we accept MAEO, then case 5 is true: AB exploits its student athletes, but AG does not. If we accept MAEF, then case 4 is true: AG and AB both exploit student athletes. Which is the better account of exploitation? There are several reasons for thinking that MAEO is to be preferred. First—although this is technically not a reason for preferring MAEO—it should be noted that when critics claim that colleges exploit student athletes, they typically imply that when colleges provide genuine educational opportunities, they are not engaged in exploitation.27 And MAEO distinguishes between AG and AB with respect to exploitation whereas MAEF does not. Second, MAEF implies that type (a) (surplus and big-time) institutions exploit their students whereas type (c) and (d) (nonsurplus) institutions do not, even though students at type (c) and (d) institutions may be worse off than students at AG (and even AB). It requires us to say that whether B is exploited turns not on the value of the educational resources that B is receiving, but on how well the institution is doing. And on the assumption that B is receiving a solid package of educational resources, it might seem mean-spirited to claim that B is 27 Consider Representative Collins's statement: "Colleges [must] not exploit our youngsters for victories and profits. Instead, colleges must provide an opportunity for a real education." Quoted in USA Today, August 17, 1992, p. 10A.

90

CHAPTER THREE

exploited if the institution also happens to receive a benefit from his or her performance, but that B is not exploited just because the institution does not. A third argument for preferring MAEO to MAEF is that our very conception of college athletics seems to bar a university from sharing its surplus with student athletes. We may think that student athletes should receive no less than X if they are to be treated fairly, but that they should receive no more than X if we are going to remain true to the ideal of the amateur student athlete. Perhaps the "amateur" ideal cannot be defended. Many think that it would be less hypocritical if universities were to treat their student athletes as (quasi-)professionals and simply pay them for their services.28 But if the amateur student athlete represents a better ideal, then it would seem that AG should not share its surplus with student athletes. And, it may be argued, we can hardly claim that AG is exploiting its student athletes when it does not do what it should not do. So MAEO must be the better account of exploitation. Is that correct? I am not sure. First, that MAEF does not allow us to distinguish between AG and AB with respect to exploitation is not a persuasive reason for preferring MAEO. We may simply have to confront the possibility that AG and AB are both exploitative, although it would not follow that there are no other important moral distinctions between AG and AB. Is MAEF mean-spirited because it makes exploitation turn on the gain to the institution rather than the value of the educational resources that are provided to B? I do not think so, at least not unless a concern with justice is mean-spirited. Perhaps, as Hume remarked, justice is the "cautious, jealous virtue."29 Perhaps it would be a better world in which B does not worry about A's gain and simply asks whether his or her own situation is satisfactory. But if we think that there are reasons to worry about the distribution of a social surplus, then we can hardly reject MAEO because it reflects such a concern. Moreover, and to emphasize that exploitation is only one moral concern, even if AG exploits student athletes (under MAEF), and thus is 28 ' T h e scandal . . . is n o t that s o m e o n e paid [Ramsey] $100 here a n d there. H e e a r n e d m o r e a n d deserved m o r e . " David Kindred, The Sporting News, December 7, 1972, p. 5. It should be noted that this change would not leave everything else as it is. Intercollegiate athletics is immensely popular "because it can still be sold with the idea that these athletes really d o represent a n d care about their universities. Otherwise, they're the French foreign legion." Underwood, "Reading, Writing, a n d Remuneration," N e w York Times, September 11, 1994, sec. 8, p. 13. 29 David Hume, An Inquiry Concerning the Principles of Morals (1777; reprint, New York: Hafner Publishing Co., 1948), sec. Ill, pt. I.

EXPLOITATION OF STUDENT ATHLETES

91

worse with respect to exploitation than type (c) and (d) institutions (which, by definition, cannot exploit their student athletes), it does not follow that exploitative "good" institutions are worse than nonexploitative "bad" institutions, all things considered. The third argument for MAEO appeals to the ideal of amateur intercollegiate athletics. The view that AG should not share its surplus with B can be understood in two ways. It might mean that B has no moral claim whatever to share in AG's surplus, that X represents both the minimum and the maximum of what a student has any legitimate claim to receive. On a second interpretation, the argument concedes that B may have some moral claim to share in AG's surplus, but maintains that there are overriding moral reasons for not allowing such sharing to occur. On the latter view, we could maintain that MAEF represents the better understanding of exploitation and still maintain the amateur ideal. Which is the better account? I do not think that we are in a position to make a compelling case for either MAEF or MAEO at this point. But the analysis does underscore the need to distinguish between the truth conditions of the claim that someone is exploited and the moral weight and moral force of that claim. To the extent that one is exploited when one does not receive a fair share of the surplus one helps to generate, there may be less moral weight to exploitation than meets the eye. First, the athletic talent that enables student athletes to contribute to a surplus is arguably arbitrary from a moral point of view, at least to the extent that it is a function of natural abilities as contrasted with the sorts of efforts that have a greater claim to be morally relevant. Second, the extent to which any individual athlete's performance serves to generate a surplus depends on background factors for which an individual athlete cannot plausibly claim credit, for example, the performance of the other athletes, the coaches, the reputation of the institution (which may have allowed it to successfully recruit other excellent players), and so on. Third, and more generally, whether B's performance generates a surplus is a function of the society's demand for such performance, and that, too, is at least somewhat arbitrary from a moral point of view. Consider men's basketball and women's basketball. Because the public's demand for women's basketball is significantly less than its demand for men's basketball, even successful big-time women's basketball programs may not generate a surplus. Is it morally troublesome that some male basketball players are exploited whereas virtually no female basketball players athletes are exploited? Perhaps, but probably not very much. Indeed, it might be argued that it is undesirable, nay, unjust, that student athletes in surplus programs should fare better than

92

CHAPTER THREE

student athletes in nonsurplus programs. We may think it more important to ensure equality between male and female student athletes than to eliminate this form of exploitation.30 THE BARGAINING PROCESS

We have seen that even if student athletes are not harmed by their college experience, they may be exploited nonetheless. Under MAEO, B is exploited if he receives less than X. Under MAEF, B is exploited if he or she receives less than F. How do student athletes find themselves in such a position? If universities do exploit student athletes, why are they able to do so? Does it occur voluntarily? One commentator has suggested that colleges are in a "superior bargaining position" with respect to student athletes.31 Is that so? Suppose that MAEO is the better view. Suppose as well that B is a high school senior and is being recruited by several major institutions, including AB and AG.32 It would seem that B should have some room for choice. And if B wants to attend an institution (AG) that promises genuine educational opportunities, it would seem that B could do so. Even though B may be in a position to choose AG over AB, there is a problem of knowledge and foresight. B may not know that AG is superior to AB with respect to educational opportunities or that he or she will be better off attending AG. Second, there is a problem of desire. Even if B could choose AG, B may not want to do so, preferring an institution that will make fewer academic demands. That B chooses AB over AG may do something to mitigate the force of the claim that B has been exploited by A, but it is arguably unreasonable to expect that an impressionable seventeen-year-old will make the choice that will best promotes his or her overall life prospects, particularly when the better long-term choice may be the less pleasant short-run choice. If we do not want students to be exploited in this way, then the best strategy is to (paternalistically) require all institutions to provide appropriate educational opportunities, even if students are willing to settle for less. But suppose that we believe that MAEF represents the better account of exploitation, that student athletes are exploited if they receive less than F. Here there are two questions: (1) What is the best 30

See Chapter 1, text accompanying note 93. Johnson, "Educating Misguided Student Athletes," p . 111. 32 After all, if B has not been heavily recruited, h e or she is not likely to b e a major contributor to the institution's financial success, and so will not be exploited even if h e or she does not receive (at A B ) t h e appropriate educational opportunities. 31

EXPLOITATION OF STUDENT ATHLETES

93

account of F? and (2) Why do they receive less than F? It is possible that the answers to both questions may be located in the bargaining structure, in the fact that prospective students cannot bargain for more than X because NCAA rules do not permit institutions to offer more than X. It would, I think, be a mistake to describe this situation as an "inequality of bargaining power," at least as between any particular institution and any particular athlete, for there is no bargaining at all. Just as students cannot bargain for financial incentives (including a share of the institution's surplus) when they might otherwise want to do so, institutions cannot (legitimately) offer special incentives to lure prospective students when they might want to do so.33 To say that there is no inequality in bargaining power as between institutions and student athletes is not to say that the bargaining structure is morally acceptable. No bargaining at all may be worse than an inequality in bargaining power. In effect, NCAA rules require that the contract between institutions and student athletes be made at a fixed maximum price. By creating a cartel among the universities, NCAA rules solve a collective action problem among institutions and make possible a surplus that might otherwise not exist. Although each institution might prefer to be able to offer a financial incentive in a given case, all institutions prefer a regime in which none are allowed to offer financial incentives to one in which all are allowed to offer financial incentives. Just as the cartelization of the relationship between universities and students may explain why students receive less than F, it also presents us with a way to understand the value of F itself. We can define F as the (hypothetical) value that B would receive if institutions were free to bargain with prospective students without regard to limitations on financial incentives or rewards. In other words, we can define F as the value that students would receive in a competitive market. The students would not be exploited because they would be getting an appropriate return on their contributions (as defined in this manner), and—what amounts to the same thing—the institution would no longer be receiving an exorbitant surplus on student athletes7 services. Now, a competitive market in student athletes might not be desirable, all things considered. But if, among other things, a competitive market in student athletes provides the criterion for de33

Because institutions cannot compete with one another by bargaining, they attempt to differentiate themselves in other ways, by emphasizing their high (or low) commitment to academics, the likelihood of national exposure, the size of the stadium or arena, proximity to family, the quality of coaching, the likelihood of becoming a starting player early in one's career, and so on.

94

CHAPTER THREE

fining a fair wage or fair price (at least under certain conditions), then we have learned something important about the relation between competitive markets and exploitation. The parenthetical clause in the previous sentence is critical. The moral quality of the wages and prices generated by a competitive market depends, at least in part, upon the moral quality of the background conditions that are in place. In evaluating the claim that student athletes are exploited, we have assumed a background in which student athletes exchange their athletic services for educational opportunities (admission and/or scholarships) that would otherwise not be available to them. Given those background conditions, a competitive market may give student athletes a certain value for their athletic services—a value that is likely to be greater than what some students now receive (those whose contribution is particularly high) and lower than the value that other students now receive (those whose contribution is relatively low). But the background conditions might be different. Suppose that we assume hypothetical background conditions in which all qualified students were able to attend a university without charge.34 Under these background conditions, a competitive market would generate a different "wage" (or value for F) for student athletes than they would receive under the present background conditions, just as the wages that workers receive in competitive equilibrium will turn on whether society provides a safety net for the unemployed and, if so, at what level. The background conditions are important even if we reject the view that students should receive the value that would be generated by a competitive market. Suppose that we assume that student athletes should not be paid for their services, but that we also assume hypothetical background conditions in which all qualified students are able to attend a university without charge or are guaranteed sufficient financial resources to pay for a university education. I think it fair to say that at least some students who enroll as student athletes under existing background conditions would not enroll as athletes under these hypothetical background conditions. If this is so, then intercollegiate athletic programs are importantly parasitic on the existing background conditions that make it attractive to enroll as student athletes. If the society is morally required to provide the hypothetical background conditions described above, then intercollegiate athletic programs are taking advantage of an unfair situation. 34

For example, the society (or government) might pay for their college education up front on condition that they pay back a share of their future income as a fair return on society's investment in their education.

EXPLOITATION OF STUDENT ATHLETES

95

Would that constitute (wrongful) exploitation? I am inclined to think that the answer may be no with respect to the universities but yes with respect to society. I think that the answer is no with respect to the universities, because it is not always wrong or unfair for A to enter into a mutually advantageous transaction with B even though B would not enter into such a transaction under fair background conditions, particularly if A has not caused B's unfair background conditions and has no special responsibility to remove them. We must distinguish between taking advantage of unfairness and taking unfair

advantage (of unfairness). And it does not follow that just because the universities are taking advantage of a background injustice to those who become student athletes that the university is taking unfair advantage of those student athletes. Of course, even if the individual institutions are themselves not acting unfairly given the extant background conditions, it would remain the case that the system of intercollegiate athletics is built upon these unfair background conditions. It would be society's responsibility (and here we could expect universities to play a major role) to change the background conditions so as to eliminate this systemic form of exploitation. On this view, student athletes may be exploited, but it may not be the universities that are solely or even primarily to blame.

Chapter Four

COMMERCIAL SURROGACY

emerged in the mid-1970s.1 By 1981, about 100 children were known to have been born to "surrogate mothers."2 By 1986, the number had reached 500. It then appeared that surrogacy might become a socially acceptable alternative to adoption for the large number (3.5 million—approximately 16% of couples of childbearing age) of infertile couples.3 While infertility has apparently increased because many women are deferring attempts at conception to ages at which they are less fertile, the supply of children available for adoption has declined for several reasons, including the legalization of abortion and the reduced social stigma attached to keeping a child born out of wedlock. In addition, commercial surrogacy is more attractive than adoption to those couples who want a child with whom they have at least a partial genetic relationship.4 Moreover, the prospect of serving as a surrogate seemed to appeal to a reasonable proportion of women. Of those questioned in a 1987 Gallup poll, 15 percent said that they would consider serving as a surrogate mother for the standard $10,000 fee.s Then came the famous legal contest between Mary Beth Whitehead and Elizabeth and William Stern over the custody of the child known COMMERCIAL SURROGACY

An earlier version of this chapter, titled "Two Questions about Surrogacy and Exploitation/' appeared in the Summer 1992 issue of Philosophy & Public Affairs. 1 Although it is important to distinguish between commercial and noncommercial (unpaid) surrogacy, I shall drop the adjective "commercial" in what follows except in those places where I want to draw the distinction. 2 Martha Field, Surrogate Motherhood (Cambridge: Harvard University Press, 1989), p. 5. Because it is in widespread use, I shall use the term "surrogate mother," even though she is typically the genetic and gestational mother and thus not a "surrogate." There are some cases in which the surrogate has served as a gestational but not genetic mother, where a fertilized ovum has been implanted in the surrogate. 3 See Peter H. Schuck, "Some Reflections on the Baby M Case," 76 Georgetown Law Journal 1793 (1988), at 1796. 4 Although surrogate mothers are most commonly impregnated through artificial insemination, in which case the infertile couple would have only a 50 percent genetic relationship with the child, some women can generate a fertilized ovum that will not become successfully implanted in the womb. In such cases, the genetic mother's fertilized ovum can be implanted in the surrogate, in which case the couple would have a complete genetic relationship with the child. 5 Field, Surrogate Motherhood, p. 5.

COMMERCIAL SURROGACY

97

as "Baby M."6 Public debate and opposition to commercial surrogacy intensified in the wake of the "Baby M" trial. Four states banned surrogacy contracts within a year, more have followed suit, and still other states, although not banning surrogacy contracts, will not enforce them.7 Although the practice of surrogacy appears to be on the run in the face of strong attacks on the political, legal, and scholarly fronts, the practice of surrogacy raises a host of deep and difficult issues of general importance to moral and political philosophy. Surrogacy has been criticized on many grounds. It has been argued that surrogacy is baby-selling, that it is harmful to the children born to surrogates, that it is harmful to many other children, whose sense of security is undermined by the practice, and that it is harmful to women as a class.8 Another line of arguments maintains that surrogacy involves the wrongful "commodification" of persons or relationships, or that it violates the Kantian maxim that persons should never be treated merely as means but always as ends in themselves.9 In addition to and often intermixed with these (and other) arguments, it is frequently said that surrogacy exploits the surrogate mothers: "[O]nce money enters into the arrangement [the] possibilities of exploitation are everywhere."10 "One of the most serious charges against surrogate motherhood contracts is that they exploit women."11 "The prohibition on payments may be understood as protecting . . . women—especially poor, single women—from being exploited . . . 6 The trial court ruled that Whitehead had entered into a valid contract. In re Baby M, 217 NJ. Super. 313 (1987). On appeal, the New Jersey Supreme Court sustained the trial court's decision with respect to the custody of Baby M, but held that unpaid surrogacy arrangements were permissible only if the surrogate mother were not under a binding agreement to surrender the child and that commercial surrogacy is a form of babyselling that violates the New Jersey statute on private adoption. In the Matter of Baby M, 109 NJ. 396, 537 A.2d 1227 (1988). 7 New York Times, March 12, 1989, sec. 1, p. 38. 8 For an argument that focuses on the effects of surrogacy on women as a class, see Debra Satz, "Markets in Women's Reproductive Labor," 21 Philosophy & Public Affairs 107 (1992). Apart from other effects on women as a class, the legitimation of surrogacy might impose psychic costs on many women, especially poor women, who would then have to treat the decision not to be a surrogate as a constant forgone opportunity to earn additional income for their family. See Henry Hansmann, 'The Economics and Ethics of Markets for Human Organs," 14 Journal of Health Politics, Policy and Law 57 (1989). 9 Elizabeth Anderson, "Is Women's Labor a Commodity?" 19 Philosophy and Public Affairs 71 (1990). For a critique of this view, see Richard J. Arneson, "Commodification and Commercial Surrogacy," 21 Philosophy and Public Affairs 132 (1992). 10 Peter Singer and Deane Wells, The Reproductive Revolution (New York: Oxford University Press, 1984), p. 125. 11 Field, Surrogate Motherhood, p. 25.

98

CHAPTER FOUR 12

paid 'breeding stock. . . / " "But the core reality of surrogate motherhood is that it is both classist and sexist: a method to obtain children genetically related to white males by exploiting poor women/' 13 "[T]he danger of exploitation of one human being by another appears . . . to outweigh the potential benefits in almost every case/' 14 Despite the frequency with which these exploitation claims are made, they are typically advanced without much analysis or argument. In this chapter, I propose to extend our analysis by taking a close look at these claims. I shall ask under what conditions surrogacy would be a case of harmful exploitation or mutually advantageous exploita-

tion and whether, as some argue, there are good reasons to think that surrogacy arrangements are typically nonconsensual. The discussion of commercial surrogacy will allow us to explore the relation between the claim that a practice is immoral for reasons that do not seem to involve allegations of exploitation and the claim that it is exploitative. It will also provide us with an opportunity to sharpen the distinction between the truth conditions of an exploitation claim and the moral force of exploitation. For even if surrogacy is exploitative, we must ask whether such exploitation provides good reasons for prohibiting or refusing to enforce surrogacy contracts. Let me say that I shall not, for the most part, be concerned with evaluating the critiques of surrogacy that do not involve allegations of exploitation or with defending any public policy positions. Nonetheless, and with respect to the moral force of exploitation, my analysis will suggest that it is more difficult to justify the prohibition or nonenforcement of surrogacy agreements on grounds of exploitation than many commentators seem to believe. Is SURROGACY EXPLOITATIVE?

Putting aside the possibility that surrogacy is not exploitative at all, there appear to be two possible answers. On one view, surrogacy is a case of harmful exploitation. The intended parents gain from the transaction while the surrogate is encouraged "to make a grave selfsacrifice to the broker's and adoptive couple's advantage."15 On a 12 Alexander Capron and Margaret J. Radin, "Choosing Family Law over Contract Law as a Paradigm for Surrogate Motherhood/' in Larry Gostin, ed., Surrogate Motherhood (Bloomington: Indiana University Press, 1990), p. 62. 13 George Annas, "Fairy Tales Surrogate Mothers Tell," in Gostin, Surrogate Motherhood, p. 43. 14 Mary Warnock, A Question of Life (Warnock Report on Human Fertilisation and Embryology) (Oxford: Blackwell, 1985), p. 46. 15 Anderson, "Is Women's Labor ," p. 87.

COMMERCIAL SURROGACY

99

second view, surrogacy is a case of mutually advantageous exploitation. Both the intended parents and the surrogate gain from the transaction but the distribution of those gains is unfair to the surrogate, say, because, "the interests of wealthy contracting couples are better served than those of the surrogates/' 16 On some versions of both of these views, surrogacy is exploitative because the intended parents gain from a transaction that is fundamentally immoral, perhaps because the relationship involves an exchange of radically incommensurate values, or because the transaction wrongly commodifies procreational labor. For want of a better term, call this moralistic exploitation.17 Because moralistic objections to surrogacy must involve some sort of harm to the surrogate if they are to ground a claim of exploitation, I shall consider harmful exploitation under the rubrics of nonmoral harm and moral harm before going on to consider mutually advantageous exploitation. SURROGACY AS HARMFUL EXPLOITATION

Nonmoral Harm

Rosemarie Tong says that surrogacy is harmful in straightforward nonmoral terms: "Since there is evidence that surrogacy arrangements . . . harm contracted mothers . . . a ban on commercial surrogacy needs to rely only on the harm principle [as opposed to legal moralism]."18 As it stands, this is a non sequitur. After all, if surrogacy is harmful but consensual, a ban on commercial surrogacy cannot rely only on some version of the Millian harm (to others) principle; it would need to rely on a principle of paternalism. But the present question is not whether the surrogate is harmed with her consent, but whether she is harmed at all. And here we must be careful. In deciding whether surrogacy is harmful, we must adopt an all-things-considered and ex ante conception of harm. There are, after all, negative elements in virtually all uncontroversially beneficial transactions. Paying money for a good that is clearly worth the price is still a negative element in the transaction. It would be better to get it for free. Similarly, we do not say that a worker is harmed by employment, although the worker may prefer 16

"Baby M and the Surrogate Motherhood Controversy," 37 American University Law Review 1013 (1987), at 1024. 17 I do not mean "moralistic" in its derisive sense, in which it connotes a narrow moral attitude; I use the term to refer to arguments that go to the intrinsic immorality of a practice rather than the gains and losses to the parties. 18 Rosemarie Tong, "The Overdue Death of a Feminist Chameleon: Taking a Stand on Surrogacy Arrangements," 21 Journal of Social Philosophy 40 (1990), at 47.

100

CHAPTER FOUR

leisure to work. We assume that the benefits received from employment are greater than the costs. Thus the question is not whether surrogacy has harmful elements, but whether it is harmful, all things considered, or a net harm. We know that some surrogates, such as Mary Beth Whitehead, have regarded their experience as a surrogate as harmful, all things considered.19 But this is largely irrelevant even if we assume that retrospective judgments are accurate indicators (and they may not be) in an individual case. For the question is not whether surrogacy is a net harm in a given case, but whether it is typically or ex ante a net harm. If a worker is severely injured on the job, such that employment is a net harm to that worker, we do not say that such employment is harmful as a practice. Similarly, the question is not whether an individual surrogate, such as Whitehead, is harmed, but whether the expected value of surrogacy is negative, where the expected value is a function of the probability distribution of the various outcomes. Surrogacy presents a particularly difficult context for making an "all things considered" judgment because some of the crucial benefits and harms are difficult if not impossible to measure and vary substantially from person to person. On the positive side of the ledger, we have to put the value of the monetary compensation, whatever psychological gratification the surrogate obtains from bringing a child into the world, and the happiness that she has brought to the intended parents. These benefits may or may not be considerable. There is much that might appear on the debit side of the ledger: the risk of physical harm or death resulting from the pregnancy or delivery, restraints on the surrogate's choices during pregnancy, the inconveniences and discomfort associated with a normal pregnancy, and perhaps most important, the psychic cost of the surrender of the baby to the intended parents.20 At this point, the defender of surrogacy might reply that the ex ante value of the surrogacy arrangement simply could not be negative, for if that were so, women would not agree to serve as a surrogates. But that is false. That a woman agrees to serve as a surrogate does not show that an accurate judgment of her ex ante value is positive; it only shows that she thought it would be positive. She may have miscalcu19

See Mary Beth Whitehead, A Mother's Story—The Truth about the Baby M Case (New

York: St. Martin's, 1989), a n d Elizabeth Kane, Birth Mother ( N e w York: Harcourt Brace Jovanovich, 1988). 20 A surrogacy contract may stipulate that the surrogate agrees to refrain from alcohol, tobacco, illicit drugs, and other risky activities. Although these restrictions are not independent harms to the surrogate—they may be beneficial—they do represent restrictions on her freedom and in that sense can be understood as costs.

COMMERCIAL SURROGACY

101

lated, perhaps because surrogates are unable to make accurate predictions of their future psychological reactions. Thus it is possible that surrogacy is harmful to most surrogates even though most surrogates believe it will be beneficial. In addition, it might be argued that most surrogates are objectively harmed by their experience, even if they do not feel harmed. It is a commonplace that B's interests—as B defines them—can be harmed even if B is unaware that the harm has occurred. For example, B's selfacknowledged interest in her reputation or the fidelity of her spouse can be damaged by libel or infidelity, even if she is unaware that either has occurred. More controversially, it may be argued that a person's objective interests can be harmed even if she does not now and never will regard these interests as her interests. She may suffer from false consciousness. Given all this, is surrogacy ex ante harmful, all things considered? In view of our limited factual knowledge and unresolved theoretical controversies over what counts as objective harm, I am inclined to think that we should now remain agnostic. Depending on the answers to these questions, surrogacy could be a case of harmful exploitation, a case of mutually advantageous exploitation, or not exploitative at all. Suppose that most surrogates are worse off for the experience, all things considered. Could surrogacy be made advantageous, all things considered, if the monetary compensation were high enough? If we are operating in the territory of the surrogate's nonmoral interests, I think it is entirely possible, nay, inevitable, that a sufficiently large increase in compensation would convert a net harm into a net benefit for many women. If surrogacy is typically harmful, all told, because "for what amounts to very little money for a nine-month, twentyfour-hour-a-day 'job,' the contracted mother risks and usually experiences a variety of physical but especially psychological harms," 21 this can be rectified by imposing an adequate minimum wage. Moreover, if surrogacy is typically mutually advantageous but unjust, this unfairness can be remedied in the same way. Yet, in contrast to other contexts in which it is uncontroversial that exploitation can be negated by increasing the value received by the exploited party, it is rarely argued that surrogacy would be less exploitative if the surrogate were paid more. In fact, unpaid surrogacy is often thought to be less exploitative than paid surrogacy. Among the critiques of surrogacy, higher pay is the dog that doesn't bark. 21 Tong, "The Overdue Death," p. 47. If the surrogate were compensated at the rate of $5.00 per hour for 24 hours per day for 270 days, the fee would be $32,400.

102

CHAPTER FOUR

Moral Harm

Why does the dog not bark? I suspect that a wage increase is rarely advanced as a solution to exploitation because it is thought that procreative labor should not be exchanged for money, because the cornmodification of procreative labor is immoral.22 And if procreative labor should not be exchanged for any money, it will not improve things to exchange procreational labor for more money. But if, as I have argued, the wrong of exploitation always involves a defect in the values exchanged, the task is to see whether this perspective can be related to exploitation. There are two ways in which the connection might be drawn: one focuses on the incommensurability of the values exchanged; the other focuses on commodification. It might be argued that commercial surrogacy involves the purchase of a good (a child) or service (procreational labor) that cannot be appropriately valued on a monetary scale, that the values exchanged are radically incommensurable. But an appeal to incommensurability, as such, does not take us very far toward supporting a claim of exploitation. First, even assuming that the values exchanged in a transaction are incommensurable, and it is not clear what that would mean in the present context, this sort of incommensurability does not establish that a market transaction is wrong. The permissibility of market transactions does not require that the goods exchanged be commensurable on a single metric. It requires that the parties transact voluntarily and, perhaps, with a (reasonable) belief that with respect to the parties' own criteria, the value received is at least as great as the value given. One can, and arguably should, be able to buy or sell a "priceless" painting without claiming that its value is commensurate with the money that is paid. Second, even when considerations of incommensurability provide a reason for thinking that a transaction is wrong, incommensurability does not entail that a party's interests are negatively affected by the transaction. Thus more will have to be said if incommensurability is going to support the claim that surrogacy is exploitative. The commodification argument maintains that whereas some goods and services (e.g., automobiles, houses, books, and at least some forms of labor) are appropriately exchanged for money, other goods and services (e.g., citizenship, human beings, criminal justice, marriage rights) should not be exchanged for money. To use Michael 22 It is also possible that higher pay is not advanced as a solution because the very receipt of monetary compensation may actually cause some of the psychological harm experienced by surrogate mothers, say, because they feel that they are doing something "sleazy/' If so, increasing the compensation may only make things worse on this score.

COMMERCIAL SURROGACY

103

Walzer's phrase, such transactions should be treated as "blocked exchanges/' 23 On this view, surrogacy is exploitative not because it comes too cheap, but because it commodities a form of labor that should not be exchanged for money at all.24 Now, I suspect that some of the rhetorical force of the commodification argument derives from linguistic aesthetics. Who could support something as ugly as the com-mod-i-fi-ca-tion of procreational labor? But let us assume arguendo that it is wrong to commodify procreational labor. It does not follow that surrogacy is harmful to the surrogate. To use one of Walzer's examples, let us assume that it was wrong to commodify exemptions from military service during the American Civil War, when citizens were allowed to pay $300 to purchase the services of a substitute.25 Although the commodification of exemptions may have led to a harm to the substitutes (I assume that the commodification was not harmful to those who bought exemptions), it is much less clear that the commodification of exemptions constituted a harm to them, qua commodification. The substitutes were harmed primarily because they were injured or killed, not because they were commodified. Similarly, even if a surrogate is harmed by the effects of the commodification of her labor, it does not follow that she is harmed by commodification qua commodification. Indeed, even if she experiences psychological harm because she feels commodified or degraded, it would not follow that she is harmed because she is commodified or degraded. Thus on the assumption that the commodification of procreational labor is wrong, we must ask whether the commodification of the surrogate's labor is (1) harmful to the surrogate's objective nonmoral interests, (2) harmful to the surrogate because she is participating in an immoral activity, or (3) wrong but not harmful to the surrogate. It might be thought that surrogacy is harmful to the surrogate's nonmoral interests because she is treated as a commodity. It might be claimed that a person can lose the respect of others or be degraded in their eyes even if she does not lose self-respect or become degraded in her own eyes. Thus to the extent that a person has an interest in the 23

See Michael Walzer, Spheres of Justice ( N e w York: Basic Books, 1983), p p . lOOff. "Commercial surrogacy attempts to transform what is specifically women's labor—the work of bringing forth children into the world—into a commodity. It does so by replacing the parental norms which usually govern the practice of gestating children with the economic norms which govern ordinary production processes. The application of commercial norms to women's labor reduces the surrogate mothers from persons worthy of respect and consideration to objects of mere use." Anderson, "Is Women's Labor," p. 80. 25 See Walzer, Spheres of Justice, p p . 98-99. 24

104

CHAPTER FOUR

way she is regarded by others, surrogacy may injure those interests. But that raises at least two points. First, it is not clear that surrogacy actually does have these effects. Second, to the extent that these effects stem solely from the way surrogacy is regarded by the society— as a matter of fact and without separate normative justification—it is not clear that they represent a basis for condemning the practice rather than a basis for condemning society's reaction. After all, homosexuality certainly was (and, to a lesser degree, still is) a basis for a loss of social respect, but this provides no reason to condemn homosexuality. Still, it might be argued that surrogacy harms the surrogate's objective interests because it because it violates her rights, independent of any other physical, economic, psychological, or social harm. A trespasser harms the property owner by violating the property owner's rights to exclusive use of the property, even though there is no "ordinary" harm to the property.26 A man violates a woman's rights by fondling her without her consent, even if the touching causes no physical pain or lasting psychological damage. Thus if a woman has a right not to have her labor commodified, then surrogacy is harmful precisely because it is a violation of her rights. The problem here, of course, is that many acts that would violate B's rights if done without B's consent do not violate B's rights if done with B's consent. It is no violation of a woman's rights if her property is entered with her consent or if she is fondled with her consent. To put the point in Kantian terms, and bracketing worries about false consciousness, we do not treat a person merely as a means rather than an end-in-herself if we treat her in a certain way only if she consents to be treated in that way. So the commodification of procreational labor is no obvious violation of the Kantian maxim if the commodification of the surrogate's labor is consensual—absent some additional argument, for example, that the rights involved are inalienable or that the consent is defective in some important respect.27 Suppose that we go so far as to grant that the commodification of procreational labor constitutes an objective harm to the surrogate. Still, it would not follow that surrogacy is harmful to the surrogate, all things considered. Surrogacy would produce a net harm to the surrogate only if the degree of harm that results from the commodification 26

See Joel Feinberg, Harm to Others (New York: Oxford University Press, 1984), ch. 1. It might be argued that surrogacy commodifies and violates the Kantian maxim with respect to the child, who, after all, does not consent to the arrangement. Although I do not see how surrogacy could be harmful to the child, who, after all, would not otherwise exist, the present discussion is confined to the claim that surrogacy is harmful to the surrogate mother. 27

COMMERCIAL SURROGACY

105

of her procreational labor is greater than the benefits that she received from the compensation and that has not been shown. Moreover, increasing the compensation would arguably offset such harm. We have been considering the claim that surrogacy is harmful to the surrogate because she is treated wrongly. But surrogacy might involve a different sort of objective or moral harm to the surrogate. It may be thought that surrogacy is injurious to the surrogate not because she is treated wrongly, but because she participates in something wrong. On this view, surrogacy is bad for the surrogate because it is bad for her character. But here we must be particularly careful. It does not follow that just because it is bad to participate in something wrong that it is bad for the person's interests to participate in something wrong. It is wrong to cheat on one's taxes, but it does not follow that it is bad for the cheater.. But it might be. To the extent that one does have an interest in one's moral character, then knowingly (or negligently) participating in an immoral activity is harmful to that person.28 The structure of the argument from moral harm might look something like this: 1. Surrogacy is wrong, say, because it is wrong to commodify procreational labor. 2. Because surrogacy is wrong, it is immoral for a woman to serve as a surrogate. 3. Participating in an immoral activity is bad for the participant because it harms her moral character. 4. Combining (1), (2), and (3), serving as a surrogate harms the character of the surrogate. 5. Because the surrogate is harmed by the transaction for the benefit of the intended parents, surrogacy is exploitative.

Two points about this line of argument. First, to the extent that commodification is harmful to the surrogate because it harms her character, it seems that it should also be harmful to the intended parents. So even if this argument supported the claim that the surrogate is harmed, it might not support the claim that she is exploited, because exploitation requires that the exploiter can reasonably expect to gain, which this view may seem to deny. And this raises the second point. The argument from moral harm allows us to claim that there are aspects of surrogacy that are harmful 28

Now, even if this sort of moral harm is possible, it is not clear to whom it applies. On the one hand, we might say, with Plato, that all persons have an interest in being moral whether or not they themselves believe they have such an interest. Or we might say, with Feinberg, that only those that desire to be good persons have an interest in being good persons and can therefore be morally harmed. See Feinberg, Harm to Others, pp. 65-70.

106

CHAPTER FOUR

to the surrogate. But assuming that surrogacy is otherwise advantageous to the surrogate, it is not clear that the harmful moral elements necessarily outweigh the beneficial nonmoral elements.29 Robert George, for example, argues for (3) when he says that "Every immoral choice . . . corrupts the chooser; every such choice integrates moral evil into the chooser's will, thus unreasonably damaging that aspect of the chooser's own well-being which consists in establishing and maintaining an upright moral character."30 But even if we were to accept George's claim that every immoral choice is damaging to the chooser's own well-being, even he implies that it is damaging only to "that aspect" of well-being that consists in upright moral character. So unless it is assumed that the moral aspect of well-being always trumps the nonmoral aspects of well-being, this leaves open the possibility that the nonmoral aspect in which the immoral choice is beneficial to the chooser's well-being is more weighty than the moral aspect in which it is not.31 In the final analysis, I am inclined to think that it is very difficult to argue that surrogacy is harmful to the surrogate because surrogacy is wrong qua commodification. I also think that there is something very odd about the attempt to understand commodification as a harm to the surrogate's character and to assess its place in overall harm and benefit to the surrogate. Commodification may be better understood as a basis for thinking that surrogacy is wrong for reasons unrelated to the interests of the surrogate and, therefore, unrelated to worries about exploitation of the surrogate. After all, the commodification of procreational labor may have undesirable effects on the offspring, women (as a class), or society as a whole. But suppose that I am wrong. Suppose that the commodification argument supports the claim that surrogacy is exploitative. The moral force of that argument would still have to be resolved. The commodification argument maintains that specific human goods, such as procreational labor, should be valued in particular ways and, in particular, that procreational labor should not be valued by the norms of the marketplace. Now I am prepared to grant as probably true, and 29 It is also not clear whether an increase in compensation to the surrogate would yield an increase in the amount of moral harm or whether the degree of moral harm is inelastic with respect to price. 30 Making Men Moral (Oxford: Clarendon Press, 1993), p . 168. 31 We could think that moral considerations t r u m p nonmoral considerations with respect to what B should do without thinking that moral considerations trump nonmoral considerations with respect to what is good for B. See Thomas Nagel's discussion of the connection between "living right7' and "living well" in The View from Nowhere (New York: Oxford University Press, 1986), pp. 189-207.

COMMERCIAL SURROGACY

107

certainly for the sake of argument, that the commodification of procreational labor fails to reflect the best "conception of human flourishing/' 32 But to anticipate the discussion of the moral force of exploitation, we should be cautious about using such an argument to support policies that interfere with surrogacy arrangements. As an empirical observation, I think it fair to say that people may not all share the same understanding of those goods and the way in which they should be valued. That does not mean that all understandings are equally right. But in a society committed to pluralism and tolerance, we should be reluctant to use what may be the best understanding of a human good to deny people "the opportunity to advance their ends by entering into voluntary relationships with others." And this even if their understanding of that good is incorrect.33 SURROGACY AS MUTUALLY ADVANTAGEOUS EXPLOITATION

Suppose that the typical surrogate is not harmed by surrogacy or would not be harmed if the compensation were higher. All things considered, surrogacy is or would be a mutually advantageous transaction. Still, surrogacy might be a case of mutually advantageous exploitation if the transaction were unfair to the surrogate. The problem, of course, is that if we are going to say that a transaction is exploitative because it is insufficiently beneficial to the exploited party, we may reasonably be asked to specify the criteria by which we are making this assessment. And as I have noted in previous chapters, this is a difficult matter. It is, for example, frequently said that a fair transaction is one in which both parties gain (roughly) equally. On that account, we can say that a surrogate is exploited if she receives less value from the transaction than do the intended parents. Unfortunately, this definition of a fair transaction is clearly wrong. If a physician performs a procedure (for a normal price) that saves a patient's life, we do not say that the physician has been exploited because the patient gained far more from the transaction than the physician did.34 In any case, and notwithstanding that I do not have an alternative account to propose at this point (I shall say more in Chapter 7), I see no reason to think 32

Anderson, "Is Women's Labor," pp. 72-73. J. D o n a l d M o o n , Constructing Community (Princeton: Princeton University Press, 1993), p . 19. 34 We also cannot say that a person is not exploited just because he or she receives more value from the transaction than does the other party. If a physician charges an exorbitant price for the procedure, we might say that the physician exploits the patient even though the patient receives more value from the transaction than the physician. 33

108

CHAPTER FOUR

that a mutually advantageous transaction cannot be unfair. So I prefer to suppose that the surrogate may be exploited even if surrogacy provides a net benefit to her. On the other hand, if the compensation is (or could be made) adequate, then we will have to conclude that the surrogate is not (or would not be) exploited, whatever else we may want to say about surrogacy. CHOICE

I suspect that there is another reason why an increase in compensation is rarely mentioned as a solution to the alleged exploitation of surrogacy, and that relates to its effect on the quality of the surrogate's consent. It may be thought that the lure of compensation compromises the surrogate's consent by coercing or forcing her to enter into the transaction or by distorting her judgment about the effects of surrogacy on her well-being. And if the lure of the standard compensation compromises the quality of the surrogate's consent by coercion or by generating cognitive errors, increasing the compensation will only make things worse. Now, resolving the question of whether the surrogate gives appropriately voluntary consent may not be crucial to resolving the question of whether she is exploited if, as I shall argue, consensual exploitation is possible. But it would still be of importance to determine whether there is a defect in consent. With respect to moral weight, a defect in consent would make the exploitation worse. With respect to moral force, it will be easier to justify the prohibition or nonenforcement of surrogacy agreements if they are nonconsensual. So let us consider whether the consent of the surrogate is typically defective because of coercion or cognitive errors. Coercion

Are surrogates coerced? Many commentators seem to think so: "[A monetary offer] may be difficult for a person of little financial means to refuse and would, in that case, be coercive."35 On this quite familiar view, poor women are coerced (or forced) to become surrogates because surrogacy represents an improvement over an unacceptable 35

Ruth Macklin, "Is There Anything Wrong with Surrogate Motherhood?" in Gostin, Surrogate Motherhood, p. 146. "To portray surrogacy contracts as representing meaningful choice and informed consent on the part of the contracting surrogate mother, rather than to see her as driven by circumstances . . . fails to take account of realities." Field, Surrogate Motherhood, p. 27.

COMMERCIAL SURROGACY

109

status quo.36 The equally familiar response maintains that surrogacy offers women an additional option to their present menu of choices, and the addition of options to one's menu of choices is always freedom enhancing rather than coercive. Which view is correct? On any standard account of coercion, surrogacy is simply not coercive. In general, A coerces B to do X only if A proposes (threatens) to make B worse off with reference to some baseline condition unless B does X, but A does not coerce B to do X if A proposes (offers) to make B better off with reference to some baseline condition if B does X.37 Whatever else we might want to say about the intended parents' proposal to a potential surrogate (offers can be unseemly—"indecent proposals''—without being coercive), they do not propose to make her worse off if she turns it down. True, she might decide that becoming a surrogate is more attractive than her other options. But that is true for all decisions in which one accepts an offer. And we cannot say, as some have argued, that women are forced to become surrogates just because they would not prefer to serve as a surrogate were it not for the compensation, for on the assumption that many persons would prefer not to work were it not for the compensation, then virtually all labor is coerced.38 It may be thought that the preceding view fails to acknowledge that some offers are simply "too good to refuse," and that nonrefusable offers are coercive even if they add to one's present menu of options.39 Now an offer may be nonrefusable for two reasons, only one 36 "To s a y t h a t a w o m a n " c h o o s e s ' t o d o t h i s . . . i s s i m p l y t o s a y t h a t w h e n a w o m a n is forced to choose between poverty and exploitation, she sometimes chooses exploitation as the lesser of two evils." Tong, "'The Overdue Death," p. 45. For present purposes, I ignore the distinction between being coerced, or forced, or acting under duress. 37 Specifying the appropriate baseline against which to measure the proposal can be a complicated question. See my Coercion (Princeton: Princeton University Press, 1987), chs. 12 and 13. 38 And so say Marxist critics of capitalist society. Given that the proletariat has two alternatives—work or starve—it follows that they are coerced into working. If this were so, surrogacy might exemplify a more general economic coercion, but there would be no reason to regard surrogacy contracts as particularly coerced or involuntary. It is often noted, in this connection, that we allow men and women to be employed in positions that are unpleasant and dangerous without barring such employment on grounds of coercion. Indeed, an important recent Supreme Court case—hailed as a victory by feminists—was based on the right of women to be employed under work conditions that pose a serious risk to their offspring. See United Auto Workers v. Johnson Controls, 499 U.S. 187 (1991). 39 Joel Feinberg argues that a proposal can be both coercive and "freedomenhancing" in Harm to Self (New York: Oxford University Press, 1986), p. 233. I discuss this view in Coercion, pp. 232-33.

110

CHAPTER FOUR

of which is problematic. Suppose that A offers B an opportunity (a new job) that would render her so much better off than her eminently acceptable status quo that it would be irrational for B to refuse. I see no reason to refer to such offers as coercive or to think that they compromise the voluntariness of B's choice. In other cases, however, the short-term benefits contained in A's offer may be so tempting or irresistible that they cause B to overlook the long-term harms. A does not threaten any adverse consequence if B declines the offer, but A's offer serves to distort B's judgment. I am inclined to think that this kind of irresistible offer is not best described as coercive, because it does not make any preexisting options less eligible, but it may well compromise the voluntariness of B's choice by introducing cognitive errors into B's decision. Let us bracket worries about cognitive errors. Suppose that B's choice is eminently reasonable, given (or precisely because of) her background conditions. Still, it might be said that if B's status quo is highly unsatisfactory, then A's offer may be coercive even if B makes a perfectly rational choice in accepting A's offer. Consider one of Robert Nozick's examples (which I condense): The Drowning Case. A comes upon B, who is drowning. A proposes to rescue B if B agrees to pay A $10,000. A and B know that there are no other potential rescuers.40

Here we might want to say that A's offer is coercive even though B makes a perfectly rational decision to pay $10,000 to be rescued. Why is this so? It might be thought that A's proposal is coercive because B has no "acceptable" alternative. I do not think that this is the right account.41 If A (a physician) should say to B (a patient), "You can choose to have this leg amputated or you will die," we do not say that B's decision to have his leg amputated is coerced just because death is an unacceptable alternative. Rather, we seek B's informed consent to the procedure, something we could not do if such consent were impossible. I suggest that A's proposal in the drowning case may be properly described as coercive not because B has no acceptable alternative, but because A may have an obligation to improve B's situation without compensation, or at least for much less. Although A proposes to improve B's situation relative to B's status quo, A actually proposes to make B worse off as compared to what B has a right to expect from A 40 Robert Nozick, "Coercion," in Sidney Morgenbesser, Patrick Suppes, and Morton White, eds., Philosophy, Science and Method (New York: St. Martin's, 1969), p . 447. I discuss this example in Wertheimer, Coercion, p . 207. 41 For a more extended discussion, see Wertheimer, Coercion, ch. 11.

COMMERCIAL SURROGACY

111

(or what A has a duty to do for B). Relative to a conception of B's baseline that reflects those moral judgments, A's putative offer can reasonably be regarded as a coercive threat. But this model of coercive offers is of no help to the argument that surrogates are coerced. For, unlike A in the drowning case, the intended parents have no special obligation to help potential surrogates without demanding anything in return. It seems, then, that the intended parents' proposal is an offer, and offers do not coerce. And that, I think, is the truth. But it is not the whole truth. Let us assume what is often alleged but has not been shown by the critics of surrogacy, namely, that women choose to serve as surrogates only because they are in dire economic straits. It may be argued that when background conditions provide an inadequate range of opportunities, the moral quality or significance of one's choice is diminished, even if the background conditions do not compromise the "voluntariness" of the choice, strictly speaking.42 But even if this is so, it is arguable that it is the background conditions that are the problem and not the offer that allows B to improve on those background conditions. The offer is still a positive good. But, it might be said, suppose that the potential surrogate has a right to a better range of opportunities, even if it is not a right against the intended parents. Does it not follow that such agreements are coercive because society has a duty to provide options that would make surrogacy unattractive? We might want to say this, but whatever we decide to say, a fundamental problem will remain. If a woman can reasonably regard surrogacy as improving her overall welfare given that society has unjustly limited her options, it is arguable that it would be adding insult to injury to deny her that opportunity. Whatever label we use to describe her choice, we must still decide whether she should be allowed to make such a choice. And referring to such choices as "coerced" will not resolve that substantive moral question. Cognitive Errors

Considerations of coercion aside, I suggested that a woman's decision to serve as a surrogate may not be appropriately consensual if the lure of the financial gain motivates her to make a decision that she will regret, or would regret if she thought objectively about its effect on her life. And, it may be said, to knowingly or negligently take advan42 See T. M. Scanlon, "The Significance of Choice/' in Tanner Lectures on Human Values, Vol. 8 (Salt Lake City: University of Utah Press, 1987).

112

CHAPTER FOUR

tage of a given woman's tendency to make such mistakes is a form of exploitation. But does the typical surrogate's choice reflect a cognitive error? If the typical surrogate agrees to surrogacy because she believes that the ex ante value of surrogacy to her is positive and the ex ante value of surrogacy to her is, in fact, positive, then there is no cognitive error (bracketing miscalculations of the magnitude of the benefit). On the other hand, the typical surrogate may come to regard surrogacy as harmful, all things considered. In this case, we would have good reason to doubt whether surrogates were giving sufficiently informed or rational consent. Given that we do not know whether surrogacy is typically harmful or advantageous, I do not think that we can say whether the typical surrogate makes this kind of cognitive error. And do not say that a woman's judgment cannot be appropriately consensual just because she cannot fully anticipate what it will be like to give up the child. People can voluntarily consent to sterilizations, sex change operations, abortions, and plastic surgery, and (shall we add?) marriage— where one cannot or frequently does not have any experience with the consequences of the decision. By comparison with some such decisions that we do allow, the problem of miscalculation in surrogacy may be relatively small and more amenable to preventive measures: restricting surrogacy to women who have given birth and therefore have personal knowledge of the bonding process (which would not have excluded Whitehead) and careful psychological screening (which might have excluded her).43 In any case, if the typical surrogate will regret her acceptance of A's proposal because the intended parents fail to disclose relevant information about the likely consequences of her decision, then they have exploited her capacity as a decision-maker or chooser. The intended parents would be treating the surrogate's decision-making capacity merely as a means to their own ends. Suppose instead that the woman's inability to anticipate the harmful consequences is not due to a lack of "external information" or opportunity for deliberation, but is the result of her own cognitive and emotional limitations. She simply underestimates the degree to which she will feel attached to the child.44 Would that constitute exploitation? It might. If the intended 43 It appears that an infertility center had performed a psychological evaluation of Whitehead and that she had "demonstrated certain traits that might make surrender of the child difficult." Unfortunately, it also appears that neither she nor the Sterns were given this information. Matter of Baby M, 537 A.2d 1227 (N.J. 1988), at 1247. 44 John Lawrence Hill argues that offers are exploitative when they intentionally or negligently take advantage of a "significant impairment of the rational-emotional ca-

COMMERCIAL SURROGACY

113

parents correctly believe that the surrogate is likely to regret her decision, it may well be unfair for them to knowingly or negligently take advantage of the surrogate's inability to anticipate that regret. The intended parents may have a quasi-fiduciary obligation to protect the potential surrogate from making choices that she will subsequently regret. The claim that the surrogate makes a cognitive error would be more controversial if her error derived from her failure to take proper account of objective or moral harms to her interests. Suppose that the surrogate will be degraded or corrupted by surrogacy, but that she does not care about these forms of harm. Or suppose that she would regard herself as harmed if she were degraded or corrupted, but she does not think that surrogacy is degrading or corrupting. Do such women suffer from cognitive defects that constitute a defect in the quality of their consent? I am not sure what to say about this, but even if we correctly believe that the surrogate's choice is nonconsensual for just this reason, it is less clear what would follow. Here we return to the moral force of exploitation. THE MORAL FORCE OF EXPLOITATION

Suppose that we come to the view that surrogacy is typically exploitative. What follows with respect to societal intervention with commercial surrogacy? It depends. It depends, in part, on whether surrogacy is a case of harmful exploitation, moralistic exploitation, or mutually advantageous exploitation. If surrogacy is a case of harmful exploitation, we need to ask whether paternalistic restrictions would be justified, although here our task will be relatively easy, particularly if we have reason to think that there are defects in the quality of the surrogate's consent. If surrogacy is a case of harmful cum moralistic exploitation, we need to ask whether restrictions can be justified on grounds of moral paternalism. And that would be at least somewhat harder. For it is one thing to interfere with a woman's decision if she miscalculates the effects of her decisions in terms of her own values or stable preferences; it is quite another to interfere when we believe (even correctly) that she has failed to understand what constitutes a harm to her objective interests. It is a commonplace that a liberal democracy is not justified in prohibiting transactions just because the transactions are morally suspect or fail to incorporate the best conception of human flourishing. And this not because we lack confidence parity" of the decision-maker. "Exploitation," 79 Cornell Law Review 631 (1994), at 68384.

114

CHAPTER FOUR

that the transactions are truly wrong, but because there are reasons— good moral reasons—to prefer a political regime that does not regard such wrongness as a sufficient justification for invoking the coercive powers of the state. If we are justified in prohibiting or refusing to enforce surrogacy agreements that are cases of harmful exploitation, it is not clear what role the exploitation is playing in the justification of the restriction as contrasted with the harm. So we are in a better position to consider the moral force of exploitation if we focus on mutually advantageous and consensual exploitation. Here it seems that we should start from the assumption that consenting adults should be free to engage in mutually advantageous transactions in the absence of compelling reasons to the contrary, then it does not follow that just because it is wrong for A to exploit B, we should prohibit A from exploiting B. Perhaps it is true, as Robert Goodin has maintained, that "there are some uses to which people should not be put even with their consent, even if it is full, informed, genuine consent." And it may be true, as he suggests, that this includes a decision to "sell a cornea or rent a womb." 45 But it does not follow that just because there are some uses to which people should not be put even with their consent that it is right to deny them the opportunity to be so put. So if exploitation is to serve as an independent basis for social restraints, we would need to explain how the exploitativeness of a transaction justifies interfering with a transaction from which the exploited party benefits, particularly if prohibition is the only viable option. Reducing Exploitation

But prohibition may not be the only viable option. Suppose that surrogacy would be less exploitative (or nonexploitative) if the compensation were higher or the contract terms were somewhat different. Might this justify societal intervention? It might, but not in order to rectify the effects of an inequality of bargaining power. Unlike some other contexts of alleged exploitation that are thought to involve such inequalities, I see no reason to believe that the compensation received by surrogates reflects a structural inequality of bargaining power between the intended parents and the surrogate.46 It seems to me that the potential surrogate 45

Reasons for Welfare (Princeton: P r i n c e t o n University Press, 1988), p . 141. Allen Wood has maintained that prospective adoptive couples are typically in a "significantly stronger bargaining position" in "Exploitation," 12 Social Philosophy and Policy 136 (1995), at 144. I see no reason to think that this is so. 46

COMMERCIAL SURROGACY

115

should be in a very strong negotiating position. Because the intended parents may desire the services of the surrogate at least as much as she desires the money, she should be in a position to command a relatively high price for her services. If this is so, why don't surrogates receive greater compensation? It is possible, I suppose, that the value of children to the adoptive parents may greatly exceed their financial capacity to reflect that value in their offer. More important, I suspect that surrogates receive rather low compensation precisely because surrogacy is viewed with such moral skepticism. Information and competition are low. And unlike many other bargaining contexts, in which bargainers believe they can legitimately press for a better deal, social norms may suggest that it is inappropriate for a surrogate to hold out for a higher wage. As a result, the terms contained in surrogacy contracts may be much closer to the reservation price of the surrogate than to the reservation price of the intended parents. Somewhat ironically, the widespread acceptance of the moral norms to which the critics of surrogacy often appeal may contribute to this dimension of its exploitativeness. If we do not effectively allow an open market in surrogacy, then potential surrogates may face a collective action problem. Left to their own devices, most potential surrogates might be able to negotiate an agreement that is less desirable than would be negotiated if all were permitted only to negotiate more favorable agreements. It is entirely possible that many surrogacy agreements that would occur at a $10,000 fee would also occur at a $50,000 fee if the government were to establish that as the minimum compensation level. Some agreements would not occur. Some intended parents would be unwilling or unable to pay the higher price. But many would.47 We would then face a problem of moral trade-offs. We would need to weigh the moral importance of reducing the exploitation of those who would be helped by this form of minimum wage legislation against the cost to those who would be excluded from the market because people refuse to purchase their services for the higher minimum wage. Yet the latter point highlights a moral consideration that many critics of surrogacy have ignored. Any restrictions that discourage surrogacy arrangements hurt those who would have benefited from such arrangements and arguably also interfere with their autonomy. If we are to respect the autonomy of potential surrogates in the world in which they find themselves, it is important that we not automatically apply to them the moral principles that define our conception of an 47

Indeed, it is possible that there would be more surrogacy agreements at the higher price, because it would increase the supply of potential surrogates.

116

CHAPTER FOUR 48

ideal world. As Margaret Radin has argued, even if a principle of "market—inalienability would protect all things important to personhood'' in a world in which background conditions are just, it may be morally preferable to allow the alienation of things important to personhood, such as procreational labor, in our actual world.49 Autonomy

The question arises then as to whether considerations of autonomy would preclude using exploitation as grounds for restricting mutually advantageous and consensual surrogacy arrangements. An adequate answer to this question would require a more fully developed theory of autonomy than I can offer. But I do want to argue that two strategies—one conceptual, one legal—that have been advanced as solutions to this problem are less than fully satisfactory. The conceptual strategy maintains that since a woman's "true" freedom or autonomy is violated by surrogacy contracts, the prohibition of such contracts does not constitute a violation of her freedom or autonomy.50 In slightly different terms, it is said that if we adopt a "positive" rather than a "negative" conception of freedom, treating a woman's procreative labor as inalienable does not diminish her freedom.51 To the contrary, it enhances her freedom. I do not want to deny that there is something to the notion of positive freedom or, perhaps more accurately, to the values it attempts to capture (for those values may be better captured by other terms). But even if we say that "true" freedom includes proper self-development, and if we think that surrogacy "would detract from the ideal of human flourishing that society should seek to foster," it remains an open question whether the right to choose not to be positively or truly free is itself a crucial dimension of one's autonomy.52 48

For the distinction between ideal theory and nonideal theory, see John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971), pp. 8ff. 49 M a r g a r e t R a d i n , "Market-Inalienability," 100 Harvard Law Review 1903 (1987). A s Radin adds, "If we think respect for persons warrants prohibiting a mother from selling something personal to obtain food for her starving children, we do not respect her personhood more by forcing her to let them starve instead." Ibid., pp. 1910-11. 50 A s Elizabeth A n d e r s o n puts it, prohibiting surrogacy contracts d o e s n o t violate the a u t o n o m y of w o m e n because " t h e content of t h e surrogate contract itself c o m p r o mises t h e a u t o n o m y of surrogate m o t h e r s . " A n d e r s o n , "Is W o m e n ' s Labor," p . 9 1 . 51 "if w e a d o p t a positive view of liberty that includes p r o p e r self-development as necessary for freedom, then inalienabilities needed to foster that development will be seen as freedom-enhancing rather than as impositions of unwanted restraints on our desire to transact in the markets." Radin, "Market-Inalienability," p. 1899. 52 A l e x a n d e r C a p r o n a n d Margaret J. Radin, " C h o o s i n g Family Law over Contract Law a s a Paradigm for Surrogate M o t h e r h o o d , " in Gostin, Surrogate Motherhood, p . 64.

COMMERCIAL SURROGACY

117

If we are tied to a maximizing notion of freedom in which we aggregate the total package of negative and positive freedoms (assuming such aggregation is possible), then we may promote a woman's overall freedom by limiting her freedom to serve as a surrogate. But we may have reason not to adopt a maximizing view. We may think that one should be free to choose not to maximize one's freedom. Even if Mill is right that one should not be free to sell oneself into slavery, he does not defend a maximizing view. Perhaps we should not be free to make decisions that involve an irrevocable and total abdication of our freedom or capacity to act as an autonomous moral agent.53 But as a general proposition, it simply is not true that we should not be free to give up our freedom. There is another way to argue that prohibiting women from choosing to become surrogates does not compromise their autonomy. Elizabeth Anderson has argued that surrogacy "takes advantage of motivations—such as self-effacing 'altruism'—which women have formed under social conditions inconsistent with genuine autonomy."54 And, the argument goes, we do not interfere with a woman's genuine autonomy if we prohibit decisions that result from nonautonomously acquired motivations. I shall pursue the issue of false consciousness in more detail in Chapter 8. Suffice it to say, for present purposes, that even if the surrogate's motivations are not autonomous in this sense, it is not clear what would follow. We certainly do not think that we can justifiably prevent religious organizations from soliciting or accepting donations just because we (even rightly) believe that the motivations that give rise to those donations may have been formed under social conditions inconsistent with genuine autonomy.55 Respect for a person's autonomy sometimes requires that we respect choices that reflect values that the person presently accepts, even if we are rightly worried about the way he or she acquired those values. So much for the conceptual strategy. A popular legal strategy for resolving the tension between respecting autonomy and prohibiting surrogacy agreements is to permit surrogacy transactions, but to "make the arrangement performable or not at the option of the mother."56 On this view, it is one thing if the surrogate wants to con53

On Liberty, c h . V. Anderson, "Is Women's Labor/' p. 91. 55 I have in mind contributions to "normal" religious organizations, not contributions that are clearly the product of fraud. Interestingly, to the extent that surrogacy takes advantage of altruistic preferences, it does not represent a problem of cornmodification. After all, it is one thing to maintain that surrogacy does not represent a "gift" relation, and another thing to assume that it does, but that women have been involuntarily socialized into giving such gifts. 56 Field, Surrogate Motherhood, p . 78. 54

118

CHAPTER FOUR

summate the transaction; it is quite another if the surrogate irrevocably commits herself to surrendering the child. It might be thought that this strategy would give us the best of both worlds: it would preserve the freedom of women to enter into surrogacy arrangements, but would also preserve their freedom to keep the child if they so wish.57 Now, it is entirely possible that the "unenforceable contract" solution will turn out to be the preferred public policy.58 Even so, it would not entirely resolve the autonomy problem. There are two related ways in which it fails to do so. First, it fails to acknowledge that the ability to enter into a binding agreement is itself a crucial dimension of one's autonomy. If B wants to enter into a surrogacy arrangement with A, but A will not do so unless B's agreement is enforceable, a decision not to enforce such agreements constrains B's ability to make a decision she prefers. This may be a justifiable restraint on B, but it is undeniably a restraint on B. Second, it may be argued that in trying to protect women from having to surrender a child against her strong maternal desires, we do not express the appropriate respect for women as autonomous and responsible persons.59 Here the worry is not that such policies compromise the autonomy of surrogates, although that may also be true. Rather, the thought is that even if many potential surrogates would benefit from a more protective public policy, such a policy would effectively "exclude women from full-fledged membership in human society."60 Whether or not this line of argument is ultimately decisive, and it is entirely possible that it is not, it suggests that the unenforceable contract solution does not fully resolve the tension between protecting surrogates from decisions that they may come to regret and respecting the autonomy of women. 57

"One attractive feature of this solution is that it helps avoid resolution of the debate, which is currently dividing feminists, about whether surrogacy exploits women or liberates them. It recognizes the truth of both positions. Surrogacy is still available when the surrogate mother desires ultimately to carry out the contract. . . . But it avoids one of the most troubling features—a contract severing the maternal bond when the mother is unwilling to relinquish her child." Ibid. 58

See Michael Trebilcock a n d R o s e m i n K e s h v a n i , " T h e Role of Private O r d e r i n g in Family Law: A Law a n d Economics P e r s p e c t i v e , " 41 University of Toronto Law Journal 533 (1991).

59 "Her state of mind at the moment of agreement is not to be taken seriously because it is subject to change during the performance of her undertaking, due to the nature of pregnancy. The insinuation is that it is unreasonable to expect her to keep her promise because her faculty of reason is suspended by the emotional facets of her biological constituency." Carmel Shalev, Birth Power (New Haven: Yale University Press, 1989), p. 121. 60

Ibid., p . 122.

COMMERCIAL SURROGACY

119

Justice

Finally, I wish to consider a set of arguments for interfering with surrogacy contracts that appeals to considerations of justice or inequality (and I shall not distinguish between them here), the values that are at the core of our concern with exploitation. I say the set of arguments "appeals to" considerations of justice, because these arguments variably assert that surrogacy instantiates injustice (i.e., that surrogacy arrangements are unjust), that surrogacy derives from injustice, that surrogacy symbolizes injustice, that surrogacy reinforces injustice. Note that some of these verbs suggest that surrogacy has a causal impact on the social world, whereas others do not, and it may be crucial to establish just which verb is most apt.61 Surrogacy could instantiate an injustice, derive from injustice, or symbolize injustice without causing the world to be more unjust. By contrast, the claim that surrogacy reinforces or creates injustice is to claim that the world would be less unjust if surrogacy were prohibited. With these considerations in mind, there are at least three different lines of justice-based arguments that we might consider. First, it might be argued that we should prevent transactions that instantiate unjust distributions even if, given the unjust background conditions, the transactions are beneficial to all concerned. On this view, it is wrong to allow unjust transactions to occur, even when (as contrasted with the world in which they do not occur) these particular transactions are not bad for anyone. Thomas Nagel puts it this way: "Inequality, even if it harms no one, counts as something bad in itself. . . [and] this may provide a reason to reject a Pareto-superior alternative because the inequality it permits is too great to be outweighed by other advantages/' 62 I do not think it unreasonable for an individual to refuse to participate in a transaction that will improve her welfare on the grounds that the transaction is unjust. But the question is not whether an individ61 "[Distributive justice requires that society's benefits and burdens be distributed fairly among different social classes. . . Since women who are less well off will almost always be the ones to serve as surrogates for wealthier or professional women, the distribution is not fair. . ." Ruth Macklin, "Is There Anything Wrong with Surrogate Motherhood?" in Gostin, Surrogate Motherhood, p. 147. "Surrogate motherhood is a method to obtain children genetically related to white males by exploiting poor women . . . it subverts any principled notion of economic fairness and justice, and undermines our commitment to equality. . . . " George Annas, "Fairy Tales Surrogate Mothers Tell,"

in G o s t i n , Surrogate Motherhood, p . 4 3 . 62 Equality and Partiality ( N e w York: O x f o r d U n i v e r s i t y P r e s s , 1991), p . 107. It is n o t

clear whether Nagel's claim is meant to apply to particular practices, such as surrogacy, or only to what Rawls calls the "basic structure of society."

120

CHAPTER FOUR

ual can herself reasonably refuse to participate in a beneficial transaction because it is unjust, but whether society can justifiably prevent her from participating in such a transaction on the grounds that it is unjust. Indeed, after suggesting that inequality is bad even when it works to the benefit of the worse-off, Nagel himself quickly adds that this unfairness may be overridden by "countervailing factors," including benefits to the worse-off. So we cannot assume that the intrinsic badness of inequality would justify the prohibition of surrogacy. A second justice-based argument focuses on the relational dimension of surrogacy. It is one thing, it might be said, if the rich can buy material goods that the poor cannot; it is quite another if the rich can purchase this highly personal service from the poor. On this view, the problem is not that surrogacy results in different "holdings," but that surrogacy instantiates highly asymmetrical and unjust personal relations. Relational inequality is a serious matter, and it must be included in any "all things considered" assessment of whether a woman would be benefited or harmed by entering into a surrogacy transaction. But if a woman can still plausibly maintain that she would benefit by such an arrangement, it is hard to see why such transactions should be prohibited on the grounds that the relation is unjust—at least if the welfare of the potential surrogate is the focus of our concern. And that gives rise to a third line of argument, one that focuses on the negative externalities of commercial surrogacy. It is, for example, often argued that pornography has harmful effects on virtually all women because it reduces their perceived value as persons, even if pornography is not harmful, all things considered, for those women who are portrayed in pornography and who are compensated for their services.63 Similarly, it may be argued that surrogacy has harmful effects on women as a class, because it reinforces inequalities of gender, although it might not be harmful, all things considered, to the surrogates themselves, who are compensated for their services.64 I want to make three points about this line of argument. First, if we argue that a (potential) surrogacy arrangement between A and B 63 "The mass production of pornography universalizes the violation of the w o m e n in it, spreading it to all w o m e n , w h o are then exploited, used, abused, and reduced as a result of men's consumption of it." Catharine MacKinnon, Toward a Feminist Theory of the State (Cambridge: Harvard University Press, 1989), p. 247. 64 A s Debra Satz puts it, "[Surrogacy] contracts will turn women's labor into something that is used and controlled by others and will reinforce gender stereotypes that have been used to justify the unequal treatment of w o m e n , " "Markets in Women's Reproductive Labor," pp. 123-24.

COMMERCIAL SURROGACY

121

should be prohibited because a policy of allowing such arrangements would have harmful effects on third parties, then we are not claiming that it is the exploitation of B that justifies its prohibition. Second, even if the prohibition of surrogacy has egalitarian consequences, we would be imposing a burden on one class of women (those who would be benefited by serving as surrogates) in order to benefit a larger class of women. We are often justified in imposing costs on some persons in order to realize gains to other (present or future) persons. But it is harder to justify imposing such costs when they are incurred by those who are thought to be among the least well-off. In addition, when the costs are incurred in the present and the gains are to be realized only in the future, we should be reasonably confident that such gains will be realized before such costs are imposed. That brings me to the third point, namely, that we have no empirical evidence to support the claim that the prohibition of surrogacy actually has egalitarian consequences or—what amounts to the same thing—that the permissibility of surrogacy would reinforce or perpetuate social inequalities. For all we know, the legitimation of commercial surrogacy might serve to empower women in their relations to men.65 Indeed, it is arguable that prohibiting surrogacy might simply shield our eyes from the background inequalities from which it derives, while allowing surrogacy would forcefully bring those inequalities to our attention and motivate us to change them. Suppressing mutually advantageous exploitative relationships may be akin to removing the homeless from public places; we may feel better, but the problem persists. CONCLUSION

Where does this leave us with respect to commercial surrogacy? Actually, nowhere. This is as good a place as any for me to remind the reader that this part of the book is not concerned with saying whether a particular practice is exploitative or defending a definite position on legal or public policy questions. Rather, my goal is to use the analysis of exploitation claims within specific contexts to shed light on several theoretical issues about the truth conditions of exploitation claims, their moral weight, and their moral force. In the context of the present chapter, then, I am not concerned with whether commercial surrogacy is typically a case of harmful exploitation, mutually advantageous exploitation, or not exploitative at all. To defend any such claim would require us to resolve a host of factual 65

See Shalev, Birth Power.

122

CHAPTER FOUR

and theoretical disputes that are beyond my enterprise. Still, I hope to have shed some light on the nature of the factual and theoretical disputes that would have to be resolved if we were to defend such claims. I have also not defended any position with respect to whether commercial surrogacy agreements should be permitted or enforced. Not only would such a defense require that I resolve the justmentioned unresolved disputes, but it would also require us to consider many arguments about commercial surrogacy that I have entirely ignored. Still, I hope to have shown that even if surrogacy is a case of mutually advantageous and consensual exploitation, the moral force of such exploitation is much less clear than is often supposed.

Chapter Five

UNCONSTITUTIONAL CONDITIONS

of this chapter is to continue the discussion of the moral force of exploitation. I do so by capitalizing on the structural similarities between mutually advantageous exploitation and an important doctrine in American constitutional law. The doctrine of "unconstitutional conditions'' maintains that it may be unconstitutional for a government to transact with a (potential) recipient of governmental benefits under conditions that appear to be Pareto superior. More specifically, the doctrine maintains that there are cases in which it is unconstitutional for the government to provide a benefit only on condition that the recipient waive a constitutional right even though the government would violate no constitutional right if it chose not to provide this benefit at all. For example, the doctrine of unconstitutional conditions maintains that a state cannot provide welfare assistance on condition that the recipient agree not to support a Democratic candidate even though the state has no constitutional obligation to provide welfare assistance in the first place.1 Justice Holmes, the most famous opponent of the doctrine, argued that if the government is not required to provide certain benefits, then the government must have a right to provide those benefits on condition that the recipient waive a constitutional right: "Even in the law the whole generally includes its parts/' 2 In effect, Holmes claims that there are no reasons to interfere with mutually advantageous and consensual transactions between the government and recipients of governmental benefits, even if they are exploitative. Although Holmes's view has the ring of plausibility, it is (virtually) universally rejected as constitutional law. By considering the reasons that might justify rejecting Holmes's claim, and here we can draw upon a rich body of cases and commentary, we can deepen our understanding of the reasons for characterizing some mutually advantageous transactions as exploitative and the reasons that might justify prohibiting or restricting such transactions.

THE MAJOR PURPOSE

1

Edward J. Fuhr, "The Doctrine of Unconstitutional Conditions and the First Amendment," 39 Case Western Law Review 97 (1989), at 107. 2 Western Union Telegraph Co. v. Kansas, 216 U.S. 1, 53 (1909) (dissenting).

124

CHAPTER FIVE THE ANALOGY REFINED

To develop the analogy between exploitation and unconstitutional conditions, it will be useful to think of a standard case of mutually advantageous exploitation as one in which B waives, transfers, or suspends a right in exchange for a benefit provided by A. For example: 1. Exploitative Wage. (1) A has the right not to employ B at all. (2) B has a right not to work for A. (3) A offers B employment at a low wage. (4) B accepts, thereby suspending his or her right not to work for A in exchange for a (low) wage. 2. Exploitative Price. (1) A has the right not to provide a service to B (e.g., fixing B's car). (2) B has a right not to give money to A. (3) A proposes to fix B's car for an exorbitant price. (4) B accepts, thereby transferring his or her (ownership) right to the money in exchange for A's services. 3. Student Athletes. (1) A has the right not to offer an athletic scholarship to B. (2) B has a right not to play football for A. (3) A offers B an athletic scholarship on condition that B play football for A. (4) B accepts, thereby suspending B's right not to play football in exchange for the athletic scholarship. 4. Surrogacy. (1) A has the right not to enter into a surrogacy arrangement with B. (2) B has a right to custody of any child she bears. (3) A proposes that B enter into a surrogacy arrangement with A, on condition that B waive her custody right to a child who results from the insemination. (4) B accepts, waiving her right to a child to whom she gives birth. We might say that A's transaction with B is exploitative in some or all of these cases, even if we concede that (1) is true in each case. Just because A violates no right of B if A refuses to offer a benefit to B (wage, service, scholarship, money), it does not follow that A can properly give or that A should be allowed to give such a benefit to B on any terms that are mutually advantageous and consensual. Structurally similar relations can occur between the government and (potential) recipients of governmental benefits. A government may extend what appears to be a discretionary benefit to the recipient on condition that the recipient waive or not exercise a constitutional right. For example: 1. Insurance Co. v. Morse. A Wisconsin statute required a foreign corporation (i.e., a corporation chartered in another state) to agree not to resort to federal courts if it wished to transact business within the state. Let us assume that (1) a state may refuse admittance to foreign corporations entirely, and (2) a corporation has a right to remove suits to the federal courts (when appropriate). Yet it was argued that Wisconsin could not

UNCONSTITUTIONAL CONDITIONS

125

require a corporation to waive its right to the use of federal courts as a precondition of transacting business in the state. 3 2. Speiser v. Randall. California provided property tax exemptions to all veterans on condition that they sign a loyalty oath. Let us assume that (1) California has the right not to provide tax exemptions to veterans, and (2) Speiser has the right not to sign a loyalty oath. 4 Yet it was argued that California could not provide a tax exemption on condition that Speiser waive his constitutional right not to sign a loyalty oath. 3. Sherbert v. Verner. South Carolina provided unemployment compensation for persons who were willing to accept suitable employment. Sherbert, a member of the Seventh Day Adventist Church, was discharged from her job because she would not work on Saturday. Because she was unwilling to accept employment that required Saturday work, she was denied unemployment compensation. Let us assume that (1) South Carolina has the right not to provide unemployment compensation, and (2) Sherbert has the right not to work on Saturday.5 Yet it was argued that South Carolina could not make receipt of unemployment compensation contingent on Sherbert's being available to work on Saturday because South Carolina's policy compromised her "free exercise" of religion. 4. South Dakota v. Dole. In 1984, Congress enacted a statute directing the Secretary of Transportation to withhold a portion of federal highway funds from states that sell alcoholic beverages to persons under age twenty-one. Let us assume that (1) the federal government has the right not to subsidize highway construction, and (2) under the Twenty-First Amendment, a state has the right to set its drinking age below the age of twenty-one. Yet it was argued that the federal government cannot make receipt of highway funds contingent on a state's setting its drinking age at twenty-one. 6 5. Lyng v. Automobile Workers. A 1981 amendment to the Food Stamp Act states that no household shall become eligible for food stamps during the time that any member of the household is on strike, even if, because of the loss of income due to the strike, the household would otherwise qualify for food stamps. Let us assume that (1) the government has the right not to provide food stamps at all, and (2) a worker has the right to 3

87 U.S. 445 (1974). The Court held that the condition was unconstitutional. 357 U.S. 513 (1958). The Court held that the condition was unconstitutional. 5 347 U.S. 398 (1963). The Court held that the condition was unconstitutional. 6 483 U.S. 203 (1986). Section 2 of the Twenty-First Amendment provides: 'The transportation or importation into any State, Territory, or possession of the United States for delivery or use therein of intoxicating liquors, in violation of the laws thereof, is hereby prohibited." This provision is often understood to entail that each state has the right to set its own laws with respect to the consumption of liquor. South Dakota permits persons nineteen years of age or older to purchase beer containing up to 3.2 percent alcohol. The Court held that the condition was constitutional. 4

126

CHAPTER FIVE

strike. Yet it was argued that the government could not make receipt of food stamps contingent on not exercising the right to strike.7 6. Harris v. McRae. The "Hyde Amendment" to Title XIX of the Social Security Act, commonly known as the "Medicaid" Act, provided that federal funds, otherwise available to a state for medical procedures, could be used to fund the cost of prenatal care and delivery, but could not be used for abortions unless the failure to perform an abortion would endanger the life of the mother or the pregnancy had resulted from rape or incest.8 Let us assume that (1) the federal government has the right not to provide any funds for medical care, and (2) a woman has the right to obtain an abortion. Yet it was argued that the federal government cannot make funds for medical care available on condition that a woman choose not to exercise her right to obtain an abortion.9 An appeal to the doctrine of unconstitutional conditions maintains that the government's policy may be unconstitutional even if we concede that (1) is true in each case. The doctrine maintains that just because the government violates no right of the recipient if it refuses to offer the benefit, it does not follow that the government can offer a benefit on condition that the recipient waive or not exercise a constitutional right. As it is sometimes said, the government cannot motivate the beneficiary to "surrender by agreement rights that the government could not take by direct action/' 1 0 Needless to say, the Court has hardly been unanimous about these issues. Whereas a majority thought the Hyde Amendment was perfectly constitutional, a minority vehemently disagreed. But my aim is not to resolve disputes as to the best interpretation or application of the doctrine. Rather, I seek to consider the arguments that have or might be offered in defense of the doctrine in order to shed light on the concerns of this book. THE DEVELOPMENT OF THE DOCTRINE

The doctrine of unconstitutional conditions had its origins in several comparatively mundane nineteenth-century cases. One set of cases involved an attempt by some states to prevent corporations from removing legal controversies from state courts to federal courts. 11 Al7

485 U.S. 360 (1988). The Court held that the condition was constitutional. 448 U.S. 297 (1980). 9 The Court held that the "Hyde Amendment" was not unconstitutional. 10 Richard A. Epstein, "Unconstitutional Conditions, State Power, and the Limits of Consent," 102 Harvard Law Review 4 (1988), at 6. For an expanded version of this argument, see his Bargaining with the State (Princeton: Princeton University Press, 1993). 11 See Maurice H. Merrill, "Unconstitutional Conditions," 77 University of Pennsylva8

nia Law Review 879 (1929).

UNCONSTITUTIONAL CONDITIONS

127

though this was an objective that no state could accomplish directly (Wisconsin could not simply prohibit a corporation from seeking to have its case heard in federal court), a state could try to accomplish this objective indirectly. Since, ex hypothesi, a state had the "greater power" to refuse admittance to a corporation chartered in another state, it could exercise its "lesser power" to exact an agreement not to resort to federal courts as the price for doing business in the state. In Insurance Co. v. Morse, the Supreme Court rejected Wisconsin's claim that the "greater power includes the lesser power." Why could not Wisconsin extend a benefit to a corporation (the permission to do business in Wisconsin) only if the corporation would trade the right for the benefit?12 Because, said the Court, A man may not barter away his life or his freedom, or his substantial rights . . . he may omit to exercise his right to remove his suit to a Federal tribunal, as often as he thinks fit. . . . He cannot, however, bind himself in advance by an agreement . . . to forfeit his rights at all times and on all occasions. . . ,13 Although the Court appears to invoke the concept of inalienable rights here—"A man may not barter away . . . his substantial rights"—we should not take this rhetoric too seriously. The Court did not claim that states could never "require a surrender of constitutional privilege in return for favors granted."14 Rather, the Court seems to be arguing that it will not allow Wisconsin to entice a corporation to trade this right (the right to sue in federal court) for this privilege (the corporation's ability to do business in Wisconsin). Why not? The Court's language suggests that a rights waiver of limited duration might be constitutional, but that an irrevocable waiver is different. But it is by no means clear that the Court would have accepted Wisconsin's policy if the waiver were renewable on an annual basis. On closer inspection of the opinion, the Court seems more concerned with consequences than with rights. It appears that the Court held the practice unconstitutional because the use of state bargaining power to discourage use of federal courts was "inconsistent with the 12

A dissenting opinion argued that the company could and did bargain away its rights: "[T]he right to impose conditions upon admission follows, as a necessary consequence, from the right to exclude altogether. . . . The insurance company accepted this condition. . . . Having received the benefits of its renunciation the revocation comes too late." 87 U.S. 445, at 458, 459. 13 Ibid., at 451. Here, the Court's argument parallels a distinction between adoption and surrogacy. Although we may think it permissible to allow a mother to agree to waive her custody right to her child at the time of the adoption, we may think it wrong to allow her to bind herself in advance to do so, to forfeit the right in its entirety. 14 Merrill, "Unconstitutional Conditions," p. 881.

128

CHAPTER FIVE

efficient functioning of the governmental system set up by the Constitution."15 Although the relationship between the undesirable consequences and the unconstitutionality of a policy is not entirely clear, the Court was similarly concerned with the consequences of state policies in other cases involving "exchanges" between states and corporations.

Consider Western Union Telegraph Co. v. Kansas.16 A Kansas statute re-

quired that a foreign corporation that wished to do business in Kansas pay a tax on its total capital (including assets that were not held or employed in Kansas) "for the benefit of the permanent school fund." Although Kansas could not have levied a tax on Western Union if the company had not been operating in the state, Kansas asserted that its greater power to deny the corporation entry into the state entailed its lesser power to tax the corporation on its non-Kansas assets as a price for doing business in the state. The Court held that the statute constituted an unacceptable burden on interstate commerce: "It is easy to see that if every State should pass [such] a statute . . . the freedom of interstate commerce would be destroyed, the decisions of this court nullified, and the business of the country thrown into confusion. . . ,"17 Because Western Union was already operating in Kansas when the statute was passed, the statute could have been viewed as an attempt to retroactively "confiscate" the property of an ongoing business rather than as a demand for an exorbitant price to enter the state to do business. Yet the Court does not appear to rest its decision on the "confiscation" interpretation, for its concerns about the statute's effects on interstate commerce would also arise under an "entry price" interpretation. But Justice Holmes did not see the problem under either interpretation: Now what has Kansas done? She had not undertaken to tax Western Union. She has not attempted to impose an absolute liability for a singledollar. She simply has said to the company that if it wants to do local business it must pay a certain sum of money. . . . If the state may prohibit, it may prohibit with the privilege of avoiding the prohibition in a certain way.18

Holmes is famous for claiming that "the life of the law has not been logic: it has been experience."19 But here it might seem that he is ap15

Ibid. 216 U.S. 1 (1910). 17 Ibid., at 27. 18 Ibid., at 53 (emphasis added). 19 Oliver Wendell Holmes, The Common Law 1 (1881). 16

UNCONSTITUTIONAL CONDITIONS

129

pealing to logic alone. For Holmes, the question is not whether such statutes, if allowed, would have undesirable consequences. The question is whether (1) Kansas's right to exclude a corporation entails that (2) Kansas has the right to admit that corporation on any basis it chooses. If (2) follows from (1), and Holmes could not see how this could be denied, there is no reason not to allow Kansas to demand that the corporation waive or trade this right in exchange for a benefit that Kansas was under no obligation to provide: "I confess my inability to understand how a condition can be unconstitutional when attached to a matter over which a State has absolute arbitrary power. "20 It might be thought that Holmes was more reluctant to endorse the "greater includes the lesser" principle with respect to fundamental civil liberties than with respect to economic matters. Not so. When a policeman had been fired for violating a regulation that restricted his political activities, Holmes maintained that the state's greater power not to hire the man as a policeman included the lesser right to hire him on condition that he waive the constitutional right to engage in political activity. As Holmes quipped from his earlier position on the Massachusetts bench, "The petitioner may have a constitutional right to talk politics, but he has no constitutional right to be a policeman/'21 Suppose we shift our perspective to the recipients of such proposals. Rather than ask whether the state has the right to demand that a recipient waive a right as a condition of receiving a discretionary benefit, do we not trivialize a citizen's constitutional rights when we allow the citizen to barter them in exchange for benefits? Holmes thought not, even when contemplating something as important as one's right to talk politics: "There are few employments for hire in which the servant does not agree to suspend his constitutional right of free speech, as well as of idleness, by the implied terms of his contract."22 To treat constitutional rights as inalienable, says Holmes, is simply absurd: "In order to enter into most of the relations of life, people have to give up some of their constitutional rights."23 And so it would appear that Holmes's logic entails that we not find unconstitutional conditions in any of the cases noted above.24 20

Western Union Tel. Co. v. Kansas, 216 U . S . 1 (1910), a t 54. Mccauliffe v. Mayor of New Bedford, 155 M a s s . 216, 220, 29 N . E . 517, 517 (1892). 22 Ibid., at 29 N.E. at 517-18. 23 Power Mfg. Co. v. Saunders, 247 U . S . 490 (1927), a t 497. 24 As one commentator has remarked, the Constitution may not have enacted Mr. Herbert Spencer's Social Statics, but it seems to have enacted pareto optimality. Albert J. Rosenthal, "Conditional Federal Spending as a Regulatory Device," 26 San Diego Law Revieiv 277 (1989), at 284. Rosenthal is, of course, referring to Holmes's famous dissent in Lochner v. New York, 198 U.S. 45, 75 (1905). 21

130

CHAPTER FIVE THE CHALLENGE

We have seen that there are two polar views about the possibility of unconstitutional conditions. On the Holmesian view, the right of the state not to grant a benefit must include the right of the state to grant a benefit on terms that it wishes and to which the recipient gives consent. On this view, the doctrine of unconstitutional conditions is a nonstarter. On the second view, the government "cannot do indirectly what it cannot do directly/' If the state cannot require a citizen to waive or not exercise a constitutional right, then it should not be able to motivate the citizen to do so indirectly or by the extension of benefits. Both views are overstated. Consider first the second view. There are many objectives the state can legitimately pursue "indirectly" (noncoercively) that it cannot pursue "directly" (by coercion). I take it to be uncontroversial that the government can encourage people to have children or buy homes by providing tax deductions for children or for interest on mortgages, although such policies serve to discourage the exercise of one's right not to reproduce or not to buy a home. Moreover, and from the beneficiary's perspective, one of the values of having a right is precisely that it can be "traded" for other goods. As Richard Epstein puts it: "If an individual values his constitutional rights less than he does the privilege that the state is offering, then why not allow the bargain to go forward? Persons have a constitutional right to keep property, yet they surrender it all the time in order to obtain alternative goods that they value more highly."25 So there is no reason to assume that there should be a general bar against policies that discourage the exercise of constitutional rights. If the second view is overstated, it is also true that the Holmesian view has been decisively rejected as a matter of constitutional law. Showing how and why that view can be justifiably rejected is a major task of the analysis to follow, and I will not try to state it simply here. But there are two moves for rejecting Holmes's view (one set forth by Kenneth Simons, the other by Cass Sunstein) that I will reject at this point. Simons has argued that we can reject the doctrine that the greater includes the lesser because the "lesser" power is "greater" in one significant respect—it more greatly affects the exercise of a constitutional right.26 Although Simons is importantly right to note that the "lesser" power to provide a benefit with conditions is a more efficacious mech25

Epstein, "Unconstitutional Conditions," p. 48. "Offers, Threats, a n d Unconstitutional Conditions," 26 San Diego Law Review 289 (1989), at 293. 26

UNCONSTITUTIONAL CONDITIONS

131

anism for altering the behavior of citizens than the "greater" power not to provide a benefit at all, the "greater includes the lesser" doctrine is not an empirical claim about the measurement of power. It does not claim that the "greater" power is more powerful. Rather, it is a logical or normative claim about the relations between powers of various scope, namely, that the right to exercise the "greater" or "larger" power entails that one has a "right" to exercise the "lesser" or "smaller" power. So the problem remains. A second way to solve the problem is to alter the background assumptions that give it form. Sunstein has argued that the doctrine of unconstitutional conditions is an anachronism because it wrongly assumes that the "common law and criminal prohibition [provide] the basic legal ordering framework, and that regulatory and spending programs are occasional and somewhat jarring additions."27 On this view, it is false and ahistoric to assume a baseline of the government's providing no benefits and then ask whether the government can provide benefits in a way that encourages people to waive their constitutional rights. Instead, we should jettison the doctrine of unconstitutional conditions and take a more direct route in which we examine the "nature of the incursion on the relevant right" and "the legitimacy and strength of the government's justification for such an incursion."28 There are two reasons for not following Sunstein's suggestion, at least at this point. First, even if the doctrine of unconstitutional conditions should be abandoned in the final analysis, we learn something about the justifications for barring pressure on those rights by examining the arguments for the doctrine. Second, even if the unconstitutional conditions doctrine is anachronistic in the context of the modern state, the arguments for the doctrine may prove relevant to the nonconstitutional contexts where common law entitlements do obtain, for those are the contexts in which most allegations of exploitation seem to occur and in which we are the most interested. And so we are back with Holmes. How can we justify limitations on the right of the government to extend benefits only on condition that constitutional rights be waived? How can we justify limiting the freedom of potential recipients to waive constitutional rights in order to attain discretionary benefits? And how can we distinguish the cases in which government can constitutionally extend benefits in exchange for the waiver of constitutional rights from those in which it cannot do so? 27

" W h y t h e Unconstitutional Conditions Doctrine Is a n A n a c h r o n i s m , " 70 Boston University Law Review 593 (1990), at 595. 28 Ibid, at 605.

132

CHAPTER FIVE UNCONSTITUTIONAL CONDITIONS AS COERCION

In Frost and Frost Trucking Co. v. Railroad Commission, 29 the Court pre-

saged a line of argument to which it appealed in several subsequent cases, to wit, that the government's offer of a benefit on condition that the recipient waive a constitutional right is unconstitutional when it coerces the prospective recipient into waiving that right. The situation was this. A California statute provided that commercial use of the highways be regulated by the railroad commission, which, in turn, required that any carrier secure a permit as a "common carrier for compensation." Frost and Frost Trucking Company was a private carrier; it transported citrus fruit for a single client between fixed locations. It was charged with violating California law because it had not secured a permit as a public carrier. The state conceded that Frost was a private carrier, but maintained that since the state had the authority to prohibit all trucking—including private trucking—on the highways, it could extend the privilege of using the highways to Frost on condition that Frost dedicate its property to the quasi-public use of public transportation. Frost was free to reject the state's offer, of course. But then it would not be "entitled to the privilege of using the highways." In arguing that California's policy was unconstitutional, Justice Sutherland granted that the state has the "greater power" to regulate the use of public highways, but also maintained that California could not use this power of exclusion to make the use of the public highways contingent on acceptance of public carrier status, for such a policy is inherently coercive. If the law were to stand, then constitutional guarantees, so carefully safeguarded against direct assault, are open to destruction by the indirect, but no less effective, process of requiring a surrender, which, though in form voluntary, in fact lacks none of the elements of compulsion. . . . In reality, the carrier is given no choice, except a choice between the rock and the whirlpool— an option to forego a privilege which may be vital to his livelihood or submit to a requirement which may constitute an intolerable burden.30 Schematically, Sutherland's argument is this:

1. The state cannot constitutionally coerce a recipient into waiving a constitutional right. 2. Some government offers of benefits are coercive. 3. When a government offer is coercive, the offer is unconstitutional. 29

271 U.S. 583 (1926), at 593. Ibid., at 593. And, said Justice Sutherland, "If the state may compel the surrender of one constitutional right as a condition of its favor, it may, in like manner, compel a surrender of all." Ibid., at 593. 30

UNCONSTITUTIONAL CONDITIONS

133

The logic of the argument is perfectly valid. The problem, of course, is to determine if the second proposition can ever be true and, if so, when. Is Sutherland right to argue that the California law was coercive? There are two reasons to think that he is not. First, if the argument were sound (and not only logically valid), the doctrine of unconstitutional conditions would be otiose. If the state's proposal coerced people into waiving their rights, it would be unconstitutional on that ground alone. There would be no need to develop a distinct doctrine of unconstitutional conditions. But the Supreme Court was right to think that it could not reject the California law simply on the ground that it was coercive, because— and this is the second problem—the argument from coercion is extremely difficult to sustain. In claiming that the California statute lacked "none of the elements of compulsion," Sutherland's argument adumbrated three difficulties that have characterized the appeal to coercion in unconstitutional conditions cases: (1) it identifies coerciveness with motivational efficacy; (2) it assumes that what I shall call hard choices (where one must choose between "a rock and a whirlpool") are coercive; and (3) it ignores the distinction between threats and offers. First, motivational efficacy. Sutherland claims that offering a discretionary benefit in exchange for the waiver of a right is an "indirect, but no less effective, process of requiring a surrender" of that right. Similar arguments have been made in other cases. In Speiser, Justice Brennan maintained that the "deterrent effect" of the policy of not providing tax exemptions to those who refuse to sign a loyalty oath "is the same as if the State were to fine them for this speech . . . the denial . . . will have the effect of coercing the claimants to refrain from the proscribed speech."31 In Harris, Brennan (dissenting) argued that the Hyde Amendment "serves to coerce indigent pregnant women to bear children that they would otherwise not elect to have. . . ." 32 And this distribution of funding, he said, is coercive because the distribution of governmental benefits "can discourage the exercise of fundamental liberties just as effectively as can outright denial of those rights through criminal and regulatory sanctions."33 Although it is difficult to discern precisely what is being argued in these cases, several Justices can be understood as claiming that the coerciveness of a policy is a function of its motivational efficacy. As it 31

357 U.S. 513, at 518, 519 (emphasis added). 448 U.S. 297 (1980), at 330. 33 Ibid., at 334 (emphasis added). Or, as Justice Marshall (dissenting) put it, "the differential distribution of incentives . . . can have precisely the same effect as an outright prohibition." Ibid., at 347 (emphasis added). 32

134

CHAPTER FIVE

stands, this claim simply begs the question of whether a proposal that is as effective at motivating behavior as a paradigmatically coercive proposal is similarly coercive by virtue of that fact. That is precisely what is at issue, and I think that claim should be rejected. To see why coercion is not a simple function of motivational efficacy, let us consider the distinctions among several forms of interpersonal power. For present purposes, and as a rough approximation, let us say that A exercises power over B when A gets B to do something that B would otherwise not do. A can exercise power over B in several ways. A exercises coercive power when A proposes to make B worse off than some baseline condition if B performs or fails to perform some act. A exercises inducive power over B when A proposes to make B better off than some baseline condition if B performs or fails to perform some act. A exercises persuasive power over B when A gives B reasons to perform or not perform some act. Coercive proposals seem to reduce the options available to B, whereas inducements and persuasion do not. When A makes a coercive proposal, "Give me your wallet or I will break your arm/' A makes a present option (keeping one's wallet) less eligible than it previously was. If A induces B to do X by offering a reward, then B now has an option (doing X plus the reward) that was previously unavailable, but A does not make any present option (not doing X) less eligible.34 If A persuades B to stop smoking by giving B reasons ("Smoking will give you cancer"), A has not affected the options available to B for better or for worse. A has simply given B reason to prefer one option (not smoking) to the other (smoking). There is no reason to think that inducements and persuasion are always less effective in changing behavior than coercion. Or so say the expressions "You can catch more flies with honey than with vinegar" and "The pen is mightier than the sword." But these expressions do not suggest, much less logically entail, that there are no significant moral differences between inducements (honey) and threats (vinegar), between the pen (persuasion) and the sword (coercion). If A can get B to waive his or her right not to mow A's lawn by offering B an attractive wage, this does not show that A might just as well have threatened to break B's arm if B refuses to mow the lawn. If A gets B to stop smoking by writing a pamphlet detailing the harmful effects of smoking, it does not follow that A might just as well have fined B for smoking. So even if it were true that the government's offer of re34 This is not to say that B would always welcome inducive offers or that people always prefer more freedom. They do not. People may sometimes prefer not to have to decide among options, and they may be corrupted or exploited by offers that they freely accept.

UNCONSTITUTIONAL CONDITIONS

135

wards were more effective than the threat of penalties in motivating people to waive a constitutional right, that would not show that the offer of such rewards is coercive. And that brings us to the second component of the argument that unconstitutional conditions are coercive, to wit, that when the state puts the potential beneficiary to a particularly hard choice, the recipient's decision is voluntary "in form/' but not "in reality." In Frost, Justice Sutherland made much of the fact that the company had"an option to forego a privilege which may be vital to his livelihood or submit to a requirement which may constitute an intolerable burden." 35 A similar problem impressed the dissenters in Harris, namely, that indigent women had to choose between (1) carrying the pregnancy to term (thereby receiving medical care), in which case they would have to raise the child or place the child for adoption, and (2) trying to secure an abortion without the necessary funds. And if indigent women choose (1) because (2) is intolerable, then it makes sense to regard the choice of (1) as having been coerced. As we have seen, the brute fact that a party faces "a choice between the rock and the whirlpool," where both choices are unattractive and the less attractive choice is extremely unattractive, does not entail that the resulting choice is coerced.36 Consider a criminal defendant who would prefer not to testify at trial, say, because he or she may have to implicate a friend (or commit perjury). After hearing the prosecutor's very strong case, the defendant comes to the conclusion that the only chance of avoiding conviction is to testify. We understand what the defendant means by "I was forced to testify." But we would certainly not say that the prosecutor coerced the defendant into testifying against him/herself or that the prosecutor had violated the defendant's Fifth Amendment rights.37 For constitutional purposes, the choice to testify was perfectly voluntary.38 35

271 U . S . 583, a t 593. As I suggested in the previous chapter, although we understand the sense in which a patient has "no choice" but to undergo surgery, given that death is an intolerable alternative, we do not think that this fact coerces the patient into consenting or invalidates the patient's "informed consent." Also see my Coercion (Princeton: Princeton University Press, 1987), chs. 10-13. 36

37

T h i s p o i n t w a s m a d e b y t h e S u p r e m e C o u r t in Brady v. U.S., 397 U . S . 742 (1970), at

750. Moreover, the friend whom the defendant testifies against might also not accept the claim that the defendant had been coerced. See Harry Frankfurt, "Coercion and Moral Responsibility," in Ted Honderich, ed., Essays on Freedom of Action (London: Routledge and Kegan Paul, 1973), p. 65. 38 Because this paragraph may seem harsh, let me say that I do not deny that there may be important moral distinctions between certain sorts of "hard choices" and choices between an array of acceptable options. I do maintain that the simple fact that

136

CHAPTER FIVE

A third line of argument for the view that unconstitutional conditions are coercive implicitly accepts the importance of the distinction between threats and offers, but maintains that the government's proposal should be understood as a threat to penalize the assertion (or nonwaiver) of a right. In Speiser, Justice Brennan wrote that to deny an exemption to those who refuse to sign the loyalty oath is "in effect to penalize them for such speech/' 39 In Sherbert, he wrote that to ask a citizen to abandon "one of the precepts of her religion" in order to qualify for the benefit "puts the same kind of burden upon the free exercise of religion as would a fine."40 By translating the offer of a governmental benefit to those who waive a constitutional right into a threat to fine or penalize the assertion (or nonwaiver) of a right, it can be argued that the government is coercing citizens into waiving their constitutional rights. Does this argument work? One might try to make the argument go by appealing to the notion of opportunity costs. As economists are fond of noting, if we think of B's choice situation as including all of the opportunities that are available, then a decision not to accept an offer has genuine opportunity costs. If A offers B $25 to mow A's lawn, then the cost to B of not accepting A's offer is $25. Relative to that baseline, B is being penalized for not mowing A's lawn. But this example does not show that opportunity costs are penalties. Rather, it shows that the opportunity cost argument proves too much. Although the notion of an opportunity cost captures important features of decision problems, B is simply not penalized for not mowing B's lawn. And that is because we have a noncontroversial "baseline" from which to understand A's proposal, namely, B's status quo. Relative to that baseline, A's proposal is an offer. If offers and threats were intertranslatable, we could view the paradigmatic armed robber case as one in which, having threatened to kill B, the robber is now offering B the benefit of life in exchange for B's money. But here, too, we have a noncontroversial baseline—B's situation prior to encountering the robber. And relative to that baseline, the robber's proposal is a threat. Baselines are not always so uncontroversial. We can argue over the best view of B's baseline. Still, if we are to argue that the government's policy is coercive, we cannot just apply a conclusory label, such as "penalty" or "fine" to the government's actions; we need to one faces a choice between two undesirable options does not establish that one's acceptance of a proposal is coerced in the sense that invalidates one's consent or violates one's moral or legal rights. 39 357 U.S. 513, at 518. 40 347 U.S. 398, at 404.

UNCONSTITUTIONAL CONDITIONS

137

argue for an understanding of the beneficiary's baseline from which it makes sense to think of the government's action as a penalty or a fine.41 If the argument goes through, then we can say that the policy is coercive. If not, not. How do we establish the best view of a person's baseline? I have argued elsewhere that for the legal and moral contexts in which the validity of B's consent is at issue, the specification of B's baseline is a moralized enterprise. Although the matter is actually somewhat more complicated, the crucial question is whether A's declared unilateral plan—what A proposes to do if A's terms are not met—is something that A has the right to do. 42 From this perspective, the robber's proposal is coercive not because B has to choose between two undesirable options, but because the robber's declared unilateral plan (to kill B) is something the robber has no right to do. Relative to what B has a right to expect from A, A proposes to make B worse off. In some cases, however, A's proposal to allow B to improve upon B's status quo can also be understood as a threat. Recall the discussion of Nozick's Drowning Case, in which A proposes to rescue B if B agrees to pay A $10,000.43 If A is morally required to rescue B in this case, as I am inclined to think, then A is proposing to harm B relative to the baseline to which B is entitled, and hence A's proposal is coercive. Thus, if B were to have signed an agreement in response to A's proposal, the agreement would not be valid because the contract would have been made under duress. Let us apply this framework to some of our cases. If California is required to allow Frost to use the highways as a private carrier, then its proposal is coercive. If not—and that was the opinion of Justice Holmes—then the proposal is not coercive. If, as Justice Clark argued (dissenting) in Speiser, California was simply refusing to extend "legislative largesse" to veterans who would not sign the loyalty oath, say, on the ground that the government should not feed the hand that bites it, then its policy was not strictly coercive.44 If, as Justice White argued in Lyng, the government's policy does not make strikers worse off, but "merely declines to extend additional foodstamp assistance," 41

Kathleen Sullivan, "Unconstitutional Conditions," 102 Harvard Law Review 1413 (1989), at 1420. Sullivan argues at some length that a focus on coercion "is unlikely to be helpful" precisely because setting the baseline is a "peculiarly elusive task." 42 Wertheimer, Coercion, esp. chs. 12 and 13. I borrow the term "declared unilateral plan" from Vinit Haksar, "Coercive Proposals," 4 Political Theory 65 (1976). 43 See Chapter 4. 44 "Neither fine nor imprisonment is involved . . . appellants are free to speak as they wish, to advocate what they will. If they advocate the violent and forceful overthrow of the California government, California will take no action against them . . . it will refuse to take any action for them." 357 U.S. 513, at 540-41.

138

CHAPTER FIVE

then the government's policy is not strictly coercive.45 If, as Justice Stewart argued in Harris, the "Hyde Amendment leaves an indigent woman with . . . the same range of choice . . . as she would have had if Congress had chosen to subsidize no health care costs at all," then that policy, too, is arguably not coercive.46 As I have suggested, we can argue over the best understanding of one's baseline. As I noted above, it can be argued that it is a mistake to start from the "no benefit" baseline, thereby regarding all benefits as governmental largesse. This is fair enough. Still, there is a good deal of space between the claim that a benefit should be understood as beneficent largesse and the claim that recipients are entitled to that specific benefit. So even if we should reject a pure largesse baseline, there is no reason to assume that we can substitute an entitlement baseline in all of the cases in which we want to question the government's right to condition a benefit on the waiver of a constitutional right. Interestingly, it is not clear that the Justices who invoked the language of coercion in unconstitutional conditions cases really meant it. Although the cases are replete with references to "coercion" and cognates, such as "penalties," "fines," and "deprivations," those words are usually accompanied by modifiers, such as "in effect" or "the same kind of burden," suggesting that the courts do not regard the supposed coercion as the real McCoy. Of course, if the sort of coercion implicated in unconstitutional conditions is not the genuine article, then the Court has also been right to think that the doctrine is not otiose. To assume that only coercive interferences with constitutional rights can be unconstitutional is to assume what the doctrine of unconstitutional conditions denies. The task is to see on what basis it can be denied. UNCONSTITUTIONAL CONDITIONS AS A BARGAINING PROBLEM

The Holmesian "greater includes the lesser" doctrine can be understood as a doctrine about the discretion of the state. It can also be understood as a variant of the principle of Pareto permissibility. This view maintains that if parties are entitled to remain in their status quo position in a bargaining situation, we should not prohibit them from engaging in Pareto-superior transactions. In this section, I propose to consider the doctrine of unconstitutional conditions as a bargaining problem. I shall consider a set of structural and distributional prob45 46

458 U.S. 360, at 369. 448 U.S. 297, 317 (1980).

UNCONSTITUTIONAL CONDITIONS

139

lems in bargaining that may justify rejecting Pareto permissibility in the current context. The principle of Pareto permissibility is most at home in the context of a competitive market with low transaction costs and relatively perfect information. When no individual buyer or seller can affect the prices at which goods and services are sold, there will be no negotiations between buyers and sellers: a buyer has no choice but to pay the market price; a seller has no choice but to sell at the market price. Because there will be a unique (equilibrium) price at which transactions will occur, transaction costs will be relatively low. Parties will not have to spend resources to secure a favorable price, because no one will be in a position to influence the price at which transactions will occur. Not all bargaining situations are competitive. Consider the nineteenthcentury "rescue" case, Post v. Jones.47 A whaling ship, the Richmond, full of its cargo of whale oil, ran aground in the Bering Strait in 1849. The crew was able to get to shore by boat and was in no immediate danger. The problem was the ship's cargo and equipment. Fortunately, two other whaling ships, the Frith and the Panama, were nearby. Rather than take on the Richmond's cargo in exchange for a court-awarded salvage fee, the masters of the Frith and the Panama proposed an auction to buy the Richmond's cargo. Because the Frith and the Panama had a common owner, the auction was not, shall we say, completely competitive. In any case, the master of the Richmond accepted the offered price of $1.00 per barrel for some of the oil, $.75 a barrel for the rest—a price at which the buyers could expect to earn a much greater profit (the selling price of the oil minus the purchase price and other costs) than it would have received as a salvor.48 The owners of the Richmond subsequently sued for the return of their cargo. In defending the validity of the contract, the buyers made several arguments, two of which are relevant to our present concern. First, they argued that by taking on the Richmond's cargo, they lost an opportunity to complete their own cargo of whaling oil. The Court conceded the validity of the appeal to opportunity costs, but concluded that their actual opportunity costs were virtually nil.49 Second, the buyers argued that they were not bound to save the Richmond's 47

60 U . S . 150 (1857). This case w a s b r o u g h t to m y a t t e n t i o n b y J. Bickenbach's r e -

v i e w of m y b o o k Coercion in 20 Canadian Journal of Philosophy 577 (1990). 48 If the Frith and the Panama did not have a common owner and did not collude in the auction, then it might have been a highly competitive market. 49 The Court noted that the whaling season was virtually over and that the salvors would have to be leaving the area shortly. 60 U.S. 150, at 159.

140

CHAPTER FIVE

property, and that relative to what they had an obligation to do, the contract was Pareto superior and therefore valid. In effect, they argued that their greater power not to save any of the cargo entailed a lesser power to save at any price to which both parties could agree. The Supreme Court disagreed: It has been contended . . . that the sale was justifiable and valid, because it was better for the interests of all concerned to accept what was offered, than suffer a total loss. But this argument proves too much, as it would justify every sale to a salvor . . . The general interests of commerce will be much better promoted by requiring the salvor to trust for compensation to the liberal recompense usually awarded by courts for such services.50 Note that the Court conceded what it could have denied, namely, that the Frith and the Panama did not have an obligation to take on the Richmond's cargo. The Court could have argued that the best rule of law requires that potential salvors make an easy rescue for a reasonable fee, and that relative to that baseline, the salvors had made a coercive threat. Instead, the Court argued that despite the appearance of a competitive auction, there was no reason to depart from the principles and procedures for compensating salvors that had developed under admiralty law. Although admiralty courts "will enforce contracts made for salvage service . . . where the salvor has not taken advantage of his power to make an unreasonable bargain . . . they will not tolerate the doctrine that a salvor can take the advantage of his situation and avail himself of the calamities of others to drive a bargain."51 Under admiralty law, then, an acceptable contract is judged by the fairness of its terms. It must not be an "unreasonable bargain." It is not sufficient that there be no fraud or duress. If an acceptable contract is not reached, admiralty courts will award compensation that takes account of the effort and opportunity costs of the salvors and the risk that the salvage attempt would be unsuccessful (since compensation is paid only to successful rescuers).52 Given that such laws were available, the Court saw no reason to defer to the ordinary principles of contract law, where the terms of a contract are (mostly) irrelevant, and where courts will invalidate a contract only on grounds of problems in its formation—fraud, duress, mutual mistake, and so on.53 50

60 U.S. 150, at 160 (emphasis added). 60 U.S. 150, at 160. 52 See William M. Landes a n d Richard A. Posner, "Salvors, Finders, Good Samaritans a n d O t h e r Rescuers: A n Economic Study of Law a n d Altruism," 7 Journal of Legal Studies 83 (1978). 53 A similar principle underlies the c o m m o n law doctrine of "conditional privilege," 51

UNCONSTITUTIONAL CONDITIONS

141

Why did admiralty law develop such rules? Despite references to principles of fairness ("unreasonable bargain"), Post suggests that the Court's motivations for rejecting the results of the auction were primarily consequentialist and economic—to protect the "general interests of commerce." But that raises the question of why such rules are efficient. Assuming that the state does not want to require potential salvors to make (easy) rescues, why is it not more efficient to allow the market to work, to allow the parties to fashion their own agreements? One problem is this. This sort of rescue situation represents a bilateral monopoly. The rescuee faces only one available seller of rescue services; the rescuer has only one available buyer of its services. (Although Post appears to involve two buyers, the collusion created a monopoly.) In the absence of a standard price for rescues, there is a very large bargaining range between the parties. We can imagine, for example, that it would have been beneficial for the owners of the Richmond to receive any price greater than (say) $.01 per barrel (something is better than nothing) and beneficial for the owners of the Frith and the Panama to pay any price less than $10.00 per barrel. Any price in the range between $.01 and $10.00 is within their zone of agreement.54 When bargaining ranges are very large, there are three principal problems. First, there are questions about the fairness of the distribution of the social surplus. Second, transaction costs increase. Each party has incentives to engage in strategic bargaining—"to conceal preferences, to lie, to haggle, to hold out, to dicker, to build coalitions, and to price discriminate, all of which cost money but which produce no real gain."55 In effect, strategic bargaining gives rise to a standard prisoner's dilemma problem. It is in each party's interests to engage in strategic bargaining, but incurring the costs associated with strategic bargaining is collectively disadvantageous to both parties.56 which allows a shipowner to use a dock under necessity conditions (e.g., a severe storm) if it pays a fair rental value and compensation for damages. No prior agreement is necessary. The shipowner may use the dock whether or not the dockowner agrees, thus violating a property owner's normal right to exclude. In effect, the law supplies a contract for the parties rather than relying on the parties to make one. Epstein, "Unconstitutional Conditions," p. 18. 54 Note that the large bargaining range is a function of the lack of competition and not the urgency of the need. If there had been several potential rescuers genuinely competing to take on the Richmond's cargo, there would probably be relatively little variance among the prices offered. 55 Epstein, "Unconstitutional Conditions," p. 17. 56 Mancur Olson, Jr., has argued that the prevalence of interest groups and strategic bargaining is a major cause of economic decline. See The Rise and Decline of Nations (New

142

CHAPTER FIVE

When such costs are factored into each party's net gain from the transaction, each may be better off with a lower gross gain if these transaction costs could be reduced or eliminated.57 Third, when both parties have incentives to engage in strategic bargaining, there is a danger that Pareto-superior agreements will not be reached at all. For example, in attempting to bargain for a better price, the master of the Richmond might well have said, "Forget it—I'd rather let the stuff sink to the bottom of the sea than give it to you for that price/' even though it would not have been in his interest to allow that (deadweight?) loss. And if such statements are to be at all credible, parties must believe there is a reasonable chance that they will be carried out, and this may be impossible unless the parties find some way to commit themselves to carrying out such threats.58 Moreover, if there is a reasonable probability that feasible rescues will not occur because bargaining breaks down, then potential rescuees have an incentive to invest in preventive measures to lower the probability of the necessity for rescue. It may be more efficient—in the "general interests of commerce"—to facilitate such rescues rather than to encourage unnecessary investments in preventive measures. Thus we have identified a consequentialist justification for a rule that attempts to solve the bargaining problem ex ante. The law imposes a price or judicial process for determining the price and allows the parties to accept or reject it.59 On the assumption that A (a salvor) does not have a duty to save, the law gives A two alternatives: (1) do not save, or (2) save at a (roughly) specified price. And given this choice, we assume that A—stripped of the opportunity to bargain for an exorbitant price—will prefer (2) to (1) if the price is appropriately set. Note that this approach does not paternalistically assume that the parties are mistaken about their interests. Rather, the law attempts to help the parties better realize their interests "in a world in which high transaction costs prohibit their realization through private transactions."60 In slightly different terms, the law attempts to imitate (more Haven: Yale University Press, 1982). For an analysis of contemporary American politics that relies on this model, see Jonathan Rauch, Demosderosis (New York: Random House, 1994). 57 Suppose that the gross gain from a transaction is 140 to A and 160 to B, but that each incurs 30 units in transaction costs, leaving A with a net gain of 110 and B with a net gain of 130. Under these conditions, any specified gain over 110 is better for both. For example, if the law specifies that A and B each receive 150, A is better off by 40 and B is better off by 20 than they are under the bargaining regime. 58

See T h o m a s C. Schelling, The Strategy of Conflict ( C a m b r i d g e : H a r v a r d U n i v e r s i t y P r e s s , 1963). 59

60

Landes and Posner, ''Salvors, Finders," p. 100. Richard Epstein, "Unconstitutional Conditions and Bargaining Breakdown," 26

San Diego Law Review 129 (1989), at 194.

UNCONSTITUTIONAL CONDITIONS

143

or less) the results that a competitive market would have provided in the actual noncompetitive context in which the parties find themselves.61 The following question arises. If we want to be reasonably confident that A will rescue B in such cases, why do we not simply require A to rescue B, rather than set a price? The justification of duties of rescue is a complex question, which I cannot pursue here. But note that in most contexts in which we think it plausible to argue for such a duty (as in Nozick's Drowning Case), the rescuer is not providing a service for which there is a normal market price or where we would think it appropriate for the rescuee to pay. In addition, assuming that the specified rescue price is sufficient to motivate potential rescuers to provide their services, it is much less costly to enforce a specified price than it would be to determine whether an alleged potential rescuer was culpable for failing to rescue. How does this discussion of rescues at sea help us understand the doctrine of unconstitutional conditions? The state is surely no salvor. Nonetheless, like the Frith and the Panama, the state is a monopoly provider for many goods and services. With respect to some goods, such as tax exemptions or civil service jobs, the state has a natural monopoly. No one else can provide these benefits. With respect to other goods, such as highways, it is not feasible to provide them on a private basis. In still other cases, the state provides services that could be but often never were provided privately on a widescale basis, such as unemployment compensation, schooling, welfare, or public defenders.62 The state's position as a monopoly provider gives rise to all of the problems of "strategic bargaining" that we have noted. First, those in positions of political power can use their ability to withhold benefits to obtain side payments (monopoly rents), to achieve personal or political objectives that might otherwise be beyond their reach. Those who do not now have such control can try to attain it by influencing those in political power or capturing those positions themselves, so that they can obtain similar benefits for themselves. Recipients have incentives to spend considerable resources attempting to secure benefits that would have been available had bargaining been impossible. If the government can bargain with citizens over the terms on which benefits will be received, then the high transaction costs that we noted in the private monopoly setting will be played out in the political arena. Second, the ability of the state to bargain with citizens over the terms on which benefits will be received may also give rise to exter61 62

Epstein, "Unconstitutional Conditions," p. 19. Sullivan, "Unconstitutional Conditions," p. 1452.

144

CHAPTER FIVE

nality problems. Recall Speiser. It would not be surprising if few veterans experienced signing a loyalty oath as a severe ideological or personal cost, either because they did not place a high value on freedom of speech, or, more likely, because they did not see the connection between signing a loyalty oath and the freedom of speech they value. Moreover, even if a veteran did experience signing a loyalty oath as a cost, he received a benefit that may have outweighed that cost to him. Still, even if the loyalty oath for a tax exemption deal is beneficial on balance to most veterans, their acceptance of that deal may impose a cost on the society as a whole. Although no individual waiver does much harm, the combined effect of such waivers may substantially weaken the framework of constitutional rights for those who wish to preserve that framework and who (unlike the veterans) receive no compensating benefit for whatever weakening occurs.63 Third, the state's position as a unitary provider of benefits to a large set of unorganized individuals can also give rise to collective action problems among the beneficiaries. Suppose that the following were true: (1) most veterans would prefer to live in a regime in which one can receive certain benefits without signing a loyalty oath; (2) all would prefer not to sign the loyalty oath if they were assured that others would also refuse to sign because they believe that if all were to refuse to sign, the government would provide the tax exemption anyway or because the benefit (to each) of living in a regime that provides stronger protection for individual rights is greater than the benefit to each of the tax exemption; and (3) lacking assurance that others will refuse to sign, each would prefer to sign a loyalty oath in exchange for a tax exemption. Under these conditions, the state would be able to effect a "benefit for rights waiver" trade with each beneficiary that it could not accomplish if the beneficiaries could solve their assurance problem. Exit Holmes. Enter the doctrine of unconstitutional conditions. In effect, the doctrine of unconstitutional conditions puts the state to an all-or-nothing choice in much the same way that potential salvors are put to a choice between accepting the specified price (or process for setting a price) or nothing. The state can provide the discretionary benefit without demanding that the recipient waive a constitutional right, or the state can choose not to provide the benefit at all. But it cannot bargain with a recipient over the terms on which the benefit will be provided, or, somewhat less strongly, the sorts of conditions it can impose are highly constrained, for example, by the principle that the rights waiver must be sufficiently germane to the benefit received.64 63 64

Epstein, "Unconstitutional Conditions/' p. 54. Whereas it would be uncontroversially unconstitutional for the federal govern-

UNCONSTITUTIONAL CONDITIONS

145

Putting governments to a constrained choice has several advantages. First, it reduces the socially costly strategic behavior that we have discussed. Consider Western Union v. Kansas once again. If put to a highly constrained choice, members of the Kansas legislature have less incentive to bargain with the company and seek (socially costly) side payments for their cooperation. Western Union need not lobby or seek to control the Kansas legislature nor need it seek ways to make a credible threat to pull out of Kansas if the state refuses to relent. Second, the doctrine of unconstitutional conditions can be understood as helping to solve a collective action problem among the recipients of a benefit. The doctrine prevents the government from accomplishing an objective on an individualized basis that it could not accomplish on a group basis. Consider how the doctrine of unconstitutional conditions parallels Mill's well-known story about the length of the workday. An employer faces an unorganized labor force, which is working nine hours a day for a day's wage. Suppose the employer says, "I'll offer you the same day's wage for ten hours' work." On an individual basis, each worker might well agree to work ten hours for a day's wage if he or she would otherwise be replaced by another worker who is willing to work ten hours for a day's wage. On a collective basis, however, it is possible that each worker would prefer not to be able to work ten hours for nine hours' pay if all were unwilling to do so. Yet because they may be unable to reach that result by voluntary action, each would prefer that all were legally prevented from working ten hours for a day's wage.65 A law requiring employers to pay overtime for work over nine hours a day solves the workers' collective action problem and makes it impossible for the employer to pursue a "divide and conquer" strategy, thereby strengthening the bargaining position of each worker. The doctrine of unconstitutional conditions helps to overcome similar collective action problems. Recall Insurance Co. v. Morse. Although ment to require that a state permit homosexual marriages as a condition of receiving federal highway funds, although such a condition would satisfy Holmes's test, the federal government argued in South Dakota v. Dole that a twenty-one-year-old drinkingage rule was germane to highway funds because both were designed to enhance highway safety. 65 "Assuming then that it really would be [in] the interest of each to work only nine hours if he could be assured that all others would do the same, there might be no means of their attaining this object but by converting their supposed mutual agreement into an engagement under penalty, by consenting to have it enforced by law." John Stuart Mill, Principles of Political Economy, ed. J. M. Robson (Toronto: University of Toronto Press, 1965), bk. 5, ch. 11, sec. 12, p. 958. See the discussion of this argument in Russell Hardin, "The Utilitarian Logic of Liberalism" 97 Ethics 47 (1986), and Arthur Kuflik's "The Utilitarian Logic of Inalienable Rights," 97 Ethics 75 (1986).

146

CHAPTER FIVE

it might be in the interest of each corporation to waive its right to use federal courts, it would be in the interests of all corporations to refuse to waive that right if the state would then prefer to allow them entry without the waiver than to exclude them all entirely. It is only because the state can "pick off" the corporations on a one-by onebasis that it has any feasible prospect of achieving its goal. From this perspective, the lesser power to give a benefit to some recipients is greater than the allegedly greater power to give it to none, because giving it to none is not a politically feasible alternative. If the government can take advantage of the recipients' inability to solve their collective action problem, it will be obtaining rights waivers on the cheap. And as Seth Kreimer writes, "If we are to permit the sale of constitutional rights, it seems inappropriate to allow the state to purchase them with fools' gold."66 The government should pay in hard currency. To put the point in slightly different terms, we can understand the doctrine of unconstitutional conditions as representing the hypothetical baseline that would obtain if the government were not able to impose the condition in question. If, under those conditions, the government would not provide the benefit, then the "no benefit" baseline is secure. But if the government would still provide the benefit if it were not able to impose the condition, then we can understand the recipient's baseline as including the benefit. Consider Harris from the latter perspective. Justice Stewart said, "The Hyde Amendment leaves an indigent woman with at least the same range of choice in deciding whether to obtain a medically necessary abortion as she would have had if Congress had chosen to subsidize no health care costs at all."67 There are two problems here. First, Stewart's claim may not be true, for if Congress had chosen not to subsidize any health care costs, it is possible that other forms of health care for the indigent would have developed. Second, even if Stewart is right, it can be argued that it begs the question to assume that the only relevant baseline is the one in which Congress subsidizes nothing: if—put to an allor-nothing choice—Congress would have subsidized medically necessary abortions rather than subsidize "no health care costs at all," then it is effectively attempting to purchase the waiver of a constitutional right for nothing. We should not assume, of course, that if the state were put to an allor-nothing proposition, it would always choose all rather than noth66

"Government 'Largesse' and Constitutional Rights: Some Paths through and around the Swamp," 26 San Diego Law Review 229 (1989), at 238. 67 448 U.S. 297 (1980), at 317.

UNCONSTITUTIONAL CONDITIONS

147

ing.68 Suppose that the government could not constitutionally require a physician to practice medicine in a rural area, but that the government proposes to pay for a student's medical education in exchange for a commitment to practice for five years in an underserved rural area. There is no reason to think that the government would choose to fund all medical education if it could not use its funding power in this way. Thus the state is paying something (the benefit it is extending) for the benefit it is receiving (the rights waiver) and those who are waiving their rights in exchange for a benefit are getting something for their rights waiver—the benefit that they would otherwise not have received. Similarly, if California would not have extended tax exemptions to veterans if it could not require them to sign a loyalty oath, then it has paid something (the tax exemption) for its gain (the loyalty oath), and the beneficiaries have gained something (the tax exemption) for what they have paid (signing the loyalty oath). We might still worry about the weakening of constitutional rights, but we can assume that if the price for discouraging the exercise of constitutional rights goes up, constitutional rights will be better protected. In addition to the strategic behavior and collective action justifications of the doctrine of unconstitutional conditions, it may also help to reduce the externality problems we noted above. Consider South Dakota v. Dole from that perspective. Many states had already adopted a twenty-one-year-old drinking-age rule at the time the federal policy was enacted, and others were willing to do so without grumbling. Nonetheless, by readily complying with the federal government's condition, these states shifted the burden of protecting "state's rights" to those states that cared more about the particular provision or the principles of federalism that were at stake. The less compliant states were increasingly isolated and without effective political leverage. The doctrine of unconstitutional conditions, if applied, would help to reduce such externality effects. UNCONSTITUTIONAL CONDITIONS AND REDISTRIBUTION

Up to this point, we have assumed that the government is under no obligation to provide a particular benefit. Still, even if the state is under no obligation to provide a particular benefit, it does not follow that the state's resources are its to distribute in any way it pleases. Those resources belong, shall we say, to the people, and should be 68

As Epstein puts it, if "the government is told that it cannot bargain with individuals, the empirical question arises whether government will deny them a useful benefit altogether, or grant them the benefit without the obnoxious condition." Supra note 10, at 28.

148

CHAPTER FIVE

used for public purposes that are consistent with democratic process and respect for basic constitutional rights. Even if we are not entitled to any specific benefit, we have "a presumptively equal right to share in governmental benefits/'69 It is not as if, once we have contributed our fair share to the state's resources, everything the state does for us is an instance of governmental beneficence. Consider the provision of a municipal swimming pool. A city has no obligation to provide a swimming pool for anyone. If it were to provide a swimming pool for whites but not for African Americans, it could maintain that African Americans are no worse off than if the city had provided no pool at all.70 I take it that the city could not successfully argue that its greater power to provide for no one includes its lesser power to provide for only some. Why not? There are so many problems with this argument that it is difficult to know where to start. Let me mention but a few. First, it simply begs the question to suppose that reasonable opposition to a practice requires that the practice make some persons worse off. Second, it is plausible to suppose that people are, in fact, worse off when they are excluded from a benefit on racial grounds, particularly because their sense of deprivation is rooted in justifiable resentment and not mere envy. Third, when a good improves one person's competitive position over another, providing a benefit to some makes the nonbeneficiary worse off. This may not be a serious problem with respect to swimming pools. But providing better schools to whites than to African Americans makes the latter worse off than they would be if whites had to go to equally bad schools. There is another way to put the objection to the exclusionary swimming pool, namely, that it takes resources from some citizens and gives them to others on an illegitimate basis. This is not an argument against redistributive policies as such. The state may legitimately redistribute resources from rich to poor, healthy to sick, nonveterans to veterans, the employed to the unemployed. But it may not redistribute resources from African Americans to whites, at least not without a compelling reason to do so. A redistribution may be unconstitutional because it violates a constitutional right to equal protection even if it places no pressure on people to waive a constitutional right. Along similar lines, we can also argue that redistributions are unconstitutional if they redistribute resources on the basis of an individual's exercise of a constitutional right without a legitimate justifica69

Michael W. McConnell, "Unconstitutional Conditions: Unrecognized Implications for the Establishment Clause," 26 San Diego Law Review 255 (1986), at 260. 70 I borrow this example from Seth Kreimer. See "Allocational Sanctions: The Problem of Negative Rights in a Positive State," 132 University of Pennsylvania Law Review 1293 (1984), at 1312.

UNCONSTITUTIONAL CONDITIONS

149

tion. A state may not have to provide welfare benefits. But if welfare benefits are provided to the poor as a group, then any member of that group is entitled to the benefits without regard to the exercise of a constitutional right, say, support of a Democratic candidate. The state cannot use the offer of such benefits to shift resources from Democrats to Republicans. Consider Speiser from this perspective. California was not required to provide tax exemptions to veterans. But once the state had decided to redistribute resources from nonveterans to veterans—perhaps as a token of its gratitude—that decision constituted a new "baseline for neutrality."71 Once the state had decided to redistribute resources in this way, the recipient's desire to exercise a constitutional right was not a legitimate basis for depriving the recipient of this particular benefit. Although Justice Clark was right to maintain that the policy was not a "criminal proceeding," that "neither fine nor imprisonment is involved," Speiser was being penalized relative to the new baseline precisely because he exercised a constitutional right. We might say the same thing about South Dakota v. Dole. Once the federal government had decided to provide funds for highways, it was arguably not permitted to redistribute those resources from a state that chose to exercise its constitutional right to set its drinking age under twenty-one, at least not without an appropriate justification.72 This view does not deny that the government may be justified in requiring the waiver of a right in exchange for a benefit. But the baseline for evaluating those justifications will have shifted. One advantage of the redistribution interpretation of unconstitutional conditions is that it helps to explain why a potential beneficiary can object to a condition even if the government's offer is not a form of effective pressure. Suppose, for example, that no Seventh Day Adventist would be prepared to work on the sabbath just to qualify for unemployment compensation. On the redistribution thesis, Seventh Day Adventists would be receiving less than their fair share of social resources because they exercised a constitutional right, even though none are motivated or coerced to waive their constitutional right.73 It might be thought that the redistribution account of unconstitutional conditions is virtually indistinguishable from the coercion interpretation that I have rejected, that I have let coercion in through the 71

Supra note 69, at 260. See Justice O'Connor's dissent, 483 U.S. 203 (1987), at 212-18. 73 In Harris, Justice Marshall makes both the coercion and redistribution arguments. He argues that "the differential distribution of incentives . . . can have precisely the same effect as an outright prohibition" (at 347). He also argues that the "appellees have been deprived of a governmental benefit for which they are otherwise eligible solely because they have attempted to exercise a constitutional right" (at 346). 72

150

CHAPTER FIVE

back door. It may be argued that if the state's decision to provide benefits establishes a new baseline for evaluation, then I have translated the state's noncoercive offer to provide those benefits if the recipient waives a constitutional right into a coercive threat to withhold those benefits unless the recipient waives a constitutional right. I believe that this objection has some force, but I do not think that the wrongness of the government's proposal is best captured by the notion of coercion. For even if we say that the state's decision to provide a benefit establishes a new entitlement baseline for claims of coercion, there remains an important distinction between two sorts of entitlement baselines. In some cases, the baseline rests on what we might call an absolute entitlement that the government is obligated to fulfill, such as the right to a trial by jury in a criminal case. In other cases, the baseline rests on a comparative entitlement, where the government is obligated to provide a benefit to B given that it is providing that benefit to others. In the standard case of coercion, the government threatens to violate an absolute entitlement unless one waives a right. In the cases under discussion, the government proposes not to provide a comparative entitlement unless one waives a right. Because a comparative entitlement is still an entitlement, the proposal to withhold a comparative entitlement could be couched in terms of coercion. But precisely because the entitlement derives from its availability to others, I think the source of the wrongness of the government's proposal is better understood through the lens of redistribution.74 The application of the redistribution approach raises a set of problems as to how to set the baseline for "neutrality." Recall Sherbert v. Verner. The majority argued that from a baseline in which Sherbert is entitled to unemployment compensation, the policy interferes with her free exercise of religion. It effectively redistributes resources from Saturday sabbath persons to Sunday sabbath persons. But it could also be argued that to award unemployment compensation to a Seventh Day Adventist redistributes resources from Saturday sabbath persons to (say) Saturday leisure persons, thereby violating the constitutional prohibition against the "establishment of religion."75 Similar difficulties of application arise with respect to the question as to what would constitute a legitimate basis for redistributing resources away from those who choose to exercise a constitutional 74

I thank Don Loeb for pressing me on this point. See Justice Stewart's dissent in this case, 374 U.S. 398 (1962), at 415. Justice Stewart, who concurred in the result in this case, raised this very question, arguing that on the majority's own views about the establishment clause, it should have found against Sherbert, although he—who disagreed with the majority's views about establishment— was entitled to find for her. 75

UNCONSTITUTIONAL CONDITIONS

151

right. Germaneness seems to matter.76 Suppose that the federal government were to make funds available for medical care on condition that the medical care be obtained from a licensed physician rather than in "alternative" medical care. Although the citizen may have a constitutional right to seek alternative medical care, the condition seems justifiably related to the purpose of the expenditure, that is, to promote health. The question is how tight this connection must be. Recall South Dakota v. Dole. The federal government could argue that it made highway funds available to the states in order to promote "safe travel," and that its drinking-age condition was therefore germane to the purpose of the expenditure—even though it did put pressure on the state's ability to exercise its constitutional right to set its drinking age. On the other hand, South Dakota could argue that the connection between the condition and the benefit is so attenuated as to constitute an illegitimate redistribution from compliant states to noncompliant states. RIGHTS

We have examined the doctrine of unconstitutional conditions from several different perspectives. I now want to return to the issue of rights, which is, after all, the normative framework that motivated support for and opposition to the doctrine. With respect to the rights of citizens, the case for the doctrine of unconstitutional conditions maintains that the government should not be permitted to induce citizens to waive rights that it could not deny by coercion. With respect to the rights of the state, Holmesian opposition to the doctrine maintains that if the government has a right not to provide a benefit ("the greater right") then it has the right to extend the benefit on restrictive terms ("the lesser right"). Our discussion of the doctrine of unconstitutional conditions gives us reason to reject both of these views, at least when stated so starkly. At first glance, it seems reasonable to believe that the government should not be allowed "to 'buy up' rights guaranteed by the Constitution."77 On closer examination, however, this must be incorrect. As a general matter, we do not enhance B's autonomy or freedom by prohibiting B from waiving or trading that right in exchange for a benefit.78 In principle, a right that can be sold or exchanged is worth more 76

See the discussion of the germaneness argument in Sullivan, "Unconstitutional Conditions," pp. 1473-1476. 77 Wyman v. James, 400 U . S . 309, 328 (1971) ( D o u g l a s , J., d i s s e n t i n g ) . 78 "While . . . many attempts to alienate fundamental human rights will, as a matter of fact, fail (being the result of voiding conditions such as duress or insanity), it is quite

152

CHAPTER FIVE

than a right that cannot be sold or traded.79 Consider the right to a trial by jury. Suppose that B is a defendant and that the government offers B the opportunity to plead guilty to a lesser charge, thereby waiving the right to a jury trial in exchange for testimony against another suspect. There may be good reasons not to allow the government to offer such deals, say, because otherwise guilty defendants will not receive the punishment they deserve. But this is not because we think the right to a jury trial is inherently inalienable. If that were so, then we would have to disallow defendants from pleading guilty.80 Of course, even if there is a presumption that one should be able to trade one's right in exchange for a benefit, there may be reason to prohibit such trades in certain cases. First, as we have seen, we may prohibit B from waiving a right because doing so enables B to better obtain B's own ends. That is the upshot of our analysis of strategic bargaining and collective action problems. Second, we may prohibit someone from waiving a right because her decision is less than fully voluntary under the circumstances or because we want to preserve a range of future choice in "areas of a person's life that are of the most profound consequence to his happiness and well being."81 Thus we do not allow people to enter into irrevocable marriages, to waive their right to seek a divorce even though one or both parties might prefer to do so. (Yes, I think it is possible that one might want to waive one's right to a revocable marriage). There is a third reason to prohibit citizens from trading some of their rights for benefits, namely, that such exchanges may have important negative externalities. Contrast the case in which the government proposes to pay for a student's medical education in exchange for a commitment to practice for five years in an underserved rural area with a case in which the government proposes to pay for a student's medical education in exchange for an agreement not to perform aboranother matter to argue that these rights are (in principle) inalienable, and an even more difficult matter to argue in this way while staunchly defending our natural moral freedom to control the course our lives will take." A. John Simmons, On the Edge of Anarchy (Princeton: Princeton University Press, 1993), p. 110. 79 See Frank Easterbrook, "Insider Trading, Secret Agents, Evidentiary Privileges, and the Production of Information," The Supreme Court Review 309, 346 (1981). 80 By contrast, we might not allow a defendant to trade the right to an attorney for a better deal, because the advice of an attorney may be a necessary condition of informed consent. 81 Arthur Kuflik, "The Utilitarian Logic," p. 84. As Justice Hunt observed in Insurance Co. V. Morse, there is an important distinction between the ability of a citizen to waive a right in any given transaction and the capacity "to bind himself in advance . . . to forfeit his rights at all times and on all occasions." 87 U.S. 445 (1974), at 451.

UNCONSTITUTIONAL CONDITIONS

153

tions. The first offer seems to be constitutional whereas the second does not. But why? There are several differences between the cases. Whereas the first has a specified time limit, the second does not (although I doubt it would matter much if it did). Whereas the second offer puts pressure upon a physician's ability to perform an appropriate and legal medical procedure, the first does not. But the principal difference between the two policies may not concern their respective effects on physicians, but their effects on patients. Whereas the "rural area" policy does not affect unfairly anyone's right to receive appropriate medical treatment, the abortion policy puts pressure upon the ability of potential patients to exercise their right to have an abortion.82 Consider the right to free speech. Suppose that the federal government wants to study the psychological and physiological effects of prolonged silence.83 The government selects 100 citizens at random and offers each a cash payment ($500) if they will not speak for two weeks. Now, this seems clearly constitutional, even though it discourages the exercise of a constitutional right. By contrast, it would be clearly unconstitutional if the government were to give a tax rebate of $50 to every citizen who agrees not to espouse socialist views—not because the right to free speech or association is inherently inalienable, for then that right couldn't be "sold" as in the previous example, but, say, because of the effects of such purchases on the democratic process, the quality of public discussion, and so forth.84 The general point, then, is that there is no reason to regard constitutional rights as generally inalienable. Whether a right should or should not be tradable depends, in part, upon the effects of such trades and cannot be settled by factors intrinsic to the rights involved. We may have good reason to treat some constitutional rights as inalienable or to prohibit trading some constitutional rights for some particular benefits. But a prohibition against trading constitutional rights will 82

I borrow this example from Albert J. Rosenthal, "Conditional Federal Spending as a Regulatory Device," 26 San Diego Law Review 277 (1989), at 282-83. 83 I borrow this example from Kenneth Simons, who, in turn, credits it to Laurence Tribe. See Simons, "Offers, Threats and Unconstitutional Conditions," 26 San Diego Law Review 289, 290-91 (1989), and Tribe, "Private Choices and Public Funding: Abortion Revisited," both in J. Choper, Y. Kamisar, and L. Tribe, eds., The Supreme Court Trends and Developments 1979-80: An Edited Transcript of the Second Annual Supreme Court and Constitutional Law Symposium 256 (1981). 84 See J o s h u a C o h e n , " F r e e d o m of E x p r e s s i o n , " 22 Philosophy & Public Affairs 207. We might say similar things about the right to vote. The main reason that we do not allow people to sell their votes or to agree not to exercise their right to vote in exchange for a fee is not to protect the individual, although that may be part of the story, but to protect the democratic process. See Larry Alexander, "Understanding Constitutional Rights in a World of Optional Baselines," 26 San Diego Law Review 175, 183 (1989).

154

CHAPTER FIVE

always emerge as the conclusion of an argument about the way we should treat a particular constitutional right, and not as the premise of an argument about the way we should treat constitutional rights in general. Let us now reconsider the alleged rights of the state. Despite everything we have said, is there not something to the view that if the government has a (greater) right not to extend a benefit, then it ought to have the (lesser) right to extend the benefit on certain conditions? There are two responses to this question—one conceptual, one normative. And both responses are negative. The conceptual response is that there is, indeed, nothing in the logic of the concept of rights that requires us to accept "the greater includes the lesser." It is a commonplace that rights are best understood as bundles of specific legal relationships, for example, immunities, privileges, powers, correlative obligations, and so on. And it is an open question just what sticks should be placed in any specific bundle. If A owns a piece of land, A may have the right to exclude others and to sell it, but A may lack the right to use it in any way A wishes. We do not think that A's right not to marry B gives A the right to marry B on any condition to which they might agree, for example, B's permanent waiver of her right to divorce A. We do not think that a landlord's ownership of an apartment gives him the right to rent on condition that the tenant waive a right to habitability. We do not think that an employer's right not to hire a worker gives the employer the right to hire the worker on any condition to which the worker might agree (e.g., subminimum wages, unsafe conditions) or to discriminate on racial grounds among prospective employees. The fact is, after all, that we do not accept "the greater includes the lesser" in many social contexts. The decision to reject "the greater includes the lesser" may have to be justified, but it violates no logical requirement of the concept of rights. Or if it does violate a logical requirement of the concept of a right, it does so in a host of nonconstitutional contexts that have not been thought to be particularly controversial. The longer—normative—response is that whether we should accept "the greater includes the lesser" depends upon the shape and justification of the right in question. If A has a deontologically (autonomy) grounded right not to transact with B, then there may be some reason to think that A has a right to enter into a transaction that provides a benefit to B on any terms on which both A and B can agree, although I think this will often not prove to be the case. But if the background justification of a right is consequentialist, there is no reason to think that the greater right not to transact must entail a lesser right to transact on mutually agreeable conditions. If the reason for

UNCONSTITUTIONAL CONDITIONS

155

acknowledging the greater right is that it has good consequences, then it is an open question as to whether acknowledging a lesser right will also have good consequences. We assign certain property rights to A and B and let them transact according to certain rules because the consequences (which may include considerations of autonomy, control, economic efficiency, and minimization of strife) are superior to alternative institutional regimes.85 But if the consequences of acknowledging a "greater right" not to sell one's property are good and the consequences of acknowledging a "lesser right" to sell one's property on certain conditions are bad, there is no moral reason to think that we must acknowledge the lesser right simply because we acknowledge the greater right. Whatever the grounding of individual rights, it is surely the case that the government does not have any deontologically grounded right to its resources or to untrammeled discretion to provide benefits as it wishes. Rather, we assign a certain realm of discretion to the government because we think that the package of some democratic discretion and constitutional restraints on such discretion is preferable to a package that provides for greater democratic discretion and fewer constitutional restrictions or one that provides for greater constitutional restrictions and less democratic discretion. If the reason for assigning a certain realm of discretion to the government is broadly consequentialist, then there is no reason to accept "the greater includes the lesser" if and when a different construal of governmental rights promotes our preferred ends more effectively. And so when Holmes said that he could not understand "how a condition can be unconstitutional when attached to a matter over which a State has absolute arbitrary power/' we can respond that there is no reason to think that the state has absolute arbitrary power over such matters in the first place. IMPLICATIONS

What is the import of this analysis of unconstitutional conditions for the problem of exploitation? As we have seen, it may be argued that there is no justification for prohibiting or restricting mutually advantageous transactions, even if they are—by some external criterion— exploitative or unfair. On this view, (1) if A has the right not to provide B with a benefit, then (2) there can be no moral basis for rejecting a noncoercive and nonfraudulent transaction between A and B. I have suggested that there is no reason to think that (2) follows from (1) at 85

See T. M. Scanlon, "Rights, Goals, and Fairness," in Stuart Hampshire, ed., Public and Private Morality (New York: Cambridge University Press, 1978).

156

CHAPTER FIVE

the constitutional level. We have seen that there are several reasons that might justify rejecting agreements that are Pareto superior to a no-transaction baseline, and we have uncovered conceptions of alternative baselines from which we can evaluate the acceptability of an agreement. It is less clear just how far these reasons will take us with respect to cases of mutually advantageous exploitation. Consider the argument from externalities. Although there may be very good reasons to prohibit B from selling his right to A because of the effect of the trades on C, it is less clear that these reasons have anything to do with exploitation. For we are inclined to think that if A exploits B, the exploitation is to be found in the relation between A and B and not in some effect on third parties. Now, it may turn out, on deeper examination, that we regard some transactions as exploitative precisely because of their effects on third parties as distinguished from the parties to the transaction. But at first glance, this seems to take us down the wrong track. The arguments from strategic bargaining problems and collective action problems are more obviously connected to exploitation. They show that B may be better enabled to achieve B's own ends if B cannot make an otherwise Pare to-superior agreement with A. But here the arguments cut both ways. On the one hand, the arguments illuminate those cases of exploitation that arise from a collective action or bargaining problem. On the other hand, they do nothing to illuminate those cases that cannot be explained in those ways, and they may even imply that there is nothing to worry about in those cases. Whether there is something to worry about in such cases remains to be seen. As we have seen, the doctrine of unconstitutional conditions also supports an alternative view of the citizen's baseline, or neutral, position. We might say that if the government would hypothetically provide a benefit free of charge if it could not impose a condition, then we may regard that hypothetical baseline as the standard against which we consider the government's proposal. Because the government is a nonprofit agency that works for the people, we may think that the government should generally charge no more than its reservation price for a benefit. But the relation between the government and its citizens is arguably special, and we may need another way to explain why transactions between individuals are exploitative. That A would be prepared to pay B $100,000 for a house does not mean that A exploits B when he gets B to agree to accept $95,000 any more than B exploits A's willingness to pay $95,000 because she is willing to sell the house to A for $90,000. Because A is entitled to do better than his reservation price, we need a different account of exploitative transactions.

UNCONSTITUTIONAL CONDITIONS

157

Let us now consider the argument from redistribution. I have suggested that there is some space between the claim that a citizen is not entitled to a specific governmental benefit (which may be true) and the claim that the provision of any benefit should be understood as governmental "largesse." The redistribution argument maintains that the government cannot propose to deny a citizen her fair share of resources if she fails to waive or chooses to exercise a constitutional right because that fair share of resources is presumptively hers to begin with. The question is whether there is any appropriate analogue to the "fair share" principle in private transactions. After all, the force of the distributional argument in the governmental context derives precisely from its contrast with the private contexts in which common law entitlements do provide an appropriate baseline. Thus we have two sets of issues. Just as we can discuss the fairness of an exchange between the government and citizens, we can discuss the fairness of an exchange in private transactions. We shall return to the issue of fairness in Chapter 7. And just as we can discuss the principles that would justify prohibiting some arguably mutually advantageous but unfair transactions in the public realm, we can discuss the principles that justify prohibiting such transactions in the private realm. We shall return to this issue in Chapter 9.

Chapter Six SEXUAL EXPLOITATION IN PSYCHOTHERAPY

claimed that members of various professions exploit their clients.1 It is said that lawyers exploit their clients when they bill them for unnecessary research or depositions, that professors exploit students by assigning work that mainly furthers the professors' research programs, and that physicians exploit their patients (or their insurers) when they order unnecessary but financially lucrative tests and procedures. In this chapter I focus on the sexual exploitation of patients by psychotherapists. There are several reasons for taking a close look at this problem. Sexual exploitation in psychotherapy is widely regarded as a matter of great concern by victims, by the profession, by scholars, and by the media.2 It presents us with a well-defined form of alleged wrongdoing that is virtually always couched in the language of exploitation. Because there are well-developed (and developing) legal and institutional responses to the problem, it presents a useful case for examining the arguments advanced in defense of those responses. And with respect to the theoretical aims of this project, it raises in a particularly interesting and important way several questions about the relation between exploitation and consent: Do patients consent to sexual relations with their psychotherapists? Would it be wrong for a psychotherapist to engage in sexual relations with his patient (in this chapter, I address the male psychotherapist/female patient case) if the patient gives consent? And what counts as genuine consent? Is the validity of consent an empirical or normative issue? Psychotherapists have no monopoly on sexual relations between professionals and clients. Sexual relations between lawyers and clients—particularly male divorce lawyers and female clients—are sufficiently prevalent to have been a staple subplot of the television series "L.A. Law/'3 Sexual relations between clergy and adult parishIT IS OFTEN

1 It would take too many footnotes to indicate all of the places in which Arthur Kuflik helped to refine the argument of this chapter. 2 It has occasioned several book-length accounts of individual cases, television documentaries, and a made-for-television movie. 3 John O'Connell, "Keeping Sex Out of the Attorney-Client Relationship: A Proposed Rule," 92 Columbia Law Review 887 (1992), at 887. In one nonfictional case, an attorney billed his client for the time they spent in bed together. See In re the Marriage of

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

159

ioners are not unknown, not to mention problems of sexual abuse of minors.4 And sexual exploitation of students by teachers has a long and not particularly respectable pedigree. Yet there are several reasons for thinking that the sexual exploitation of patients by psychotherapists is particularly troublesome. (I shall not generally distinguish between psychiatrists and psychologists.) It is not uncommon. It appears that between 6 and 12 percent of psychotherapists have had sexual relations with their patients (I have seen no data as to the prevalence of what are sometimes called "sexual boundary violations" on a per-patient basis.) Sexual impropriety accounts for 53 percent of malpractice suits against psychiatrists and 20 percent of the monetary value of the damages awarded.5 The harmful effects of sexual exploitation by psychotherapists are thought to be particularly severe. And the modality of psychotherapeutic treatment puts the psychotherapist at considerable "risk" for becoming an exploiter.6 If we step back from psychotherapy, it is an interesting question as to why the norm against sexual relations in the medical profession (as a whole) is both stronger and of longer standing than the comparable norm among lawyers and professors. The Hippocratic oath specifically requires physicians to remain "clear of all voluntary injustice and of other mischief and of sexual deeds upon bodies of females and males, be they free or slave."7 By contrast, although attorney-client sexual relations are frowned upon, there is no (per se) rule against attorney-client sexual relations in the Model Rules of Professional Conduct.8 Attorney-client sex has long been viewed as wrong, but Sherry Kantar and Michael Kantar, 581 N.E.2d 6 (111. App. Ct. 1991). In another wellknown case, a lawyer was asked by a woman to represent an acquaintance in a criminal proceeding. Because she could not pay the full fee, the attorney offered to exchange his legal services for nude photographs of the woman (and her niece) and sexual services. In re Wood, 256 Ind. 616, 358 N.E.2d 128 (1976). 4 See Eduardo Cruz, "When the Shepherd Preys on the Flock: Clergy Sexual Exploitation and the Search for Solutions," 19 Florida State University Law Review 499 (1991). Here I have in mind relations between clergy and adults. The sexual exploitation of children by Catholic priests has recently surfaced as a problem sufficiently serious to merit comment by the pope. 5 Kenneth S. Pope and Melba J. T. Vasquez, Ethics in Psychotherapy and Counseling (San Francisco: Jossey-Bass, 1991), at 28. 6 When I say "at risk," I do not, of course, mean to imply that the psychotherapist is a victim himself. 7 As translated in Leon Kass, Toward a More Natural Science (New York: Free Press, 1985), p. 229. 8 See O'Connell, "Keeping Sex Out." Also see Linda M. Jorgenson and Pamela K. Sutherland, "Fiduciary Theory Applied to Personal Dealings: Attorney-Client Sexual Contact," 45 Arkansas Law Review 459 (1992).

160

CHAPTER SIX

not very wrong, more than a peccadillo perhaps, but certainly less than a moral felony.9 Similarly, whereas sexual relations between teachers and students are now addressed by sexual harassment codes at many institutions, academia's "dirty little secret'' has not been particularly little or secretive or regarded as all that dirty. Of course, the medical patient is literally exposed to the physician in a way that the client and student are not exposed to the lawyer and professor. Moreover, the physician's normal practice brings the physician dangerously close to the boundary of impropriety. There is no large physical gap—however large the psychological and moral gap—between, say, the professional palpation of the patient's breast and improper fondling.10 So the history and strength of medicine's strictures may reflect the heightened dangers of sexual impropriety to which medical practice gives rise and the special wrongness of an inappropriate relationship with that aspect of the client's person (the body) which is the object of the physician's ministrations. Not all of these factors will help to explain the special concern about sexual relations between psychotherapists and their patients, for like the lawyer or teacher, the psychotherapist interacts with the patient's mind, not her body.11 I believe that psychotherapist-patient sexual relations are viewed as particularly wrong for at least four reasons. First, sexual contact between psychotherapists and patient is more likely to involve sexual intercourse as contrasted with inappropriate touching (although this does not distinguish psychotherapy from law and academia). Second, precisely because the sexual contact between a psychotherapist and a patient is more likely to involve sexual intercourse, the patient is more likely to have given some sort of consent to the sexual relations than, say, to an unwanted touching by a physician, where the patient gives no consent at all. Although this might suggest that psychotherapist-patient sexual relations are less wrong because they are more consensual, the fact that the psychotherapist has manipulated the patient's consent may constitute an additional and special wrong. Third, the underlying disorder that motivates a patient to seek psychotherapy suggests that the patient starts from a 9

I doubt that doctors would even jokingly refer to sexual relations with patients as "taking it out in trade." Ibid., p. 887. 10 This raises the question as to whether "fondling" is objective or subjective and, if subjective, whose subjectivity is crucial. For example, is B fondled by A if: (1) A experiences the touching as a fondling but B does not? (2) A does not experience the touching as a fondling but B does? I am inclined to regard (1) as a fondling but not (2), unless there is reason to think that B is experiencing something about A's intentions of which A is unaware. 11 I assume that I need take no position on the mind/body problem in this context!

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

161

position of special vulnerability. Fourth, because the relationship between the psychotherapist and patient requires a great degree of emotional trust, it also presents an opportunity for the betrayal of that trust. For whatever reasons and whatever the level of compliance, the norms prohibiting sexual relations between psychotherapists and patients (among both psychiatrists and nonphysician psychotherapists) are particularly stringent.12 In addition to well-developed professional rules, there have been a variety of legal responses to the problem of sexual relations between psychotherapists and patients. Indeed, several states have enacted criminal statutes that are specifically aimed at psychotherapists.13 The purpose of this chapter is not to add my voice to the chorus of condemnation. That would be both presumptuous and inefficacious. Rather, I want to see what can be learned about exploitation, its conceptual corollaries and its moral force. And here we encounter a not atypical phenomenon. Although the term "exploitation" is widely applied to psychotherapist-patient sexual relations, the existing literature contains virtually no attempt to explain why the phenomenon is best captured by that concept.14 This remark should not be misunderstood. In some other cases of alleged exploitation discussed in this book, I do question whether the transaction is (a) wrong at all, or (b) wrong because it is exploitative, or (c) justifiably prohibited even if it is wrong by virtue of its being exploitative. In the present context, however, I do not seriously question whether sexual relations between psychotherapists are typically wrong, exploitative, and justifia12

The code of the American Psychiatric Association says this: "The patient may place his/her trust in his/her psychiatrist knowing that the psychiatrist's ethics and professional responsibilities preclude his/her gratifying his/her own needs by exploiting the patient. This becomes particularly important because of the essentially private, highly personal, and sometimes intensely emotional nature of the relationship established with the psychiatrist." From American Psychiatric Association, Principles of Medical Ethics with Annotations Especially Applicable to Psychiatry, 1985, cited in Peter Y. Windt et al., eds., Ethical Issues in the Professions (Englewood Cliffs, N.J.: Prentice-Hall, 1989), p. 567. 13 They include Minnesota, Wisconsin, Colorado, North Dakota, California, Maine, and Florida. See Linda Jorgenson, Rebecca Randies, and Larry Strasburger, "The Furor over Psychotherapist-Patient Sexual Contact: New Solutions to an Old Problem," 32 William and Mary Law Review 645 (1991), at 672-73. A Minnesota statute creates a specific civil cause of action for injury caused by "sexual exploitation" of a patient or former patient. Minn. Stat. § 148A.02 (Supp. 1987) 14 For example, a provision in a Wisconsin statute specially labeled "Sexual exploitation by therapist," provides definitions of many terms—"physician," "psychologist," "sexual contact," "therapist," etc. But it contains no definition of exploitation. As reprinted in John Gornisiorek, Breach of Trust (Thousand Oaks, Ca.: Sage, 1995), p. 334.

162

CHAPTER SIX

bly prohibited. The question is why such relations are wrong in a way that is most naturally described as exploitation and why this form of exploitation is justifiably prohibited.

PSYCHOTHERAPIST-PATIENT SEXUAL RELATIONS

Although our purposes are conceptual and theoretical, it is important that we get some feel for the sorts of cases that figure prominently in the literature. Here are a few. Mrs. Zipkin was referred to Dr. Freeman, a psychiatrist, for treatment of severe diarrhea and headaches after her internist had been unable to find a physical cause for her symptoms.15 Her symptoms subsided after two months of treatment. She asked Dr. Freeman whether she needed to continue therapy, and Dr. Freeman persuaded her to continue, maintaining that her symptoms would recur unless the underlying problems were resolved. Mrs. Zipkin accepted Dr. Freeman's recommendation.16 Mrs. Zipkin eventually fell in love (or came to believe that she fell in love?) with Dr. Freeman.17 He invited her to accompany him on social occasions and overnight trips and to swim with him in the nude. He convinced Mrs. Zipkin to divorce her husband, move into an apartment above his own, invest in a farm that he bought, and break into her husband's house and steal his new suits. Ms. Roy, a lesbian, consulted Dr. Renatus S. Hartogs for help with sexual problems and depression.18 After several months of treatment, Dr. Hartogs induced Ms. Roy to have sexual intercourse with him, telling her that it was part of her treatment—"It would be good for you . . . if you can love me, you can love another man."19 Ms. Roy's condition worsened, and she was subsequently committed to a mental hospital on two occasions.20 Barbara Noel was being treated by Dr. Jules H. Masserman, an elder statesman of the psychiatric profession, a former president of the 15

Zipkin v. Freeman, 436 S.W.2d 753 (Mo. 1968). Indeed, Mrs. Zipkin could not recall "any direction or demand that Dr. Freeman gave to her that she did not attempt to do." Ibid., at 759. 17 Ibid., at 757. 18 Roy v. Hartogs, 85 Misc. 2d 891, 381 N.Y.S.2d 587 (N.Y. App. 1976). Also see Lucy Freeman and Julie Roy, Betrayal (New York: Stein and Day, 1976). 19 Freeman and Roy, Betrayal, p. 55. 20 Ms. Roy sued for malpractice. She was originally awarded $350,000 in punitive and compensatory damages, but the award was reduced by the New York State Appellate Court to $25,000. New York Times, February 4, 1976, p. 37. 16

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

163

American Psychiatric Association, and the author of 432 professional articles and 16 books.21 Dr. Masserman offered her the "sodium amytal interview," as he did to many of his patients. Sodium amytal ("truth serum") is highly addictive and renders its patients unconscious. On one occasion, and after eighteen years of therapy, Ms. Noel awoke from her sodium amytal treatment on Dr. Masserman's couch to find the naked psychiatrist raping her.22 Ellen Plaisil was, by her own account, a deeply troubled woman. She became a patient of Dr. Lonnie Leonard, who practiced a form of psychotherapy that was allegedly based on the "objectivist" (egoistic) philosophy of Ayn Rand.23 He engaged in a variety of sexual acts with Ms. Plaisil and several other female patients, all of whom appeared to be in awe of him and fearful of terminating their relationship. His "therapy" also included an arrangement in which his patients would pay a fee to view a videotape of the prior session.24 Carolyn Bates had sexual intercourse with Dr X. about once a month for about a year—"They were wordless, compartmentalized encounters, never more than 4 or 5 minutes long, always at the start of the session."25 Ms. Bates stated that she did not know in advance whether Dr X. would propose sexual relations on a particular occasion, or whether Dr. X. saw himself "as party to a fantasy love affair, as a helpful friend, as a therapist throwing caution to the wind, or as an exploiting individual who didn't give a damn."26 Melissa Roberts-Henry began psychotherapy with Dr. Jason Richter because she was experiencing marital difficulties and was feeling guilty about an extramarital affair. At her initiation, they started dating ten days after therapy ended and shortly thereafter began having sexual intercourse. Their sexual relationship lasted three years. As the relationship deteriorated, Ms. Henry became extremely depressed, dysfunctional, subject to panic attacks, and suicidal. She entered therapy with another psychiatrist, Dr. Martha Gay, and when Ms. Roberts-Henry sued Dr. Richter for damages, Dr. Richter sued 21

Barbara Noel with Kathryn Watterson, You Must Be Dreaming (New York: Poseidon Press, 1992). Does this suggest that he was exploiting students as well? 22 Dr. Masserman's insurers have made four out-of-court settlements with his patients, and he was subsequently suspended by the American Psychiatric Association. Minneapolis Star Tribune, F e b r u a r y 14, 1993, p . 10F. 23 Ellen Plaisil, Therapist ( N e w York: St. M a r t i n ' s / M a r e k , 1985). 24

New York Times, November 23, 1982, p. Cl. Ms. Plaisil and another patient were then awarded $250,000 in a malpractice suit settlement. 25

C a r o l y n Bates a n d A n n e t t e Brodsky, Sex in the Therapy Hour ( N e w York: Guilford P r e s s , 1989), p . 3 5 . 26

Ibid.

164

CHAPTER SIX

Dr. Gay, claiming that Dr. Gay, not he, was responsible for Ms. Roberts-Henry's deteriorating condition.27 What do these cases show or, more accurately, what do they illustrate? First, the cases illustrate that the consequences of psychotherapistpatient sexual relations may be quite severe, or, if we want to avoid a causal statement, that psychotherapist-patient sexual relations have often been followed by suicide (or suicide attempts), hospitalization, drug addiction, and marital dissolution. In addition, such patients have often found it difficult to seek therapy from other psychotherapists, and other psychotherapists may refuse to accept such patients lest they find themselves embroiled in professional or legal controversy.28 Second, although the case descriptions do not provide enough detail about the personalities and motivations of the offending psychotherapists, they do suggest considerable variation. Twemlow and Gabbard suggest that most exploiting psychotherapists can be placed in one of three broad categories: the psychotic, the antisocial, and the lovesick.29 Truly psychotic psychotherapists are probably relatively few in number.30 The antisocial psychotherapists, who are more common, typically understand that the sexual relationship is not therapeutic for the patient, although they may be deluded about the harm they are imposing on their patients. The "lovesick" psychotherapist is generally thought to be the most common type of offender. A typical scenario involves a relatively well-functioning (usually) male psychotherapist who may be experiencing difficulties with intimate relations in his own life and who commences sexual relations with a patient who is considerably younger and to whom he "portrays himself as lonely, vulnerable, and needy/' 31 One survey indicated that 65 percent of offenders believed that they were in love with their patients and over 90 percent believed that their patients were in love with them.32 Third, these cases illustrate that sexual relations between psychotherapists and patients can vary with respect to the degree or form of consent and with respect to its perceived relevance to the therapeutic 27 A s p o r t r a y e d o n a PBS "Frontline" d o c u m e n t a r y , " M y Doctor, M y Lover," aired in Burlington, Vt., o n April 11, 1993. Also see Newsweek, April 13, 1992, p . 5 3 . 28 The "Frontline" documentary stated that several psychiatrists refused to accept Melissa Roberts-Henry as a patient after her experience with Dr. Richter. 29 " T h e Lovesick T h e r a p i s t , " in G l e n n G a b b a r d , e d . , Sexual Exploitation in Professional Relationships ( W a s h i n g t o n , D . C . : A m e r i c a n Psychiatric Press, 1989), p . 7 1 . 30 One clinical psychologist at a state mental institution maintained that God told him that his semen would confer eternal salvation on his patients. Ibid., p. 73. 31 Ibid. 32 Ibid.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

165

process. In some cases (e.g., Noel), sexual relations are entirely nonvolitional; they appear to be relatively straightforward cases of sexual assault. In other cases, there is the form if not the reality of consent. In some cases (e.g., Roy) the psychotherapist claims (and perhaps believes) that sexual relations are integral to the therapeutic process and the patient may have understood the relations in that way. In other cases (e.g., Zipkin and Roberts-Henry), the therapeutic relation may have led to the sexual relation but was understood by both parties to be distinct from the sexual relation. Finally, the cases are typical with respect to gender. Surveys suggest that approximately 90 percent of offending psychotherapists are male and approximately 90 percent of the patients are female.33 Thus I shall generally refer to the psychotherapist as "he" and the patient as "she." TRANSFERENCE AND COUNTERTRANSFERENCE

The standard view is that much psychotherapist-patient sexual contact results from a mishandling of transference and countertransference, the four edges of two double-edged swords. Technically defined, transference involves the patient's "projection of feelings, thoughts and wishes onto the analyst, who has come to represent some person from the patient's past."34 More broadly, transference involves an emotional reaction to the psychotherapist that "lies in its excess, in both character and degree, over what is rational and justifiable."35 For many patients, the psychotherapist becomes an allknowing figure, the most important person in the patient's life, the salvation from all her problems. Although the process of psychotherapy will vary with the personality and theoretical approach of the therapist, the patient is typically encouraged to be entirely open and honest, to relax the boundaries that she typically places on her thoughts, feelings, and communications.36 As this process unfolds, many patients begin to experience intensely positive (and sometimes negative) feelings toward their psychotherapists, to idolize and fantasize about them, and to have 33 Jorgenson, Randies, and Strasburger, "The Furor," p. 651. Female-male and homosexual relationships have also occurred. 34 Simmons v. United States, 805 F.2d 1363 (9th Cir. 1986), at 1364, q u o t i n g Stedman's Medical Dictionary 1473 (5th Lawyers Ed., 1982). 35 S i g m u n d Freud, "The Dynamics of the Transference," in 2 Collected Papers at 312, as cited in Phyllis Coleman, "Sex between Psychiatrist and Former Patient: A Proposal for a "No Harm, No Foul' Rule," 41 Oklahoma Law Review 1 (1988), at 6. 36 Coleman, "Sex between," p. 5.

166

CHAPTER SIX

strong sexual feelings towards them, a response that is heightened by the fact that the discussion of sexual desires is often at the center of the therapeutic process.37 In some psychotherapeutic approaches (e.g., traditional psychoanalysis), the "resolution" of the patient's transference is itself a focal point for the psychotherapeutic process. "Working through" her feelings about her psychotherapist is a way for the patient to "work through" her emotional conflicts about other important persons in her life.38 From this perspective, the resolution of the transference is a necessary and perhaps sufficient condition of successful therapy. The patient is cured when she begins to react to the psychotherapist as he "really is rather than reacting to him as if he were some other person in her life."39 In other psychotherapeutic approaches, the transference process is regarded as extrinsic to the main work of psychotherapy, but it must be managed if the therapy is to be successful. Just as the psychotherapeutic relationship can be intense for the patient, it can also be emotionally intense for the psychotherapist, who may experience "countertransference." Even though a therapist is trained to believe, "She's not in love with you, stupid," it is no doubt hard for the psychotherapist to resist taking the patient's attraction quite personally. In addition, the discussion of sexuality can be a titillating and arousing experience for the psychotherapist as well as the patient.40 The ubiquity and intensity of transference and countertransference explains much of the strength of the strictures against psychotherapistpatient sexual relations. The standard view is that the transference/countertransference interaction will have serious harmful consequences if the psychotherapist acts on those feelings or permits the patient to do so.41 The patient is likely to experience all of the pain of unrequited 37 "At the moment the woman patient enters therapy and begins to sense the special interest of her therapist—an interest that goes beyond the usual concern a doctor has for a patient—she is likely to respond with a sense of excitement. The patient begins to think of nothing else; her dreams are suddenly filled with the presence of the doctor and his comforting contact with her." Sydney Smith, "The Seduction of the Female Patient," in Gabbard, Sexual Exploitation, pp. 59-60. 38 "p o r example, although the patient may be unable or unwilling to articulate feelings of having been ignored and misunderstood as a child, she will repeat her reaction to those feelings with the therapist by accusing him of misunderstanding or ignoring her." Coleman, "Sex between," p. 7. 39 Ibid., p. 8. 40 Ibid., p. 45. Moreover, if, as Henry Kissinger once remarked, "power is the greatest aphrodisiac," we can add to the brew another source of sexual stimulation. 41 "When a psychotherapist permits a patient to 'act out' transference feelings and wishes, he deprives her of the opportunity to explore them and to understand their significance in the larger context of her life. Based on infantile ideas and emotions, the

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

167

love. Perhaps more important, she may experience sexual relations with her psychotherapist as deeply incestuous, with all the associated feelings of shame, guilt, and betrayal.42 Moreover, sexual relations effectively stop the therapeutic treatment itself—"When sex begins, therapy ends." The standard view also maintains that the management of transference and countertransference is a job with a sharp division of labor. Whereas the patient is allowed or encouraged to experience the feelings of transference, the psychotherapist has a responsibility to consciously monitor the interaction, a task for which he has been trained. The psychotherapist must see to it that the appropriate boundaries are observed. Of course, not all instances of sexual relations between psychotherapists and patients are best understood as resulting from the mishandling of transference. Some psychotherapists are simply predators. Others may believe that sexual relations can be a valid part of the therapeutic process, although the sincerity with which this belief is held seems belied by the fact that the target patients are typically attractive and younger. And some patients may attempt to seduce their psychotherapists apart from any difficulty with transference. But whatever the motivation, psychotherapists are generally thought to be exploitative whenever they engage in sexual relations with their patients. It is to the analysis of exploitation that we now turn. GAIN AND HARM

I have suggested that a transaction between A and B is exploitative only when (1) A expects to gain something of value from a transaction or interaction with B, and (2) B is harmed by the transaction (harmful exploitation) or gains less than he or she should (mutually advantageous exploitation). Let us consider each of these conditions in the current context. The Psychotherapist

I believe that the language of exploitation is entirely appropriate in the present context because sexual gratification is a form of gain that is no more mysterious than financial gain, the stuff of most exploitation discourse. Such gratification also helps to distinguish exploitative sexual harassment, as when a professor proposes exchanging a grade for patient's behavior is painful because it is false and humiliating once its source is revealed." Jorgenson, Randies, and Strasburger, ""Furor," p. 656. 42 Simmons v. United States, 805 F.2d 1363 (9th Cir. 1986), at 1367.

168

CHAPTER SIX

sexual favors, from nonexploitative sexist harassment, as when a professor makes remarks that are demeaning to women. To test this view against our intuitions, suppose that a psychotherapist engages in sexual relations with a patient from altruistic motivations. He is not sexually attracted to this patient and is putting his career at risk because he believes that sexual relations will be beneficial for this particular patient, even though he accepts the profession's view that such relations are generally not beneficial. Does this view imply that the psychotherapist is not engaging in exploitation, even if the patient is harmed? I am inclined to think so, and that seems to me an implication we should not want to resist, although we might also want to condemn the psychotherapist for negligence or recklessness in his treatment. Although the first dimension of exploitation—gain to the exploiter— appears to be met here, the present context provides an opportunity to pursue that dimension in a bit more detail. We can begin by observing that to say that the psychotherapist gains from the sexual relation is not to claim that he gains, all things considered. If, for example, a psychotherapist should contract syphilis from a sexual encounter with a patient, I take it that he has exploited the patient nonetheless.43 Why? We might say that the behavior is exploitative because the psychotherapist gains from the act itself, although not from the consequences of the act. Or we might say that the motivation of the exploiter is crucial, that A exploits B (or acts exploitatively) if A expects to gain from the transaction and if that expectation is an important motivating consideration. We can also note that just as the conception of "gain" that is presupposed by exploitation is sufficiently protean to encompass nonfinancial gains, it is also sufficiently protean to encompass gains that are not self-serving in the ordinary way, that include even altruistic and conscientious purposes and desires. Suppose that Dr. A is seeking to test a new treatment for AIDS. Although the treatment is promising, it does not represent the best available mix of benefits and risks. Dr. A believes that he can get Patient B to consent to the experimental treatment by manipulating B's sense of guilt about having contracted the disease. We might say that A has exploited B for the purposes of medical science (or on behalf of other potential patients) even though it may not serve the physician's self-interests. We can see the exploitative dangers inherent in other forms of nonfinancial gain if we look more closely at intrinsic work satisfaction. Indeed, precisely because intrinsic work satisfaction and sexual plea43

See Joel Feinberg, Harmless Wrongdoing (New York: Oxford University Press, 1988), p. 193.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

169

sure represent different forms of psychic income that involve risks of exploitation, let us pause to consider them together. First, intrinsic work satisfaction and sexual pleasure are relatively "free" goods (and noncaloric, too). They can both be understood as payoffs in a positive-sum game, for one party's gain does not subtract from the other party's gain. Whereas a higher wage for A means less profits for A's employer or higher prices for customers or lower wages for A's coworkers, A's intrinsic work satisfaction does not entail any of these consequences and may even be to the benefit of others. If A's job carries high intrinsic work satisfaction, A may accept a lower wage than he or she might otherwise require, thereby leaving more financial resources available for others. Similarly, sexual relations are among the more important and easily produced sources of "free" positive value in the world. True, people make great investments in their personal relationships and put themselves at considerable risk of being rejected, disappointed, or abandoned. Nonetheless, the pleasure, or shall I say marginal utility, that a couple receives from their sexual interactions is effectively created ex nihilo, and does not subtract from the amount of such satisfaction that is available to others. Just as we may worry about the distribution of an economic or financial social surplus, we may worry about the distribution of intrinsic work satisfaction and sexual pleasure. It appears, for example, that intrinsic work satisfaction is distributed less equally than workers' satisfaction with their income, and it might be thought that we should try to monetarily compensate those whose intrinsic work satisfaction is low.44 Similarly, even if both A and B gain from their sexual relations as contrasted with their (no sex) baseline, we may think it objectionable if the amount of sexual satisfaction is highly inequitable, apart from any problems associated with the belief that the gains are unfair, which may be a source of tension that detracts from the total gain. Sexual relations could constitute a case of mutually advantageous exploitation. The third similarity between intrinsic work satisfaction and sexual pleasure, and of particular importance to exploitation, concerns their potential to generate undesirable incentives or conflicts of interest—a feature that both share with monetary gain. Suppose that Dr. A is a salaried resident in surgery at a hospital. A works long hours and 44

See Robert E. Lane, The Market Experience (Cambridge: Cambridge University Press, 1991), pp. 407ff. Lane goes on to point out that intrinsic work satisfaction is marred by several social defects. Unlike the gains from income, the gains from work satisfaction cannot be shared. They do not necessarily lead to the production of goods that others want. And unlike income, work satisfaction is not taxed, thereby making equality of sacrifice impossible.

170

CHAPTER SIX

performs numerous procedures not because she is compensated on a fee-for-service basis, but because she loves her work. Still, if A's love for surgery distorts her medical judgment and leads her to recommend and perform unnecessary and potentially harmful surgery, then we may regard A's intrinsic work satisfaction with a jaundiced eye. We may wish that she enjoyed her work a little less.45 Indeed, there is a sense in which the dangers of exploitation are greater when they derive from interactions that are plausibly regarded as positivesum than when they derive from interactions that are zero-sum at the margin. When a physician stands to gain monetarily from a transaction with a patient, he knows that his marginal gain comes at his patient's (or his patient's insurer's) expense, and a reasonably conscientious physician will be sensitive to such factors. By contrast, a physician may have no reason to think that intrinsic work satisfaction is costly to anyone, and so may not be sufficiently sensitive to the dangers it creates. The danger inherent in intrinsic work satisfaction also illustrates that exploitation can be unintentional. After all, A does not seek to derive intrinsic work satisfaction from surgery at her patients' expense. Rather, A gets intrinsic work satisfaction because she believes her surgery promotes their interests. The problem is that her love for surgery leads her to overestimate its benefits and underestimate its risks. The problems inherent in the search for intrinsic work satisfaction may be particularly acute for the psychotherapist. Harry Stack Sullivan maintained that psychotherapy was among the most difficult and potentially least rewarding of professional activities, precisely because the therapist was called upon to suppress his own needs and desires in what is otherwise an intimate human relationship.46 On this view, the standard joke about psychiatrists captures something important. The psychotherapist is right to worry about his fee, for that is all the compensation that he should expect to receive. The Patient

Unlike other forms of (alleged) exploitation that I consider in this book, the sexual exploitation of patients is virtually universally viewed as a case of harmful exploitation. I see no reason to doubt that view. The question is to decide just what it does or might mean. 45

Similar things m a y be said about lawyers w h o love to go to court. "There is n o fun in psychiatry. If you try to get fun out of it, you pay a considerable price for your unjustifiable optimism. If you do not feel equal to the headaches that psychiatry induces, you are in the w r o n g business." The Psychiatric Interview (New York: W. W. Norton, 1954), p. 10, quoted in Stuart Twemlow and Glenn Gabbard, "The Lovesick Therapist," in Gabbard, Sexual Exploitation, p . 113. 46

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

171

Psychotherapist-patient sexual relations typically involve two sorts of harm. First, they involve what we might call a harm to the person of the patient. Therapists who have treated sexually exploited patients report that 90 percent have suffered negative consequences, including hospitalization and suicide attempts.47 Second, they involve a harm to the professional relationship between the patient and the psychotherapist, in two ways.48 They affect the patient's welfare, because they reduce the prospects for a beneficial therapeutic relationship, even if they do not result in a palpable worsening of the patient's condition.49 In addition, psychotherapist-patient sexual relations violate the values underlying the professional relationship, even if they do not affect the patient's welfare, just as we might say that a physician harms the professional relationship with his patient if he violates his patient's confidentiality, even if that violation did not affect the patient's physical or mental well-being, say, because the patient never learns that this has occurred. It is difficult to gain systematic evidence about the effects of psychotherapist-patient sexual relations on patients. We cannot study the effects of psychotherapist-patient sexual relations in quite the way in which we might compare different treatments for breast cancer, say, by comparing the mortality rates of women (with comparable diagnoses and controlling for other variables, such as age) who receive total mastectomies (where the entire breast is removed) with those who receive lumpectomies (where only the cancerous lump is removed). We lack noncontroversial diagnoses (such as a "stage 1" tumor) of the pretreatment psychological status of the patient. We lack noncontroversial measures of outcome (such as recurrence or mortality). And even if we had such measures, it would be difficult to isolate the effects of sexual relations as distinguished from other dimensions of the therapeutic relationship. These difficulties notwithstanding, I see no reason to doubt the widespread view that psychotherapist-patient sexual relations are typically harmful, frequently devastating, and virtually always below an appropriate standard of skilled treatment with expected positive 47

Jorgenson, Randies, a n d Strasburger, "Furor," p. 683. A survey of practitioners suggests that at least 90 percent of patients are seriously harmed by sexual contact with their psychotherapists. See J. Bouhoutsos et al., "Sexual Intimacy between Psychotherapists a n d Patients," 14 Professional Psychology: Research and Practice 195-96 (1983). 48 See Jorgenson a n d Sutherland, "Fiduciary Theory," p. 473. 49 "[The therapist] violates a n agreement that t h e therapeutic relationship exists solely to serve the patient's needs. By stepping out of his role, the therapist betrays the patient's trust and fractures the structure of continuity and constancy essential to accomplishing the therapeutic task." Jorgenson, Randies, and Strasburger, "Furor," p. 657.

172

CHAPTER SIX 50

benefits for the patient. But rather than try to describe the effects of psychotherapist-patient sexual relations in further detail, I want to make two more general remarks. First, there is no reason to assume that sexual relations between psychotherapists and patients are seriously harmful to the patient in every instance, particularly given that we are apt to learn about only those cases that result in legal actions or become the bases of books, movies, and television shows.51 Indeed, it is possible that sexual relations sometimes produce a treatment outcome more positive than could have been attained in other ways or that they produce other beneficial outcomes, including relationships that have led to reasonably successful marriages. Second, it is not clear against what baseline we should evaluate the claim that the patient has been harmed by her psychotherapist. If we adopt a pretreatment baseline, we say that the patient is harmed only if she leaves the therapeutic encounter worse off than when she entered therapy. If we adopt a normative baseline, we say that the patient is harmed if she leaves the therapeutic encounter less well off than she could otherwise have reasonably expected. There is reason to think that the normative baseline is the better view: a patient whose condition is deteriorating has not been harmed by a treatment if it retarded the pace of deterioration more effectively than other interventions would have done, and a patient has arguably been harmed even if she did not leave treatment worse off than when she entered treatment, if only because she paid for treatment and received no benefit. Since victims of psychotherapist-patient sexual relations will typically have been harmed by reference to both baselines, the choice between the pretreatment and normative baselines is of little practical relevance in the current context. Nonetheless, this brief discussion suggests that there are good theoretical reasons to understand the notion of harm by reference to a normative baseline. CONSENT

In an old joke, a man accused of murder says, "I didn't kill anyone, and, in any case, it was self-defense. In addition, I was insane/' A similar view is sometimes advanced with respect to the role of consent in psychotherapist-patient sexual relations. One argument main50

See Gabbard, Sexual Exploitation, p. xi. "In the politically correct scheme of things, the victim is always severely h a r m e d . . . but in actuality, while many exploited patients are devastated, others emerge unscathed/' These remarks are attributed to Thomas Gutheil, Professor of Psychiatry at Harvard Medical School, in Carl Sherman, "Behind Closed Doors," Psychology Today, May, 1993, p. 64. 51

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

173

tains that psychotherapist-patient sexual relations are exploitative because "meaningful, informed consent to a sexual relationship" cannot be given under these conditions.52 Another argument maintains that psychotherapist-patient sexual relations would be exploitative even if they were consensual because "there can be no consent that legitimizes a clearly unethical and illegal act."53 Although it is contradictory to maintain that one did not kill anyone and that one killed in self-defense, it is not contradictory to maintain that (1) patients do not typically give "meaningful, informed" consent to sexual relations with their therapists, and (2) even if the patient gives "meaningful, informed" consent, it would still be wrong for psychotherapists to engage in sexual relations with their patients. Still, there are several reasons for thinking that it is important to determine whether patients do, or can, give "meaningful informed" consent. First, to say that consent cannot legitimate an unethical act begs the question as to whether an act that would be unethical if it were nonconsensual ceases to be unethical if it is consensual. Second, if A's action would be unethical even with B's consent, such consent would at least nullify the relevance of the absence of consent as a reason for regarding A's action as unethical. For example, we may say that rape is wrong because it involves nonconsensual sexual relations and that prostitution is wrong even if it involves consensual sexual relations, but the reasons for thinking that consensual prostitution is wrong must be different from the reasons for believing that rape is wrong. Third, if psychotherapist-patient sexual relations are wrong (in part) for reasons other than the absence of consent, we must determine whether these reasons support the claim that such relations are exploitative. Fourth, even if sexual relations between psychotherapists and patients could be both exploitative and consensual, the moral weight and, in particular, the moral force of that exploitation may turn on consent. The presence of genuine consent typically counts against the justification of a legal prohibition, even when it is insufficient to make the act itself (between consenters) morally acceptable or nonexploitative. It is perfectly coherent to argue that rape both is wrong and should be prohibited because it is nonconsensual, but that consensual prostitution should not be prohibited even if it is wrong and that it should not be prohibited precisely because it is consensual. So if psychotherapist-patient sexual relations are consensual, then it must 52

Jorgenson and Sutherland, "Fiduciary Theory," p. 491. Kenneth S. Pope, "Therapist-Patient Sex Syndrome: A Guide for Attorneys and Subsequent Therapists to Assessing Damage," in Gabbard, Sexual Exploitation, p. 47. 53

174

CHAPTER SIX

be shown why they ought to be prohibited, their wrongness notwithstanding. The centrality of the principle of informed consent in medical practice gives us another reason for looking closely at the quality of the patient's consent. We do not view the physician as the benevolent and paternalistic protector of the patient's interests. Rather, we typically start from the assumption that the patient is competent to choose among treatments that carry various benefits and risks.54 True, the consent of a competent patient may not be sufficient to legitimate treatments or procedures that are not medically indicated. But given our respect for the autonomy of the patient, we must ask whether psychotherapy patients can give informed consent to sexual relations with their psychotherapist, and if not, whether that view is consistent with respect for the patient's autonomy. As a first approximation, let us say that sexual relations can be consensual only if (I decline here to say if and only if) (1) the patient is competent to give consent, (2) the consent is uncoerced, (3) the consent is sufficiently informed, and (4) the patient's judgment is not distorted.55 I did not state that the patient's consent must be volitional, because the absence of volition does not represent a defect in consent, but rather the complete absence of consent. If Dr. Masserman had sexual relations with Barbara Noel when she was unconscious from sodium amytal, there is not even the appearance of (invalid) consent, much less the presence of valid consent.56 It was simply a case of rape.57 In many other cases, however, the issue of exploitation arises precisely because the patient's actions are clearly 54

Of course, we reject that assumption in some cases, as when we appoint surrogate decision-makers for children and the severely demented. See Allen E. Buchanan and Dan W. Brock, Deciding for Others: The Ethics of Surrogate Decision Making (New York: Cambridge University Press, 1989). 55 I am here concerned with considering what might count as defects in consent, not with providing a precise account of the necessary and sufficient conditions of consent. 56 That Dr. M a s s e r m a n w a s a psychiatrist does nothing to distinguish this sort of case from t h e behavior of a n anesthesiologist w h o w a s reported to have placed h i s penis in t h e m o u t h s of anesthetized patients o n more than 160 occasions. Irwin Perr, "Medicolegal Aspects of Professional Sexual Exploitation," in Gabbard, Sexual Exploitation, p . 213. Indeed, precisely because there is n o ambiguity about consent w h e n B exercises n o volition, it seems m o r e natural to refer to these situations as cases of sexual a b u s e t h a n as cases of sexual exploitation. 57 Some cases lie o n the margin. Consider Carolyn Bates's account of h e r r e s p o n s e to Dr. X: "For several meetings following [the first episode], Dr. X held m e for longer m o m e n t s at t h e e n d of each h o u r until o n e time h e pulled m y hips u p against his genitals, a n action that both frightened m e a n d confused m e . A s with everything h e did in therapy, however, I did n o t challenge his action." Bates a n d Brodsky, Sex in the Therapy Hour, p. 32.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

175

volitional. The patient may believe that she is consenting to the sexual relations or may, indeed, initiate the sexual relations. In these cases, the question is not whether there is volition, but whether the patient's consent is deficient with respect to competence, freedom, information, or judgment. Competence

Only persons who have the requisite mental competence are capable of giving valid consent. We frequently speak of the sexual exploitation of children because they do not have the ability to fully understand what is happening to them.58 It might be thought that the "underlying distress that brought the patient into treatment" is evidence that she is not competent to give informed consent to sexual relations.59 By itself, however, this argument would prove too much. A reasonably high proportion of those patients are functioning at a high level prior to therapy.60 Most patients are thought to be competent to make important choices about their lives. They may borrow money, make contracts, and get married. They can choose to enter or terminate psychotherapy. They may choose to accept or reject modes of therapy, such as tranquilizers or antidepressants.61 So if there is a defect in the patient's competence to consent to sexual relations with her psychotherapist, it is not likely to be a function of the patient's general or pretherapeutic competence. Rather, it will be a problem that is specifically connected to the psychotherapeutic process and to the fact that the object of the consent is sexual relations. Coercion

Bracketing issues of competence, it might be claimed that a patient's consent is not valid because it is the product of coercion. In examining this claim, I shall assume that the psychotherapist does not threaten 58

Or, if they do understand, they may not have the ability to resist, in which case it is coercion, not incompetence, that compromises the quality of their consent. 59 Paul S. A p p e l b a u m a n d Linda Jorgenson, "Psychotherapist-Patient Sexual Contact after Termination: A n Analysis a n d a Proposal," 148 American Journal of Psychiatry 1466 (1991), at 1469. 60 One study found that one-third of such patients were themselves training or practicing therapists and came to therapy with relatively minor problems. Supra, note 51, at 64. 61 Indeed, the previous remarks are frequently applied to those with reasonably severe psychological disorders, where we are quite reluctant to set the standards of competence too high lest that effectively nullify the right of the mentally ill to refuse commitment or other medical interventions.

176

CHAPTER SIX

to use physical force if the patient refuses to engage in sexual relations or threaten to engage in other obviously improper behavior (e.g., refusing to prescribe appropriate drugs or violating the patient's confidentiality) should she refuse. For if the psychotherapist makes these sorts of threats should the patient refuse his advances, then it would be uncontroversial that the patient's consent is coerced. Suppose, then, that the psychotherapist proposes sexual relations and the patient does not refuse. It might be argued that the psychotherapist has implicitly threatened to abandon the patient if she refuses to engage in sexual relations. Given that the patient may feel that the consequences of refusal would be disastrous to her well-being, she may believe that she has no choice but to acquiesce. The question is what this shows. Consider, for example, a patient with a gangrenous leg who must decide whether to undergo amputation. The patient has two choices: (1) amputation and probable survival, or (2) no amputation and probable death. Now, we do not say that anyone who faces a choice between (1) and (2) is effectively coerced into choosing (1) simply because (2) is unacceptable. If we regarded such choices as coerced, then we would treat the patient's consent to the amputation as unnecessary and invalid. But we do not. If the absence of any reasonable alternative to the physician's proposal is not sufficient to render consent irrelevant in the case of surgery, it is not clear that the absence of any reasonable alternative is, per se, sufficient to render irrelevant the patient's consent to sexual relations. But that misconstrues the way in which coercion undermines consent. As we have seen in previous chapters, A coerces B when A proposes to make B worse off than she is morally entitled to be with respect to the coercer. It is one thing if the less acceptable alternative (death from gangrene) is caused by factors external to the physicianpatient relation. That will typically not constitute coercion. But it is quite another thing if the less acceptable alternative (withdrawal of affection) is created by the relation itself. If the psychotherapist has explicitly or implicitly proposed that the patient will be worse off than she is entitled to be (with respect to him) if she does not engage in sexual relations, then the charge of coercion can be sustained. And I think it can be sustained in many cases. But suppose that there are no implicit or explicit threats of this sort. Still, it might be argued that there are two dimensions of the relationship between psychotherapist and patient that effectively coerce the patient into consenting to sexual relations. The first dimension involves the (now) familiar claim that there is an inequality of (bargaining) power between the parties. The second and less familiar dimen-

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

177

sion involves the effect of transference on the patient's ability to choose. Now there is no doubt that there are important asymmetries of power between the psychotherapist and the patient. The psychotherapist is the professional; the patient is the client. The psychotherapist is "healthy'/ the patient is "sick" or "has problems." The psychotherapist has extensive information about and from the patient; the patient knows relatively little about the psychotherapist.62 The psychotherapist is (in our discussion) male; the patient is female. The problem is not to show that there is an imbalance of power, but what to make of that fact. Somewhat similar inequalities of power typically exist between physicians and patients as well as between other professionals (e.g., attorneys and their clients) and with respect to the entire range of professional issues. Physicians' recommendations for treatment are often accepted uncritically by patients. Even when physicians attempt to provide patients with detailed information about the benefits and risks of various courses of treatment, patients may be interested in the answer to only one question: "What would you do?" I do not think that decisions made in such ways are incompatible with patient autonomy even if they do not approach our ideal of patient decision-making. If a patient has decided that it is in her interests—including her interest in not being burdened by having to make medical decisions—to regard her physician's advice as dispositive, then she may exercise her autonomy by treating the physician's view in that way.63 There is another reason not to require equality of power as a background condition for genuine consent, despite the frequent suggestions that meaningful consent requires such equality: if we were to say that consent is not meaningful until we rectify the background inequality of power, then there would be much less reason to seek the consent of the less powerful party whenever an inequality of power obtained. If this is right, then an imbalance of knowledge and expertise between physician and patient does not, in itself, seem to preclude genuine consent, or, if it precludes genuine consent, the (defective) consent may still be sufficient to legitimate the physician's action. If psychotherapy patients can give informed consent to the use of antidepressants and tranquilizers, despite the inequality of power, 62 Jacqueline R. Castel, "Women's Sexual Exploitation in Therapy," 49 University of Toronto Faculty of Law Review 42 (1991), at 43.1 would add, however, that there is at least one dimension of the relationship that may give more power to the patient, namely, that the psychotherapist may be at greater reputational and professional risk. 63 See A r t h u r Kuflik, " T h e Inalienability of A u t o n o m y , " 13 Philosophy & Public Affairs 271 (1984).

178

CHAPTER SIX

then the question arises as to whether they can give informed consent to sexual relations in the face of such inequalities. Here there are at least three possibilities: (1) the quality of the patient's consent is, in fact, more compromised when the object of the consent is sexual relations; (2) the standards for what constitutes valid consent are higher when the object of the consent is sexual relations; and/or (3) a patient can give suitably voluntary consent to sexual relations, but there are good reasons for thinking that the patient's consent cannot legitimate sexual relations with the psychotherapist although it can legitimate other treatment options offered by the psychotherapist. Let us consider these possibilities. Is there reason to think that the voluntariness of consent is, in fact, more compromised when the object of consent concerns sexual relations? As a general matter, it seems not. Consider sexual relations among adolescents. Suppose that A, "the popular jock," tells B that he will continue their dating relationship only if she has sexual intercourse with him. Suppose that B would most prefer (1) a dating relationship without sexual intercourse, but, given the perceived status that comes with a dating relationship with A, B prefers (2) a dating relationship with sexual intercourse to (3) no dating relationship at all. Because A prefers (2) to (3) to (1), they do have sexual intercourse. Now, we may think that A's behavior is less than admirable, but we are unlikely to say that B has not consented to sexual relations with A, inequalities of power notwithstanding. How is this case to be distinguished from that of the patient? It might be distinguished in three ways. First, we may think that the patient's consent is less voluntary than the adolescent's consent because the inequality of power is greater, just as we regard sexual relations between a thirty-year-old male and a fifteen-year-old female as not fully voluntary ("statutory rape"), whatever the appearance of consent. Second, we may think that transference compromises the informational dimension of consent, that whereas the adolescent knows what she is doing, the patient does not. I will consider that view in the next section. There is a third possibility. We may think that even if the quality of the patient's consent is no weaker than the adolescent's, the standards for valid consent are higher, or that the patient's consent cannot legitimate sexual relations, whereas the adolescent's consent can legitimate sexual relations with another adolescent (but not with an adult). If this is so, we will have to ask why the standards are higher or why consent seems to legitimate sexual relations in some cases but not in others. But we can note for now that there do seem to be realms of sexual relations in which caveat emptor seems to reign (as between

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

179

adolescents and adolescents and between adults and adults) and realms in which it does not (as between adults and adolescents). Here, the argument is not so much that the adolescent's consent is weaker when the other party is an adult, although that may be part of the story, but that adults have a (quasi-)fiduciary responsibility not to engage in sexual relations with adolescents whereas chronological equals do not have such obligations. On this view, it is not the weakness of the younger party's consent that is doing the moral work. Rather, it is the strength of the older party's obligation. Similarly, we may argue that patient consent cannot legitimate sexual relations because the psychotherapist has a duty to refrain from even consensual sexual relations with his patient. I return to this argument in the discussion of "Consent as Irrelevant" later in this chapter. Information and Judgment

Valid consent must be informed and (reasonably) rational as well as free or uncoerced. We may think that patients cannot give valid consent to sexual relations with their psychotherapists because transference undermines or skews the cognitive or informational requirements of valid consent, in two ways. First, it may be argued that the process of transference is infantalizing, that it temporarily converts a chronological adult into a psychological child who is dependent upon her therapist. From this perspective, transference generates a local but pronounced decisional incompetence or distortion in her judgment. Although the patient may be generally competent to consent to sexual relations outside of psychotherapy, and although the patient may be competent to consent to nonsexual dimensions of her psychotherapeutic relation, transference does undermine her competence to consent to sexual relations with her psychotherapist. Second, transference may distort the patient's understanding of the person with whom she is having sexual relations. By way of analogy, suppose that B sleeps with a blindfold, that her husband arranges for A to take his place in bed, and that B has sexual relations with A. We may say that B has consented to have sexual relations with the person with whom she thought she was having sexual relations, but that she has not consented to have sexual relations with A. From this perspective, the psychotherapist's use of the patient's transference to gain her consent to sexual relations may amount to a similar form of fraud.64 64

It appears that Dr. Hartogs practiced a more traditional form of fraud on Ms. Roy, when he told her that the point of transference was to develop her capacity to respond

180

CHAPTER SIX

He does and should know that the patient may find herself attracted to him because transference—by definition—undermines her cognitive capacity to perceive him for the person that he really is.65 We should not press this argument too far. Transference, broadly understood, is a part of most romantic relationships—"People in love have notoriously bad judgment about the qualities of their lovers. . . ."66 We commonly say that men "marry their mothers" and that women "marry their fathers," and if we were to take these sayings too seriously, then most marriages would not be truly consensual. There might be good reasons for thinking that the degree of transference that characterizes most (although perhaps not all) nontherapeutic relations is not sufficient to undermine the presumption that the decision-maker understands what she is doing, but that psychotherapy is different. But I doubt that this entirely explains the different stances that we take. I think we must take a different tack. Informed Consent

Up to this point, we have been largely operating within the framework of the principle of consent. We have identified several criteria (competence, absence of coercion, and information) that are requirements of informed consent and have asked whether patients can give informed consent to sexual relations with their psychotherapists. We are not in a position to say whether any patients actually give informed consent to sexual relations with their psychotherapists, but there are good reasons to think that the quality of the consent is substantially compromised in many cases. The crucial point here is that within this framework, one must show that the quality of the patient's will has been significantly compromised if we are to reach the conclusion that it is illegitimate for the psychotherapist to have sexual relations with her. That may be the right view, but it is not the only view. CONSENT AS IRRELEVANT: THE FIDUCIARY MODEL

As we have seen, it is sometimes argued that even if the patient can and does give informed consent to sexual relations with her psychotherapist, such consent cannot legitimate such relations. The most sexually to him and then transfer that capacity to other men. See Freeman and Roy, Betrayal, p. 55. 65 "Consent, under such circumstances, rather than being to a relationship with her therapist, is to a relationship with her illusion of him." Castel, "Women's Sexual Exploitation," p. 47. 66 Appelbaum and Jorgenson, "Psychotherapist-Patient," p. 1470.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

181

important route to this view maintains that the psychotherapist has a fiduciary obligation to promote the best interests of the patient, an obligation that cannot be set aside by the patient's consent—"A psychiatrist is not to . . . respond if the patient takes off her clothes and dances on the table."67 The present question is this: Does the fiduciary argument apply even if the patient gives informed consent? Or does the fiduciary argument implicitly assume that the patient cannot or does not give informed consent? To fix ideas, let us consider the role of consent in three models of business or professional relationships: (a) an adversary relationship; (b) a principal-agent relationship; and (c) a fiduciary relationship. In an adversary relationship, we assume that the parties' interests are at least somewhat in conflict and that each party is entitled to promote her own interests. A range of mutually beneficial prices between buyer and seller is possible, but it is legitimate for both parties to press for a more favorable position within the zone of agreement. In adversary relationships, consent legitimates the result, at least in the sense that B cannot claim that her rights have been violated. I believe that the adversarial model captures at least part of the traditional view of the morality of "dating" relationships (which is not to say that this is the best moral view). On that view, it is legitimate for a male to seek to have sexual relations with a female, so long as such relations are consensual. There is no assumption that A is acting on behalf of B's interests. A is not required to refrain from sexual relations with B because it would be better for B if A so refrained. By contrast, the normative structure of the principal-agent relationship assumes that the agent (A) acts on behalf of the principal (B). For example, the shareholders in a corporation are principals, and the managers are their agents; investors are principals, and stockbrokers are agents; home sellers are principals, and real estate brokers are their agents.68 As investors and home sellers well know, the normative structure of principal-agent relationships is threatened by "agency risk," by the fact that the interests of agents are not identical to the interests of principals and the "principal cannot perfectly and 67 Dr. Rose, a Colorado psychiatrist, on "Frontline" documentary. As one advisory brochure puts it, "Sexual relationships between therapists and their patients are never the fault of the patient. . . . It is your therapist's responsibility not to exploit your feelings of love and attraction." From "If Sex Enters into the Psychotherapy Relationship," a brochure published by the American Psychological Association (Washington, D.C., 1987), reprinted in Bates and Brodsky, Sex in the Therapy Hour, p. 224. 68 See John Pratt and Richard Zeckhauser, "Principals and Agents: An Overview," in John Pratt and Richard Zeckhauser, eds., Principals and Agents (Boston: Harvard Business School Press, 1985), at 2.

182

CHAPTER SIX

costlessly monitor the agent's actions and information/'69 Corporate managers (agents) may prefer a larger salary although it may not maximize the profits of the shareholders (principals); stockbrokers (agents) may have an interest in generating commissions by "churning" the accounts of investors (principals), while investors want only to pay commissions on transactions that promote their interests; a real estate broker (agent) may prefer to sell a principal's house below the going market rate because the commission on the marginal value (the difference between the proposed price and the market price) is less than the cost to the broker of the increased advertising and "showings" necessary to secure the market price. In this context, requiring the principal to consent to the agent's actions may enable the principal to reduce agency loss, although obtaining such consent is often too costly or ineffective. A fiduciary-beneficiary relationship is a variant of a principal-agent relationship that responds to the problems just noted. Although they have the same normative structure, we are apt to employ what I shall call the standard principal-agent model when the principal retains reasonably effective ability to monitor, control, and direct the activities of the agent. We are apt to adopt a fiduciary model when there are significant structural inequalities between the principal and agent with respect to competence, expertise, or power. There are several important differences between the moral and legal requirements of a standard principal-agent relation and a fiduciary relation.70 For example, whereas the "agent" may have relatively fixed obligations to the principal, a fiduciary may be assigned an "open-ended duty to act" on behalf of the beneficiary. Whereas both agents and fiduciaries may not lie to the principal, the fiduciary will be assigned an affirmative duty to disclose information.71 Whereas we can generally assume that the principal's consent to a transaction that also benefits the agent (as when the agent makes a commission on a sale) is voluntary and informed, it is sometimes assumed that a beneficiary's consent to a transaction that benefits the fiduciary is not valid unless that assumption can be rebutted, say, by the fact that the beneficiary has received independent advice with respect to the transaction.72 69

Ibid., at pp. 2-3. Agency risk is a problem for agents as well as principals, because a principal may not employ an agent (to the potential agent's disadvantage) unless the agent can satisfactorily resolve the problem of agency risk. I say home sellers rather than buyers because there is generally no representation that the agent works on behalf of the buyer. 70 See Robert C. Clark, "Agency Costs versus Fiduciary Duties," ibid. 71 Ibid. 72 "[Fiduciaries] must show that t h e benefit was conferred after fair a n d full dis-

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

183

Let us consider the physician-patient relationship in light of these three models. That relationship could be understood as incorporating elements of the three models, or, more accurately, as reflecting each of the three models at different points in the relationship. Consider B's choice of physician. Assume (as may be close to the truth) that patients are able to choose among competent physicians, all of whom practice under a similar normative structure, that is, each physician will attempt to act on behalf of the medical interests of his patients.73 Also assume (although this is less true) that each physician offers a readily knowable package of fees and expertise and that patients have incentives to economize on their fees. Under those conditions, the adversary model applies to the patient's choice of a physician, because each physician is trying to get the greatest income that he can command for his expertise and each patient is trying to pay the least possible for a given level of expertise. Once the patient has chosen a physician, however, a new normative structure characterizes the relationship—the physician is to act on behalf of the patient's interests. At the point of deciding whether to undergo a specific treatment, however, we may want to adopt something close to an agent-principal model. We believe that a competent patient should make the decision as to whether treatment serves her interests, values, or purposes and that the physician acts so as to promote that goal. I do not want to minimize the degree to which the physician can and should attempt to guide the patient's decision-making or to overestimate the (informational) competence of the patient to make such choices. Indeed, precisely for that reason, the physician has a fiduciary obligation not to recommend or perform treatments and procedures that do not promise to serve the patient's interests. Nonetheless, among those recommendations that can be reasonably regarded as serving the patient's interests, the (normally competent) patient has the authority to decide whether and what treatment to undergo. Once the patient gives her informed consent to a particular proceclosure and that it did not disadvantage their clients. At least one jurisdiction also requires that fiduciaries prove that the client could deal with the fiduciary at arm's length in regard to the matter from which the fiduciary receives benefit. Thus, even if a client understands the transaction, the court will carefully scrutinize it in order to learn how the client's participation was induced." Linda Jorgenson and Rebecca M. Randies, "Time Out: The Statute of Limitations and Fiduciary Theory in Psychotherapist Sexual Misconduct Cases," 44 Oklahoma Law Review 181 (1991), at 202. 73 It is frequently said that the increasing importance of health maintenance organizations, in which patients pay a fixed fee for health care, may be threatening this assumption.

184

CHAPTER SIX

dure, the relationship often (although perhaps not always) moves into a fiduciary mode with respect to the execution of the treatment. At this point in the process, the patient's consent is not sought and would not legitimize a deviation from accepted practice if it were sought and obtained.74 The physician must act solely in the interests of the patient because the patient is unable to evaluate or promote her own interests. If we assume that the relationship between the psychotherapist and patient is at the third stage, then the patient's consent cannot legitimate sexual relations because such relations violate the psychotherapist's fiduciary responsibility to serve the interests of the patient. But if this is so, it does not solve the problem that animated our discussion, to wit, whether the fiduciary argument implicitly denies the patient's capacity to act as an autonomous individual who is capable of giving genuine consent to sexual relations with her psychotherapist. It has been argued that the fiduciary model is attractive precisely because it simultaneously condemns psychotherapist-patient sexual relations and affirms the patient's autonomy and capacity to consent.75 But can we have our fiduciary cake and patient autonomy, too? Not in a straightforward way, or so I shall suggest. The argument for applying the fiduciary model to psychotherapy typically notes that the very structure of the treatment process encourages and even requires the patient to relax her ordinary psychological defenses, to expose herself in ways that normal social and psychological modesty would preclude.76 In response, the psychotherapist has a duty not to take advantage of the patient's vulnerabilities, to refuse to engage in any activity that is not reasonably regarded as promoting the patient's interests and which the patient would (hypothetically) reject were she in a position to protect her interests.77 But this argument does not 74 The patient may, of course, decide to withdraw consent to the procedure, a decision that the physician may or may not be required to respect. I take it that a physician should ignore the withdrawal of consent if the cessation or interruption of a procedure poses a danger to the patient. 75 "Mental health professionals have a fiduciary relationship with their patients in which deep confidences are shared and close ties may develop. Their professional status also creates a power asymmetry. A professional who sexually exploits a patient is taking unfair advantage of a position of power and trust. This model does not imply that the patient is impaired or incapable of consenting; competent, autonomous decision-makers can be unduly influenced." T. G. Gutheil, "Patient-Therapist Sexual Relations," 6 Harvard Medical School Mental Health Letter 5 (1989), at 5-6, quoted in

Jerry Edelwich and Archie Brodsky, Sexual Dilemmas for the Helping Professional (New York: Brunner/Mazel, 1991), p. 214. 76 Castel, "Women's Sexual Exploitation," p. 45. 77 Ellen Plaisil makes the following remark in light of her experience with Dr. Leonard, t h e "objectivist" psychotherapist: "Informed consent in a case such as this

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

185

assert that a patient's genuine consent to sexual relations with her psychotherapist should be treated as irrelevant, even though she is competent to consent. To the contrary. It maintains that we adopt a fiduciary model precisely because there are good reasons to regard apparent consent as insufficiently competent or voluntary or informed under the circumstances. And so the question remains as to whether we can insist that the psychotherapist refrain from sexual relations and simultaneously fully respect the patient's capacity to choose to engage in sexual relations. We can say that the therapeutic relationship exists to serve the patient's psychotherapeutic needs rather than her sexual desires, but the question would remain, Why can't the patient change that if she wants?78 We can understand the fiduciary relation as a Ulysses-type contract.79 Consider a standard example. In advance of a party, B gives A the keys to B's car and instructs A not to let B drive home should A believe that B has consumed too much alcohol. By mutual consent, A acquires a fiduciary obligation not to accede to B's subsequent request for the car keys if B should consume too much alcohol. Precisely because B knows that he may change his mind, ex post, B decides, ex ante, to weaken the force of his subsequent decisions. If A should give B the keys in response to a subsequent request, we might well hold A morally (if not legally) responsible for what happens to B. Similarly, we might say that the patient cannot change the terms of that relationship because, as a competent and autonomous person, she gives her consent to enter into a fiduciary relationship in which she anticipates that she will becomes less autonomous and less competent within the confines of that relationship. B consents to a relationship in which the force of her future consent is rendered irrelevant or subject to exceptionally high standards of validity. Although B enters into a fiduciary relationship with A because B believes that the relationship will promote B's interests, B understands that the psychotherapeutic process may render her unable to monitor certain aspects of the relawould mean permission given either prior to or immediately following the commencement of therapy. It is the informed part of informed consent that requires it to be given so early in the psychotherapeutic process. Consent given at any later time risks the possibility, the probability, of the therapist influencing the patient's judgment and ultimate decision on the matter." Therapist, p. 186. 78 Several statutes specifically preclude patient consent as a defense. Supra note 13, at 729. 79 "A Ulysses-type contract" refers to the story of Ulysses asking his sailors to tie him to the mast so that he can hear the call of the sirens (unlike the sailors, who were told to put wax in their ears) but not be lured away. See Jon Elster, Ulysses and the Sirens (New York: Cambridge University Press, 1979).

186

CHAPTER SIX

tionship. For that reason, B decides ex ante to allocate certain responsibilities to A that B's subsequent consent will be relatively (although perhaps not completely) powerless to change. To say that fiduciary relationships can be understood along the lines of a Ulysses-type contract does not, of course, show that the patient actually consents to giving up her ability to change the terms of that relationship, although I think that she does consent to something close to this.80 How many would-be patients would refuse to enter a regime in which they are forbidden to respond to another person's desires and vice versa? I do not know. I think that many patients would agree to such a regime, if they understood that the revelation of such feelings or the prospect of such revelations would undermine the prospects for successful therapy.81 Some patients might not agree to such a regime because they do not understand the power of transference. But that would not be an argument for respecting their desire to change their mind. Rather, it might be an argument for a paternalistic policy in which we do not allow patients to enter into agreements that they cannot fully understand. Here we shift from an "actual" Ulysses-type contract to a "hypothetical" Ulyssestype contract. We can consider restrictions on the funeral industry along the lines of a hypothetical Ulysses-type contract. When morticians are condemned for getting grieving families to spend large sums on funerals and caskets, the industry responds, "How can it be our fault? It's up to the family to decide how much they want to spend/' 82 The reply to this argument is obvious and no less sound for that, namely that the decision-making ability of funeral purchasers is often impaired by grief, guilt, and time pressure.83 For these reasons, it may be desirable to offer greater protection to funeral purchasers than to buyers of other goods, say, by banning undertakers from engaging in "upselling" (getting customers to buy more expensive services) tactics. It is not that any of us have actually agreed to a regime in which we are 80

After all, when the patient enters psychotherapy, the "no sexual relationship" principle is often explicitly discussed, and, even if not, it is a crucial element in the "package" of professional service that the patient is buying—a package that consists of the psychotherapist's individual expertise and the professional norms that govern the delivery of his services. I thank Arthur Applbaum for pressing me on this point. 81 This analysis of consent will not work for those cases in which the patient is in crisis before therapy and therapy stabilizes the situation, thus making it possible for the patient to give more genuine consent during therapy than prior to therapy. 82

See Jessica Mitford, The American Way of Death ( N e w York: Simon a n d Schuster, 1978), p. 4 1 . 83 See Funeral I n d u s t r y Practices, Final Staff Report to t h e Federal Trade C o m m i s s i o n a n d P r o p o s e d Trade Regulation Rule, Bureau of C o n s u m e r Protection, June 1978, p. 1.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

187

protected from the risk of impaired decision-making in buying funeral services, but that we understand the need for such protection and that it does not disparage our self-understanding as competent and autonomous persons to think that we might need such protection under the circumstances. Reconstructing fiduciary relations along the lines of a Ulysses-type contract does seem to imply that the patient's capacity to give informed consent to sexual relations during psychotherapy is diminished.84 Perhaps the fiduciary model "does not imply that the patient is impaired or incapable of consenting" qua person, but it may imply that the patient's capacity to consent is diminished qua patient within the therapeutic process.85 I see no reason to resist this implication. There is some tension between our concern with protecting the patient from the harmful exploitation of her vulnerability and our commitment to respecting autonomy and competence. And the answer is to be found in a more complex and diachronic approach to autonomy, not the mere assertion that the tension can be easily resolved within a synchronic framework. Suppose that we reach the opposite view. Suppose that we believe that some patients can give genuine consent to sexual relations with their psychotherapists. Still, it would not follow that the psychotherapist does nothing wrong in encouraging or acquiescing in a sexual relationship with his patient. We are not "passive vendors" of services agreed to or solicited by others.86 We might say that the psychotherapist acts wrongly because his behavior is self-serving and not therapeutic. And even if the psychotherapist would not be acting wrongly toward a patient if the patient had actually consented, we might think that the psychotherapist acts wrongly because he cannot be sufficiently certain that the patient has consented. Moreover, we could say that the psychotherapist acts wrongly because his actions undermine confidence in the profession.87 After all, even if the psychotherapist does not wrong his patient if she gives meaningful, informed consent to sexual relations, given that it is difficult for others 84

In his testimony on behalf of Julie Roy, Dr. Willard Gaylin put it this way: "When somebody is ill, he places exceptionally large trust in his doctor—so much so that, however we are talking about money or sex or anything else, a doctor has a responsibility not to take advantage of a frightened and helpless patient. This is true of all medicine, not just psychiatry." Freeman and Roy, Betrayal, p. 160. 85 T. Gutheil, quoted in Edelwich and Brodsky, Sexual Dilemmas, p. 214. 86 B u c h a n a n a n d Brock, Deciding for Others, p . 143. 87 Buchanan and Brock make a similar point with respect to competence: "[A]lso important to the appropriate standard of competence is the value of maintaining public confidence in the integrity of the medical profession, so as to protect and foster the trust necessary to physician-patient relationships that function well." Ibid., p. 48.

188

CHAPTER SIX

to assess the quality of a patient's consent, A may have a responsibility to them not to undermine their trust in him or the profession. Thus it is possible to respect a patient's capacity to consent to sexual relations with her psychotherapist and maintain that it is wrong for the psychotherapist to engage in such relations. But it is less clear that we can respect a patient's capacity to consent to sexual relations with her psychotherapist and maintain that the psychotherapist has committed a wrong about which she can rightfully complain that she has been exploited. Moreover, we have seen that there may be important distinctions between the moral force of consensual exploitation and nonconsensual exploitation. It will be more difficult to justify prohibiting relationships if they are consensual. So we cannot avoid trying to answer the questions of whether and under what conditions patients can give genuine informed consent to sexual relations with their psychotherapists. STANDARDS OF CONSENT

I noted above what might appear to be inconsistencies in our judgments about consent. I noted that we are apt to regard casual adolescent sexual relations as consensual, whereas we are inclined to regard sexual relations between, say, a thirty-year-old male and an adolescent female as (somewhat) nonconsensual and even as statutory rape. I noted that we are generally inclined to think that psychotherapy patients are capable of giving or refusing consent to certain modes of treatment (e.g., electroshock, psychotropic drugs, involuntary commitment), but that psychotherapy patients cannot give genuine consent to sexual relations. Are these views consistent?88 Let us say that B gives valid consent to A's action when B's consent is sufficient to legitimate A's action, that is, when it has the right sort of normative force. On what I shall call the quality of will view, the validity of B's consent is a function of empirical facts about B's psychology.89 On the quality of will view, the judgments noted above would be consistent only if the quality of will that typifies the adolescent's consent to sexual relations with an adult is, in fact, lower than the quality of will that typifies the adolescent's consent to sexual relations with another adolescent and the quality of will that typifies a patient's consent to sexual relations with her psy88

I am not here asking whether these views are correct, for that would require more information about individual cases. I am asking whether, at a general level, they are consistent and what would render them consistent. 89 I borrow this p h r a s e from T. M . Scanlon, "The Significance of Choice," in Tanner Lectures on Human Values, Vol. 8 (Salt Lake City: University of Utah Press, 1989).

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

189

chotherapist is lower than the quality of will that typifies the patient's consent to (nonsexual) treatments. On a second view, the validity of B's consent is not a straightforward function of B's quality of will. Rather, the standards for valid consent are normative and variable. They are higher in some cases than in others. Call this the variable standard view. On the variable standard view, the validity of consent does not turn on whether B's quality of will passes some fixed standard, but whether it passes the standard appropriate for that case. I believe that a defense of the variable standard view must be something like this. It is a mistake to think of the principle of consent as resting on a freestanding concept with empirical referents, which, when ascertained, can generate the appropriate moral upshot, for example, to bar interference with B's decision. Rather, when we say that B gives informed consent to A's action, we point to a certain range of phenomena—what counts as signifying consent, certain sorts of mental states, and the absence of certain sorts of compromising conditions—that serve to support certain moral upshots. Although we may have come to express the issue in terms of whether B has given consent, and that seems like an empirical question, the central question is irreducibly normative: what sorts of phenomena are sufficient to generate the relevant moral result? And there is no reason to think that an empirical investigation will answer this question. The normative construction of a central concept in a moral argument is not idiosyncratic to the present enterprise. Consider an issue that arises in applying the (Millian) principle that the state can justifiably prohibit only conduct that causes harm to others: What constitutes harm? Should the psychic distress caused by offensive behavior or speech count as harm? Does mere trespass constitute harm? Clearly, there is a sense in which psychic distress at offensive behavior is a form of harm. It may cause no less pain or loss than many actions that are paradigmatically harmful. Just as clearly, there is a sense in which one is not harmed by mere trespass, particularly if one is not aware that the trespass has occurred. We could construct an empirical account of harm, say, one that turned on neurological measures of pain. But such an account would be of little help. As Jeremy Waldron has put it, "[T]he question is . . . not what 'harm' really means . . . but which conception answers more adequately to the purposes for which the concept is deployed."90 And on a normative view of the harm principle, one that defined harm (in part) by refer90

Liberal Rights (New York: Cambridge University Press, 1993), pp. 119-20. For a somewhat different view, see Frederick Schauer, "The Phenomenology of Speech and Harm," 103 Ethics 635 (1993).

190

CHAPTER SIX

ence to violations of the rights of the other party, it is perfectly coherent to maintain that the psychic distress caused by offensive behavior does not count as a harm because offense does not violate rights, whereas trespassing on another's property does constitute a harm because it does violate a right, even though the former is more painful or hurtful than the latter. To take an issue somewhat closer to our present concerns, Dan Brock and Allen Buchanan have defended a similar view of "competence/' On their view, to assess a patient's competence is to evaluate the patient's decision-making capacity to make a particular decision. Although the "underlying capacities and skills forming the basis of that judgment are possessed in different degrees," the judgment that a patient is competent is a normative and binary (or "threshold") judgment—the patient does or does not have the capacities to be assigned the authority to make a particular decision.91 And, they maintain, that judgment should be made in light of the appropriate balance between the values of autonomy and welfare, between the values of respecting a patient's right to decide for herself and protecting her from the harmful consequences of her own choices. Given this approach, they maintain that "no single standard of competence . . . can be adequate for all decisions."92 On the variable standard view of consent, B's quality of will is analogous to the underlying decision-making capacities noted above and the validity of consent is analogous to the decision that the patient is competent. Although the concept of informed consent may appear to refer to the agent's quality of will, because the principle of informed consent is meant to resolve a set of normative problems, the best understanding of that concept cannot be made independent of the normative tasks that present themselves. In the final analysis, we always have to reach a judgment as to whether the patient's consent legitimates the action in question, and it is unlikely that this judgment will be a straightforward function of the agent's quality of will. It will require the use of moral criteria. How should the moral criteria for valid consent in a particular context be formulated? Borrowing from T. M. Scanlon's version of contractarianism, we might say that the proper criteria for valid consent are those criteria that "no one, suitably motivated, could reasonably reject as a basis for informed, unforced general agreement."931 do not 91

Buchanan a n d Brock, Deciding For Others, p. 27. Ibid., at 51. I am not here endorsing the way in which Brock and Buchanan have balanced these criteria. For an approach that places less emphasis on the patient's welfare, see Ronald Dworkin, Life's Dominion (New York: Knopf, 1993), esp. ch. 7. 93 Scanlon, "The Significance of Choice," p . 184. 92

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

191

want to underestimate the difficulties in the way of determining whether a person could reasonably reject a certain principle. But if the parties understand that the purpose of these criteria is to allocate responsibility for actions taken under a variety of circumstances, there is no reason to think that they would insist upon a single standard. And it is entirely possible that no one could reasonably reject a principle of informed consent that allows a physician to perform a medically indicated amputation if a normally competent and informed adult agrees to the procedure, however constrained the choice, but does not allow a psychotherapist to engage in sexual relations with his patient even if a normally competent and informed patient agrees to those relations, when those relations are likely to be harmful to the patient, and when the patient may be in the grip of transference.94 And it is possible that no one could reasonably reject such criteria even if the underlying quality of will of the psychotherapy patient is not substantially different from the quality of will of the gangrenous patient. Although there is much more that would have to be said to flesh out this line of argument, here I want to note a danger with the variable standard or normative view of consent, namely, that it threatens to detach the assessment of consent from the agent's quality of will. In effect, it maintains that B gives valid consent whenever B's apparent consent legitimates A's action or has the relevant moral upshot. Consider the cognitive dimension of informed consent. Suppose that a physician carefully reviews the treatment alternatives with a woman diagnosed with breast cancer, say, the choice between a lumpectomy and mastectomy. The patient pretends to be listening to the physician's explanation. She then asks the physician, "What would you do?" and signs a consent form authorizing the recommended treatment. Has the patient given "informed consent"? If informed consent requires that the patient understand the appropriate information, then the answer must be no. Yet for legal or moral purposes, we might say that the patient has given valid consent, or perhaps more accurately, that B's apparently valid consent makes it legitimate for the physician to perform the recommended treatment.95 94 "[F]rom the point of view of an agent looking at his own actions over time, situations of choice have to be evaluated not only for what they make 'available' but for what they make it likely that one will choose. It is not unreasonable to want to have some protection against the consequences of one's own mistakes." Ibid., p. 195. 95 It could be that the doctor has done her part but the patient has still not been treated with adequate respect—getting someone who is sick and scared to understand certain pieces of information may take more work than it is reasonable to ask of any one health professional. Perhaps the hospital should provide a patient advocate in some

192

CHAPTER SIX

Consider the volitional dimension of informed consent. Suppose that the gangrenous patient signs a consent form authorizing the amputation of his leg because the only alternative is death. Has he given voluntary consent to the amputation? If voluntary consent requires a substantial degree of "decisional space/' then the answer must be no. Yet for legal and moral purposes, we may want to say that the patient's signature grants a moral permission for the physician to go ahead with the surgery. What should we say about consent in these two cases? Using the variable standard view, we could say that B has given valid informed consent, applying a relatively weak standard of consent. On ScanIon's preferred view, however, what he calls the background view, the moral force of a person's choice depends, in part, upon the value that an agent would reasonably attach to having outcomes depend "on his or her subjective responses under certain conditions."96 But it also depends upon the relevant background conditions, on whether others have "done enough" to provide the agent with reasonable choosing conditions and alternatives. When those background conditions are in place, "the fact that a person chooses a certain outcome may make that outcome one that he or she cannot reasonably complain of"—whatever the quality of will associated with that choice.97 On the background view, a deliberate choice is neither sufficient nor necessary to legitimate an outcome. It is not sufficient because the fact that a person chooses a certain outcome may not deprive the agent of a reasonable complaint if she has not been provided with the relevant background conditions. It is not necessary because if others have done enough (by some standard) to provide the relevant background conditions, then the agent may be responsible for what happens to him, whether or not he makes an informed and deliberate choice. To illustrate this point, Scanlon asks us to consider a case in which a community must move highly toxic substances through a neighborhood and takes great pains to warn citizens to avoid being outside during that period. If a person simply forgets about the danger and exercises in his yard when the trucks go by, he "is responsible for what happens to him and 'cannot complain of it' even though he made no conscious decision to take the risk."98 cases. But whatever efforts are made to enable patients to understand their options, there will be cases in which we believe that enough has been done on the patient's behalf and yet the patient does not understand the basis of her consent. 96

97 98

Scanlon, " T h e Significance of C h o i c e , " p . 184. Ibid., p . 190. Ibid., p . 196.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

193

Let us try to apply the background view to some of our examples. Suppose that the physician has done "enough" (whatever that is) to explain the alternative treatments to the breast cancer patient. Rather than say that she has given informed consent, as under the variable standard view, we would make a more complicated statement on the background view, to wit: B has not given informed consent, but we will not excuse B from what would otherwise be the force of B's consent—to legitimate the transaction or procedure—because there are good reasons to allocate the burden for this informational defect to B rather than A. In effect, we are claiming that even though the quality of B's consent may not be up to snuff, B does not have the basis for a subsequent complaint. Or we could say that the fact that B does not understand the treatment negates the fact of consent but does not negate the moral force of consent. Similarly, since there is nothing that the physician can do to create more "decisional space'' for the gangrenous patient, the gangrenous patient's consent is sufficient to authorize the physician to proceed with surgery. Consider now the case of the gangrenous patient's consent to surgery and the psychotherapy patient's consent to sexual relations in light of the distinction between the variable standard view and the background view. The variable standard view requires us to say that the gangrenous patient has given sufficiently voluntary and informed consent to the surgery, but that the psychotherapy patient has not given informed consent (of sufficient quality) to sexual relations with her psychotherapist. The background view allows us to say something like this: Perhaps the psychotherapy patient has given consent of no less quality than is sufficient to give a surgeon permission to perform an amputation on a gangrenous leg, but there are good reasons not to allow such consent to legitimate sexual relations whereas there are good reasons to allow it to legitimate an amputation. Which is the better view? The advantage of the variable standard view is that it preserves the linguistic link between informed consent and the relevant moral upshot. The disadvantage of the variable standard view is that it is arguably at variance with any reasonable account as to what constitutes informed consent. The adjustment in language seems ad hoc. The advantage of the background view is that it provides a plausible account of the moral situation. The disadvantage, or shall I say apparent disadvantage, is that it weakens the link

194

CHAPTER SIX

between consent and certain moral upshots. In effect, it converts questions of consent into questions of justice about the allocation of responsibility. As a practical matter, it may not make much difference which view we choose. The two views are likely to prove extensionally equivalent with respect to the moral force of consent. As a philosophical matter, I am inclined to think that the background view is closer to the truth. The background view has an additional advantage. By linking the issue of consent with an issue of justice—to whom should the responsibility for a decision be allocated?—we can better see the relationship between consent and exploitation, for to say that A has (or has not) taken unfair advantage of B in securing B's agreement is not simply to say that the quality of B's will does not (or does) pass muster. It is to evaluate A's and B's behavior within the context of the decision at issue. When a physician acts properly in asking the patient to agree to a beneficial amputation, then it is appropriate to assign the responsibility for that decision to the patient even though the quality of the patient's consent is arguably compromised by the lack of alternatives. On the other hand, when a psychotherapist acts improperly by engaging in behavior that is likely to be harmful to the patient and where the patient's decision-making ability is compromised by the therapeutic process, then it is appropriate to assign the burden of that harmful decision to the psychotherapist. The quality of will that underlies an individual's consent will play a significant role in our moral judgments but it is not dispositive. LEGAL AND PROFESSIONAL RESTRICTIONS

Psychotherapists who engage in sexual relations with their patients are rightly regarded as exploitative because (1) the psychotherapists gain from the relationship; (2) the relationship is typically harmful to the patient; and (3) the patient's consent is insufficiently informed and voluntary to render such relations legitimate and might not render them legitimate even if it were sufficiently informed and voluntary. What should be done about such exploitation? In this section I consider the arguments for legal and professional restrictions on psychotherapist-patient sexual relations during the pretermination and posttermination periods. My primary concern is not to defend particular policies, but to examine the arguments that have been and might be advanced in their defense and to consider whether patients can claim that such restrictions are inconsistent with respect for their autonomy.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

195

Pretermination

I see no reason to doubt the standard view that "It's never okay" is the best rule." Further questions may be raised as to the most effective legal and institutional mechanisms for enforcing that principle, including statutes that specifically criminalize psychotherapist-patient sexual relations; prosecution of offending psychotherapists under sexual assault or rape statutes; common law actions for battery, malpractice, and infliction of emotional distress; statutes that create a specific civil cause of action for sexual exploitation; professional rules, regulations, and penalties; treatment of offending psychotherapists; and professional education and training.100 It should be said, however, that even if "It's never okay" is the best rule, the justification of an absolute bar to sexual relations between psychotherapists and patients need not assume that all such relations are malintentioned, nonconsensual, harmful, or exploitative. It need only assume that psychotherapist-patient sexual relations are typically harmful or nonconsensual and that psychotherapists cannot make publicly reliable judgments as to when they are consensual and not harmful. I say publicly reliable because even if some psychothera99

This phrase is used in several publications issued by a task force appointed by the Minnesota legislature. Jorgenson, Randies, and Strasburger, "Furor/' p. 647. 100 As evidence, if any were necessary, that few things in the law are as simple as they seem, let me briefly note two issues that have arisen with respect to legal strategies. First, there is some controversy as to whether sexual exploitation should be regarded as a form of malpractice. Some psychotherapists (or their insurers) have maintained that such relations occurred outside of the therapeutic context, even if they occurred in the psychotherapist's office, and hence do not constitute malpractice. This issue appears to have been crucial to the suit against Carolyn Bates's "Dr. X." He maintained that his relations with Ms. Bates were in addition to and distinct from therapy, that he had a personal relationship with Ms. Bates apart from their therapeutic relationship, and that within the context of the personal relationship, she was free of her "patient identity and [was] able to consent to the intercourse." See Bates and Brodsky, Sex in the Therapy Hour, p. 53. It has also been argued that it is contrary to public policy to insure against intentional or criminal acts, that such insurance may give rise to a "moral hazard" problem. To the extent that psychotherapists are insured against the harms caused by sexual relations, they have less incentive to avoid acts that might impose such harms. Another problem concerns the appropriate "statute of limitations" that should apply to allegations of sexual exploitation. Unlike most crimes, where victims immediately understand that they have been harmed, a patient may not fully understand that she has been wronged by her psychotherapist until some time after the injury occurred. The offending psychotherapist may (successfully) convince the patient that she has not been injured or that she has consented, and patients may frequently go through repression and denial about the events that occurred. For these and other reasons, it has been argued that the standard statutes of limitations should not apply to such cases.

196

CHAPTER SIX

pists can make reliable judgments as to when such relations are consensual and not harmful, and this is doubtful, outside observers may have no feasible method of distinguishing those psychotherapists from the true exploiters, and their behavior may jeopardize the public's confidence in the profession. Thus, given the high probability and severity of harm resulting from psychotherapist-patient sexual relations, it is perfectly reasonable to adopt a rule that is more inclusive than the underlying phenomena require. Even if an absolute ban would preclude some psychotherapist-patient sexual relations that would be a beneficial part of the treatment process or would preclude a personal relation that would be satisfying to both (apart from the treatment process), an absolute prohibition may be defensible nonetheless. That some psychotherapists have had successful marriages with patients with whom they commenced sexual relations during therapy would not show, by itself, that an absolute ban is unjustifiable any more than the fact that some automobile passengers survive because they do not wear seat belts tells against a policy of always wearing seat belts. Interestingly, although feminists have traditionally been in the vanguard of the fight for stringent restrictions on psychotherapistpatient sexual relations, one might advance a feminist critique of the prohibitory strategy. It might be argued that the prohibitory strategy implicitly accepts and reinforces a picture of powerful and dominating males exploiting weak and compliant females, thereby perpetuating the general pattern of gender inequality across the entire society, whatever its effect on psychotherapy patients. It might also be argued that prohibitory strategies are counterproductive within psychotherapy itself, that they are less efficacious in preventing sexual exploitation than policies that would empower women to achieve greater equality and control over their lives.101 If the prohibitory strategy were bad for women as a class and for psychotherapy patients, there would be good reason to consider rejecting prohibitory policies. If it were good for one group and bad for the other, then we would have to balance the interests of female psychotherapy patients (who might be benefited by the prohibitory strategy) and the interests of women as group (who might be harmed by the prohibitory 101

"It is not in women's interests, either theoretically or practically, to concede to institutions the right to delimit the formation of personal ties among community members. Repressive policies engender more, not less repression." Frances Hoffman, "Sexual Harassment in Academia: Feminist Theory and Institutional Practice," 56 Harvard Educational Review 105-121 (1986), at 115. This argument is made in the context of sexual relations in academia. It is not clear that the author would want to extend the argument to the present context.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

197

strategy), or vice versa. In any case, there is little evidence that either claim is correct. Despite the plausible ring to statements about the reinforcement of gender inequalities, I know of no evidence that prohibitory policies have this causal effect. Indeed it is equally plausible to believe that by "taking a stand" against the exploitation of women in psychotherapy, society is reinforcing the principle of gender equality. Whereas the feminist critique of the prohibitory strategy just discussed is fundamentally consequentialist, a different critique maintains that prohibitory strategies are unjustifiable because they deny the "right of individuals to establish relationships when, with whom, and where they choose/' 102 Because the autonomy-based argument has been made more frequently in the context of prohibitions on posttermination sexual relations, I shall consider it below. Suffice it to say that if the argument is not conclusive with respect to the posttermination period, as I think it is not, it would certainly not be conclusive with respect to the pretermination period. Posttermination

The principle that psychotherapists should never engage in sexual relations with their patients during therapy is now well accepted, although there are serious problems of compliance with the principle.103 There is less unanimity with respect to the best principle for the posttermination period. Some argue that the pretermination ban should never be lifted—"once a patient, always a patient." Others argue that a determinate waiting period (say, one year) strikes the appropriate balance between protecting patients from exploitation and respecting individual choice.104 Others argue that decisions should be made on a case-by-case basis, and that where there is no harm, there should be "no foul."105 The primary argument against a permanent ban on posttermination sexual relations is that the factors that justify a pretermination ban are simply not present to the same degree in the posttermination period. If the psychotherapeutic process has been relatively successful, and if both parties can relate to each other as competent adults, then there seems to be no reason to prohibit the parties from entering 102

Ibid., at 114. 103 Psychotherapist-patient sexual relations may be permitted or even encouraged by some "nontraditional" approaches, but all mainstream professional associations maintain that "it's never okay" during therapy. 104 Appelbaum and Jorgenson, "Psychotherapist-Patient," p. 1471. 105 Coleman, "Sex between."

198

CHAPTER SIX

into a consensual relationship.106 A second argument against a permanent ban or long-term prohibition maintains that such a ban effectively denies the effectiveness of therapy itself. It demonstrates a form of professional "self-doubt/'107 A third argument maintains that a permanent ban is paternalistic because it denies the competence of the former patient to give her "informed consent" to enter a relationship just because she is a former patient and it is sexist because the rule seems primarily directed at women, and presupposes that they will always remain "under the influence of their male psychiatrists/'108 In the face of these objections, there are at least three important lines of argument for a posttermination ban. First, it may be argued that posttermination sexual relations carry at least some (if not all) of the risks of pretermination relationships. If this entails denying the complete effectiveness of the therapeutic process and the resolution of transference, so be it. A second argument maintains that there is something fundamentally improper about the posttermination relationship itself, that the pretermination history of the relationship casts a moral shadow that darkens its character. On this view, the relationship will be compromised by the imbalance of power, asymmetry of information, and transference that characterized the therapeutic relationship—"the patient stays the patient/' 109 I find the first line of argument more plausible than the second. It seems eminently legitimate for society to consider the harmful effects of posttermination sexual relations on former patients. But it is less clear that society has a sufficiently important interest in the moral quality of the posttermination sexual relationship to justify restrictions on that ground alone. It is one thing to believe that a relationship with a certain sort of history or characteristics ("May-December" romances, say) is nonideal and quite another to argue that it is permissible for society to prohibit such relationships. This is, after all, an 106 If the same potential for exploitation and harm that exists when a psychiatrist engages in sexual activity with a current patient does not exist when he engages in a similar relationship with a former patient, then an absolute prohibition against sexual activity with former patients would be unnecessary. Ibid., p. 2. 107 Ibid., p . 4. 108 Ibid., p . 4 8 . 109 So says Rina Folman, quoted in supra, n. 51, at 64. "To rationalize the acceptability of sex with patients after termination, one would need to demonstrate that transference no longer exists after termination of treatment." Glenn Gabbard and Kenneth S. Pope, "Sexual Intimacies After Termination: Clinical, Ethical, and Legal Aspects," in Gabbard Sexual Exploitation in Professional Relationships, at 116.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

199

arena in which we go a very long way toward accommodating individual choice.110 A third line of argument for a posttermination ban has a different structure. What I shall call the prophylactic argument focuses on the pretermination effects of a posttermination ban. The prophylactic argument does not deny that psychotherapists and former patients are capable of entering into genuinely voluntary posttermination relationships. Rather, it maintains that the very possibility of a posttermination relationship tends to compromise the (pretermination) therapeutic relationship. If the patient believes that a posttermination sexual relationship is a genuine possibility, the patient may be reluctant to discuss material about which she is ashamed and looking attractive and pleasing the therapist become more important.111 In addition, the resolution of transference is made more difficult if the patient has any doubt as to whether a relationship could really occur. And thus a ban on posttermination relationships will protect the therapeutic relationship from a potentially powerful contaminant. The prophylactic argument applies to the psychotherapist as well. If the psychotherapist believes there is a chance that at least some of his patients will come to be his lovers in the posttermination stage, this may encourage him to see his practice as a source of sexual partners rather than as a place of work.112 The prophylactic argument does not imply that psychotherapists would otherwise adopt an aggressively aggrandizing attitude towards their patients. It only implies that psychotherapy is a difficult process and that perfectly decent and well-meaning psychotherapists may need all the help they can get in maintaining appropriate boundaries.113 110 As Appelbaum and Jorgenson note, post-termination relationships "may involve no more problems than many relationships routinely sanctioned in our society." Supra, n. 65, at 1472. 111 "You can't do therapy when you're withholding information or keeping secrets to make a good impression because you might want to date later. . . . Symptoms may miraculously vanish when consummation awaits the end of therapy." Sherman, "Behind Closed Doors," p. 64, quoting Glen Gabbard. 112 "The expectation of a possible love affair after termination is bound to lead to self-aggrandizing behavior . . . during the course of therapy. Such limit-testing is expected of the client in any case, but it is potentially tragic when the therapist's purposes are confused by personal longings and impulses." Edelwich and Brodsky, Sexual Dilemmas, p. 99. Dennis Thompson pointed out that it is possible that psychotherapists are less likely to have sexual relations with patients during therapy if they believe that they could have such relationships when therapy is completed. If this were true, then it might be best to allow posttermination sexual relationships. in Even if the prophylactic argument is sound, one might raise a question about its moral force during the posttermination period. Suppose that a psychotherapist and

200

CHAPTER SIX

We have seen that there are arguments for and against a permanent ban on posttermination sexual relations. Perhaps an intermediate position, such as a determinate and limited ban, is to be preferred. Paul Appelbaum and Linda Jorgenson argue that a permanent ban on posttermination sexual relations is excessive because virtually all of the factors that justify an absolute proscription during treatment are "modified in the post-termination context/'114 The psychotherapist's capacity to coerce the patient into sexual relations is weaker because "the implicit threat that treatment will be withheld if the patient refuses sex is now defused."115 The psychotherapist's claim that sexual relations will not interfere with therapy seems not to apply. Moreover, at least some of the psychotherapist's fiduciary obligations to the patient end when therapy ends. 116 Finally, and with respect to the prophylactic argument, they maintain that patients are unlikely to distort their "presentations to make themselves appear more attractive as future partners" if they must wait one year to begin a relationship, and psychotherapists will not be tempted to prematurely terminate treatment if they have to wait one year before sexual relations can begin. So should there be a permanent or determinate ban on posttermination sexual relations? I am inclined to think that the character of the best policy will depend upon a set of empirical considerations about which we may not (and certainly I do not) have adequate knowledge. We can say this. Although the reasons that justify a permanent ban on pretermination sexual relations are no doubt stronger than the reasons for a permanent ban on posttermination sexual relations, it does not follow that "an absolute prohibition against sexual activity with former patients would be unnecessary."117 The case for a posttermination ban might still be strong enough. patient go through a successful psychotherapeutic process on the assumption that a posttermination sexual relationship between them is not possible and that the success of the therapy is due, in part, to the fact that they made this assumption. But once the transference and countertransference have been resolved, do they have reason to refrain from a relationship—apart from its indirect effect on others? This is, of course, analogous to a standard problem for rule-consequentialist arguments. Note that I am not asking whether it would be right to punish the psychotherapist even if his behavior was not wrong (under the prophylactic argument), but whether he has any independent moral reason to refrain from a relationship under the rubric of the prophylactic argument. 114 Applebaum and Jorgenson, "Psychotherapist-Patient," p. 1470. 115 Ibid. 116 For example, while psychotherapists must continue to honor the principle of confidentiality with respect to their former patients, they are not under an obligation to resume therapy with them. Ibid. 117 Coleman, "Sex between," p. 2.

SEXUAL EXPLOITATION IN PSYCHOTHERAPY

201

It might be objected that even if a posttermination ban produced better consequences for most patients than did other policies, society would not be justified in prohibiting relationships between consenting adults who do not have a professional relationship. Just as we reject miscegenation laws because we think the right to have an intimate relation with the person of one's choice is among our most basic rights, we should reject a posttermination ban on relationships between psychotherapists and patients for similar reasons. Perhaps the tenuousness of the patient's consent to intimate relations during therapy is sufficient to justify abridging that right during the pretermination period, but it is not sufficient to abridge that right after therapy is terminated. There are, however, at least two differences between a ban on miscegenation and a determinate or indefinite ban on intimate relations between psychotherapists and former patients. First, the restriction on posttermination relations is not based on a classification whose aim is to stigmatize. If the ban were justified by appeal to the capacity of the posttermination patient to make important choices in her life or to concerns about the moral quality of the relationship, then it might raise such suspicions. But to the extent that the justification appeals to the effects of such relationships on the therapeutic relationship itself, it does not unfairly stigmatize either the psychotherapist or the former patient. Second, there is an obvious distinction between the cases with respect to prior consent. A ban on miscegenation is not consented to, only obeyed or disobeyed. By contrast, a restriction on posttermination relationships would originate in the voluntary act of both parties to enter a psychotherapeutic relationship with such a restriction, at least if the restriction were announced in advance. Still, it might be objected that the right to have an intimate relation with another person should be regarded as inalienable. Even if the waiver of that right originates in the voluntary decision to enter psychotherapy, it is wrong to condition psychotherapy on the waiver of that right. To use an (overused) Millian example, it may be argued that just as one should not be free to waive one's right not to be a slave in exchange for a monetary benefit, one should not be permitted to irrevocably waive one's right to enter into an intimate relation in exchange for psychotherapeutic benefits. But even if Mill is right to maintain that the right not to be a slave should be regarded as inalienable, this simply raises the question of whether the right to have an intimate relation with any other (otherwise available) person should be similarly regarded. Even if the waivers have the same degree of permanence, the scope and moral quality of the loss of freedom entailed by the waivers are importantly different. It is, of course, en-

202

CHAPTER SIX

tirely possible that a less permanent ban strikes the right balance between protecting the integrity of the therapeutic process and upholding freedom of choice, but I see no reason for thinking that a more permanent ban cannot be legitimately considered as a policy option. It should not be ruled out, in advance, on the ground that it violates the rights of the parties. And if a posttermination ban does not unjustifiably violate the rights of the parties, then certainly a pretermination ban does not do so.

Chapter Seven

UNFAIR TRANSACTIONS

AN EXPLOITATIVE TRANSACTION is

one in which A takes unfair advantage of B. A engages in harmful exploitation when A gains by an action or transaction that is harmful to B where we define harm in relation to some appropriate baseline. A engages in mutually advantageous exploitation when, in relation to the same baseline, A gains unfairly or excessively by an action or transaction that is beneficial to B. Bracketing questions about fairness and exploitation, there could be disagreement as to whether a transaction is harmful or advantageous to B and, if advantageous, to what extent. This dispute may turn on empirical issues about the effects of the transaction or on moral issues about the best way to understand B's baseline. The disagreement may also turn on a philosophical issue that figured most prominently in our discussion of commercial surrogacy: how should we understand the interests of the parties to a transaction? To simplify matters, we could understand the parties' interests as subjective or objective.1 On a subjective account, a party's interests are psychological states — pleasure, happiness, desire, preference satisfaction, and the like. On an objective account, a party's interests are what is good for him or her.2 One's objective interests may, of course, include a subjective component: ceteris paribus, the satisfaction of a person's desires are good for that person. Still, on an objective account, one's interests are not reducible to one's psychological states in a straightforward way. In most cases of exploitation in which we are interested, it will not make a crucial difference which account we choose. But it can. Suppose that a professor proposes to cancel all Friday classes if the students agree.3 All happily accept the proposal. Might this be a case of harmful exploitation? On a subjective account of the (or my) students' interests, probably not. On an objective account of the students' interests, it might be argued that the proposal is bad for the students even though they all prefer it. But I shall not pursue these issues much further. For I believe that we can develop an account of the structure of 1

See, for example, James Griffin, Weil-Being (Oxford: Clarendon Press, 1986). This may include her interest in her moral character. I shall discuss that issue in a bit more detail in Chapter 9. 3 To ensure that none are pressured into agreeing, they vote by secret ballot. 2

208

CHAPTER SEVEN

exploitation that is largely neutral among different empirical findings and philosophical views of interests, good, or well-being, although such views will be crucial to determining just when people actually are exploited and whether exploitation is harmful or mutually advantageous. However these issues are resolved, I think it is clear that the terms or substance of a transaction must be unfair if it is to be exploitative. Consider a variant of one of the introductory examples. A fair-minded but not overly altruistic entrepreneur roams the highways in a snowstorm and offers to rescue stranded motorists for an eminently fair price. He takes advantage of the rescuees' plight in order to supplement his income. True, his rescuees typically feel that they have "no choice" but to accept his help at the proposed price. Still, he hardly exploits their plight in a pejorative sense. No unfairness in the terms of the transaction, no exploitation.4 But what makes the terms of a transaction unfair or wrong? We have been working with two dimensions of transaction: (1) the benefit to A, which is an ingredient in both harmful and mutually advantageous exploitation, and (2) the fairness of the transaction to B, which is the central issue in mutually advantageous exploitation (I assume that transactions that are wrongfully harmful to B are also unfair to B). I propose that we push the analysis of those two dimensions somewhat further. BENEFIT TO A

I have argued that A cannot exploit or take unfair advantage of B unless A obtains some advantage through the transaction with B. In Chapter 1 we saw that A can abuse B without exploiting B, as when A inflicts physical harm on B without benefiting from that imposition of harm. We saw that A can discriminate against B without exploiting B, as when A deprives B of an opportunity but does not gain from that deprivation in any important way. We saw that A can oppress B without exploiting B, as when A deprives B of freedom but does not gain from that deprivation of B's freedom. These observations still seem right in the light of the analyses in Part I, and I see no reason to greatly expand upon or revise that part of the discussion. But we have learned about this dimension of exploitation along the way. Certain points can now be emphasized and other points can now be developed. 4

As I noted in Chapter 1, Allen Wood appears to disagree with this view. See "Exploitation," 12 Social Philosophy and Policy 136 (1995), at 147.

UNFAIR TRANSACTIONS

209

We have seen that A does not actually exploit B unless A gains from the interaction, even if A seeks to gain from the transaction. We can distinguish between the claim that A exploits B and the claim that A acts exploitatively towards B. A university that seeks but fails to gain from its use of student athletes may act exploitatively, but the student athletes cannot claim that they are exploited. They could not (and do not) claim that they should receive a greater share of the university's profits if there are no profits to be shared.5 At several points in this book I have explored the distinction between an ex ante and ex post view of B's gain from a transaction. That distinction applies here as well. We should distinguish between the case in which A unsuccessfully seeks to gain from a transaction (as in an attempted theft), in which case A acts exploitatively but does not exploit, and the case in which A receives an ex ante but not ex post gain from a transaction with B, as when A steals a losing lottery ticket from B. In the latter case, A does exploit B even though A does not realize an ex post gain which could be shared with B.6 The distinctions among "A gains/' "A unsuccessfully seeks to gain" and "A gains ex ante but not ex post" pose interesting philosophical puzzles, and they have implications for what A owes B in the way of restitution, but I shall not pursue them further. I do not claim that A exploits B only when A intends to gain a benefit at B's expense. A may exploit B even though A believes the terms of the transaction are eminently fair—if they are not. And A may exploit B while (and perhaps even because) A is unaware of the effects of the transaction on B. Free riders frequently do not understand that they are riding free. Of course, as with other forms of wrongdoing, we can distinguish between the wrongness of the act and the agent's culpability for that act. We may regard intentional exploitation as worse than negligent exploitation, and there may even be completely nonculpable exploitation. But it would still be exploitation. Several additional points can be made about the character of A's gain. First, in the case of harmful exploitation, A's gain need not be particularly large. The exploitation need only be wrongfully harmful. 5 See Chapter 3. Of course if A should squander its benefit from the transaction with B, B could still claim that he had been exploited even though A had nothing (left) to share with B. But that would still be distinct from the case in which A realizes no benefit at all. 6 What does A owe B in this case? A might owe B the ex post value of the lottery ticket, in which case A owes B the full value if it wins and nothing if it loses. A might owe B the ex ante value, in which case A owes B either what B paid for the tickets or perhaps less, given that the ex ante value of a $1 ticket might be only $.60. Or perhaps A owes B the ex ante value if the ticket loses, but the ex post value if the ticket wins.

210

CHAPTER SEVEN

As we noted in Chapter 2, there may be cases in which there is a highly competitive market among sellers of some product, such that any given seller makes only a small profit from the transaction. If A sells appliances to people who cannot afford them, we may say that A exploits them, not because he is in a position to offer better terms (it is possible that A's profit margin is extremely low), but because such transactions are harmful to the buyers. If A cannot offer better terms, then A should not transact with B at all. Of course, and as we saw in Chapter 6, A's gain need not be financial. Psychotherapists who exploit their patients may not garner a huge benefit from the sexual encounters. They exploit because they seek an inappropriate gain from an action that is harmful to their patients. And that brings us to a second point. Although the history of exploitation discourse has focused on monetary gains to A ("surplus value"), we have seen that A's gain can take other forms. In the case of student athletes, we have the possibility that fans are obtaining entertainment value without paying for it. In commercial surrogacy, we have the gain to the adoptive parents of an infant. In psychotherapy, we have the gain of sexual pleasure. In the case of unconstitutional conditions, a government may gain political advantages by getting beneficiaries to waive a constitutional right, although it may well be acting on behalf of other citizens when it does so. And that brings us to the third point. Although the paradigmatic exploitation claims involve cases in which A seeks to promote his or her se//-interest, we need a more protean conception of what counts as a benefit to A, one that includes A's purposes, goals, and values. In some cases, for example, A exploits B on behalf of C. Call this mediated exploitation. Consider the Legal Aid lawyer who advises her client to reject a generous settlement because she wants to litigate a new legal doctrine that, if the litigation is successful, will serve the interests of a larger group. The lawyer may have exploited her client on behalf of others, but she has exploited her client nonetheless.7 It is not always clear just who is doing the exploiting in a case of mediated exploitation. In some cases, C (an individual or group) may have authorized or asked A to seek a benefit for C from a transaction with B, in which case we may hold C (as well as A) culpable for the exploitation. In other cases, perhaps illustrated by the Legal Aid lawyer, C has neither authorized nor asked A to seek a benefit for C from A's transaction with B. Although we would not hold C culpable for 7

A similar relationship may occur in medical experimentation, where participation in an experiment may not be optimal for an individual patient but will produce important knowledge. A physician who gets a patient to "consent" under such conditions may not be serving his or her own self-interest, narrowly defined.

UNFAIR TRANSACTIONS

211

any exploitation in the latter case, it is only because C gains from the A's transaction with B that we can say that A exploits B. We could distinguish mediated exploitation from what we might call misguided altruism. Consider the Laetrile Society. Laetrile is a substance derived from apricot pits that has been touted as a cure for cancer, although there is no scientific evidence that it has any therapeutic effect.8 Unlike some disreputable pharmaceutical companies, which might stand to profit from the sale of laetrile, The Laetrile Society raises money from donations and provides laetrile to cancer patients without charge. The Laetrile Society may act wrongly, but it does not exploit. It is not that it fails to exploit just because its motives are benign and does not gain directly. For that is true of mediated exploitation. The Laetrile Society does not exploit because it is seeking to promote the interests of the persons with whom it is interacting. By contrast, the legal aid lawyer exploits her client because she is seeking to promote the interests of others at the expense of her client, even if her aims are otherwise eminently admirable. The fourth general point about the gain to A concerns the baseline by which that gain is measured. When A and B consider a transaction, each party calculates his or her Best Alternative To a Negotiated Agreement (BATNA). A party's BATNA is a course of action, for example, to do nothing or to transact with another party. A BATNA has a value to each party. Call it the reservation value. With reference to that reservation value, each party determines his or her reservation price, a price that is just equal to the reservation value, the minimum threshold value that he or she is prepared to accept for entering into an agreement.9 A party gains from a transaction when he or she receives more than the reservation price. The concept of a party's reservation price is usually treated as unproblematic. For some analytical purposes, it may well be unproblematic. But for our purposes, it is not. There are several issues that might be raised, but here I want to focus on the choice between what I shall call A's actual reservation price and A's morally justified reservation price.

A party's actual reservation price is a counterfactual: it is the minimum price that the party would be prepared to accept. In some cases, one may have a relatively firm and correct understanding of one's

8 See Marion Smiley, "Legalizing Laetrile/' in Amy Gutmann and Dennis Thompson, eds., Ethics and Politics (Chicago: Nelson-Hall, 1990). Although laetrile may do no direct harm, it may cause some cancer patients to forgo potentially beneficial therapy. 9 See Howard Raiffa, The Art and Science of Negotiation (Cambridge: Harvard University Press, 1982), p. 45.

212

CHAPTER SEVEN

own reservation price, but in other cases, one will not. What is my reservation price for changing a student's grade from B to A? I do not know.10 One can have a firm but mistaken conception of one's own reservation price. Before entering into negotiations with B, A may think he is not prepared to sell his house for less than $150,000. But he may find, even to his surprise, that he is prepared to sell for less. A party's actual reservation price is arguably counterfactual in another way. It is a function of one's information about the market (in general) and the reservation price of the parties with whom one is bargaining. In most cases in which there is a market for what is being exchanged, a party's actual reservation price will not be much lower than the market price, particularly if A is aware that there is a market price for the good or service.11 One's reservation price is typically not independent of what one takes to be the price that other parties are willing to pay (or accept). Suppose that A's husband dies and leaves A with a baseball signed by Babe Ruth. A has no interest in baseball or its history. Its premarket subjective value to A is close to zero. B proposes to buy the baseball from A for $1,000—the going market price. A accepts. How much has A gained? It might be argued that A has gained $1,000, as measured against her (premarket) actual reservation price, raising the possibility that A has exploited B who, after all, may have gained little as compared with his reservation price. But this seems wrong. To see why, let us suppose that while A herself has no interest in baseball or its history, she understands that the market price for baseballs autographed by Babe Ruth is $1,000 and she is determined not to accept (much) less. Her actual reservation price is now $950 (say, because the value to A from the $50 is not worth the costs in time and inconvenience that she would have to endure in order to get the market price).12 On this view, A has gained at most $50 if she accepts B's offer of $1,000, and would not be particularly vulnerable to a charge of exploitation. But A is not vulnerable to a charge of exploitation in the previous case not just because her gain is so modest. A does not exploit B because A is morally justified in refusing to accept less than the mar10

I would like to think that there is no monetary price for which I would change a grade, but that may not be true. This is the theme of a 1990s movie, Indecent Proposal, in which a character played by Robert Redford offers a character played by Demi Moore $1 million to spend the night with him. 11 A's actual reservation price might be higher than the market price if the good or service has special value to A. 12 David Miller discusses a similar problem in slightly different terms in "Exploitation in the Market," in Andrew Reeve, ed., Modern Theories of Exploitation (Beverly Hills, Calif.: Sage Publications, 1987), pp. 149-50.

UNFAIR TRANSACTIONS

213

ket price, and it is her morally justified reservation price that seems to block the charge of exploitation. To see this, we must ask whether A is always morally justified in refusing to accept less than the market price? I think not. Consider a variant of The Port Caledonia and the Anna.13 Suppose that A, a tugmaster, is in a position to rescue only one of two vessels in distress—B or C. Although A would have charged $100 for his services under normal conditions, B and C begin to bid for A's services. B's last offer is $5,000; Cs last offer is $4,900. Given C's offer, A refuses to accept less than $4,900 from B. How much has A gained if A rescues B? If A's actual reservation price is defined in terms of the opportunity cost to A of not rescuing B, then A has gained only $100. For A's opportunity cost or BATNA of not rescuing B is the $4,900 that A would have received for rescuing C. But here is a case in which we might think that A's morally justified reservation price is less than the actual reservation price, that A is not entitled to receive "surplus rent" that is caused by this unusual situation.14 A's actual reservation price and A's morally justified reservation price might diverge for a different reason. Consider a variant of the previous case. Suppose that A would not much mind seeing B sink because A dislikes B or harbors a prejudice toward people of B's group.15 A's normal actual reservation price for such a rescue is $100, but given A's dislike for B, his actual reservation price for rescuing B is $4,900; A is genuinely indifferent between rescuing and not rescuing B for that price. How much does A gain if he rescues B for $5,000? $100? $4,900? I am inclined to think that the answer is $4,900 and that this shows that we should ordinarily understand A's gain in terms of A's morally justified reservation price. This may be no less than A's actual reservation price in most contexts, but it need not be. It should be stressed that a morally justified reservation price defines a baseline and not a ceiling. How much more than the baseline one may take without exploiting the other party depends on the scope of the bargaining range or zone of agreement. And so we may need to determine a morally justified reservation price for B as well as for A. B may wrongly believe that a transaction is exploitative because 13

See Chapter 2, note 13 and accompanying text. I say "might" because it is not clear how A should choose how to allocate his services under these conditions. It might be argued that competitive bidding is a useful allocation device, because it allocates A's services to the buyer who values them most, as revealed by "willingness to pay." But it could be argued that A ought to allocate his services by a lottery. 15 Michael Gorr considers such a case in Coercion, Freedom and Exploitation (New York: Peter Lang, 1989), p. 156. 14

214

CHAPTER SEVEN

her actual reservation price is unreasonably low. Although B may not be prepared to pay more than X for some good, we may think that B's morally justified reservation price should be higher than her actual reservation price.16 A final point about this dimension of exploitative transactions—the gain to A—concerns its moral significance. Suppose that A were to enslave P (a productive worker) and U (an unproductive one) in the hopes of generating surplus value from them. There is no significant difference in the day-to-day quality of P's and U's lives as slaves. They work equally hard. It turns out, however, that P generates value far in excess of what is required for his subsistence, whereas U generates value that is approximately equivalent to what is required for his subsistence. If I am correct, although A acts exploitatively toward both P and U, A exploits P but does not exploit U. The present question is this: Given that A exploits P but not U, is A's relationship with P morally worse than his relationship with U? Or, given that P's life and U's life are equally bad, is A's relationship with P and U equally bad? There are several possibilities. On what might be called a current time-slice utilitarian view, it might be thought that exploitative slavery is less bad than nonexploitative harm because at least someone gains (A), and a gain in utility is always good no matter who receives that utility. On what might be called an objective good view, illegitimate gains in utility do not count in calculating overall good, just as we do not count the pleasure that racists receive from acts of discrimination as mitigating the wrongness of discrimination. On what might be called the organic whole view, we may accept that the gain to A is a good, but we go on to argue that adding a good to a bad state of affairs may make the overall state of affairs worse rather than better— "The value of a whole must not be assumed to be the same as the sum of the values of its parts." 17 It may be worse if A takes (unjustifiable) pleasure in B's misery than if A feels sorrow over B's misery, even if, ceteris paribus, pleasure is intrinsically good and sorrow is intrinsically bad. Similarly, A's relationship with P may be worse than his relationship with U, even though the gain to A is a good. In my view, the current time-slice utilitarian view is simply implausible. The objective good view seems to be at least part of the truth, but I do not think it captures the whole story. There may be some truth to the organic whole view, but it seems unduly complex. We can reach a similar view by focusing on fairness rather than organic 16

I thank Arthur Applbaum for pressing me on this point. G. E. Moore, Principia Ethica (Cambridge: Cambridge University Press, 1903), ch. 1, sec. 18. 17

UNFAIR TRANSACTIONS

215

wholes. It is unfair that A receive the entire surplus of P's productivity when an alternative distribution in which P receives some or all of that surplus is eminently feasible. By contrast, it is possible that there is no alternative—fairer—distribution of U's productivity under which U could do (much) better, because U is not capable of producing more than is required for his subsistence. On the fairness view, A's relationship with P is worse than A's relationship with U because A's relationship with P gives rise to a type of unfairness that is absent in A's relationship with U. There is a fourth view about the moral significance of A's gain that should be mentioned. On this view, a form of rule consequentialism, it makes sense to regard (1), the state of affairs in which A gains by enslaving P, as worse than (2), the state of affairs in which A does not gain by enslaving U, even if (1) is not worse than (2).18 Given human behavior as we know it, it is more likely that A will engage in acts that are harmful to B if A stands to gain from those acts than if he does not. Although people do commit "senseless" acts of harm from which they do not gain in any ordinary sense, we have more reason to worry about those acts of harm from which someone does gain. On this view, the calibration of the wrongness of a person's actions is analogous to a Benthamite calibration of legal penalties. Given that legal penalties are at least partially designed to make it less likely that people will engage in unwanted behaviors, the magnitude of the punishment must reflect the positive value of the crime to the criminal and not just the negative value or harm to the victim or society.19 On this view, we regard the exploitative enslavement of P as worse than the nonexploitative enslavement of U, because we have more reason to worry about exploitative enslavement than about nonexploitative enslavement. Similarly, we may regard the sexual exploitation of patients by psychotherapists as worse than simple malpractice (e.g., prescribing incorrect drugs), not solely because the victims of sexual exploitation are harmed more than the victims of malpractice (they may not be), but because it takes a higher degree of moral censure to deter acts from which psychotherapists gain than to deter acts from which they do not. It might be objected at this point that the degree of censure necessary to deter an act plays no role in determining the wrongness of the act, that even if we have more reason to worry about exploitative harm than about nonexploitative harm, it does not follow that exploit18

See R. M. Hare, Moral Thinking (Oxford: Clarendon Press, 1981). "The value of the punishment must not be less in any case than what is sufficient to outweigh that of the profit of the offence." Jeremy Bentham, An Introduction to the Principles of Morals and Legislation (1823 edition), chs. XIV, VIII, Rule 1. 19

216

CHAPTER SEVEN

ative harm is worse (although it may be worse for other reasons). I am inclined to think that this objection is on the mark. Still, unless we assume a retributive view in which a social response to wrongness must be a direct function of the wrongness of the act, we may be justified in treating exploitative harm more harshly than we treat nonexploitative harm. WHAT MAKES A TRANSACTION FAIR?

We have examined the gain to A as an element of exploitative transactions by focusing on cases of harmful exploitation. A may also gain by an action or transaction that is beneficial to B. I have argued that if such a transaction is unfair to B, we may have a case of mutually advantageous exploitation. When are transactions unfair? Background Fairness versus Transaction-Specific Fairness

To fix ideas, we are interested in transactions with the payoff structure shown in Figure 7.1. One might object to setting the No Transaction payoffs at 0. After all, in many cases of alleged mutually advantageous exploitation, A's status quo is superior to B's. I believe, however, that we should distinguish between fairness as a principle for the distribution of social resources and fairness as a principle for transactions and that principles of fair transactions should bracket information that might be relevant to other moral purposes, such as justifying aid or redistribution. So in setting both party's payoffs in the No Transaction alternative at 0, we deliberately abstract from certain elements of their background situations, such as the parties' overall welfare. The payoffs are transaction specific. On my view, then, the greedy snowstorm rescuer may be guilty of exploitation even if he is relatively poor and the rescuee is relatively affluent. Setting both party's payoffs in the No Transaction alternative at 0 also reflects the background rights of the parties. It reflects the assumption that A has a right not to transact with B at all. As we have seen, there are cases in which that assumption is questionable. For No Transaction

Transaction

0

>0

0

>0

Figure 7.1

UNFAIR TRANSACTIONS No Rescue A

0

B