Buying Freedom: The Ethics and Economics of Slave Redemption 9780691186405, 9780691130095, 9780691130101, 2006050993

Citation preview

BUYING FREEDOM

BUYING FREEDOM: THE ETHICS AND ECONOMICS OF SLAVE REDEMPTION

Edited by Kwame Anthony Appiah and Martin Bunzl

With a foreword by Kevin Bales

PRINCETON UNIVERSITY PRESS

PRINCETON AND OXFORD

Copyright © 2007 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, 3 Market Place, Woodstock, Oxfordshire OX20 1SY All Rights Reserved Library of Congress Cataloging-in-Publication Data Buying freedom : the ethics and economics of slave redemption / edited by Kwame Anthony Appiah and Martin Bunzl. p. cm. Includes bibliographical references and index. ISBN-13: 978-0-691-13009-5 (alk. paper) ISBN-13: 978-0-691-13010-1 (pbk. : alk. paper) 1. Slavery. 2. Slaves—Emancipation. 3. Redemption (Law)—Moral and ethical aspects. 4. Redemption (Law)—Economic aspects. I. Appiah, Anthony. II. Bunzl, Martin. HT857.B65 2007 306.3
62—dc22 2006050993 British Library Cataloging-in-Publication Data is available This book has been composed in Sabon Printed on acid-free paper. ∞ press.princeton.edu Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

Contents

Foreword Kevin Bales

vii

Introduction Kwame Anthony Appiah and Martin Bunzl

1

PART I: THE ECONOMICS OF REDEMPTION

7

Chapter 1: Some Simple Analytics of Slave Redemption Dean S. Karlan and Alan B. Krueger

9

Chapter 2: Slave Redemption When It Takes Time to Redeem Slaves Carol Ann Rogers and Kenneth A. Swinnerton

20

Chapter 3: An Exploration of the Worst Forms of Child Labor: Is Redemption a Viable Option? Arnab K. Basu and Nancy H. Chau

37

Chapter 4: Slavery, Freedom, and Sen Stanley Engerman

77

Chapter 5: Freedom, Servitude, and Voluntary Contracts Jonathan Conning and Michael Kevane

108

PART II: ANTHROPOLOGICAL CONSIDERATIONS

141

Chapter 6: Slavery and Slave Redemption in the Sudan Jok Madut Jok

143

Chapter 7: Dilemmas in the Practice of Rachat in French West Africa E. Ann McDougall

158

PART III: HISTORICAL CONSIDERATIONS

179

Chapter 8: The End of Serfdom in Russia—Lessons for Sudan? Lisa D. Cook

181

Chapter 9: Conflicting Imperatives: Black and White American Abolitionists Debate Slave Redemption Margaret M. R. Kellow

200

vi

CONTENTS

Chapter 10: Frederick Douglass and the Politics of Slave Redemptions John Stauffer

213

PART IV: PHILOSOPHICAL CONSIDERATIONS

223

Chapter 11: The Moral Quandary of Slave Redemption Howard McGary

225

Chapter 12: The Next Best Thing Martin Bunzl

235

Chapter 13: What’s Wrong with Slavery? Kwame Anthony Appiah

249

Appendix: “They Call Us Animals,” Testimonies of Abductees and Slaves in Sudan Jok Madut Jok

259

List of Contributors

269

Index

271

Foreword KEVIN BALES

This multidimensional exploration of redemption is exciting, intriguing, and deeply hopeful. Slave redemption has been a controversial subject for centuries, and the controversies that have surrounded it in the past and surround it today are themselves multifaceted. Morality, economics, politics, and the proper strategies and tactics for eradicating slavery generally, all have a part in the comprehension of this issue. As often happens when lives are at stake, the discussion over redemption has, at times, become polarized, even vicious. Clarity of thought is not enhanced by passion, but when confronting monolithic systems of terror and exploitation, the passionate and single-minded adherence to one path of action may be seen as the only chance of success. I have had the uncomfortable good fortune of experiencing this controversy in person. For some fifteen years I have been exploring firsthand the ideas of slave liberation and redemption, and at the same time living in the global laboratory where these ideas are being tested. Looking into the faces of slaves, it was sometimes difficult to know the correct path of action—what intervention would be most likely to lead to a successful liberation and rehabilitation? It was always clear that a broad and careful analysis of redemption was needed. That this need might be met is the basis of the hope I feel underpins this book. There are still some 27 million of slaves in the world today. Over the next decades, many thousands of these slaves and free people may find that their lives intersect as some benefit from and others support redemption. While this book does not provide an answer to the question of redemption in any individual’s case, it extends and deepens our understanding of the subject, and makes it more likely that we can respond to the lived experience of slaves in ways that do the least harm, and perhaps opens the opportunity for liberation and true autonomy. As an example of the guidance that is so very needed, let me point to the chapter by Kellow. Working with the original and oldest anti-slavery organization, Anti-Slavery International, and establishing its sister organization in the United States, Free the Slaves, in the late 1990s, I was directly confronted with the question of real slaves. There were significant and sometimes potentially damaging consequences for slaves, the organizations, and myself, consequences that were linked to the possible

viii

KEVIN BALES

action (or inaction) that might be taken to address the needs of these slaves through redemption. After many discussions (and a number of arguments) with people directly involved in the debate over redemption in Sudan, and trying to find a position that was right both personally and for our organization, I wrote: The lesson to be drawn from these examples [from India, Sudan, Pakistan, medieval Europe, and the antebellum South] is this: redeeming slaves has a role in the anti-slavery movement, but only when other actions have failed or are impossible. In one way it is the lesser of several evils. The greatest evil is enslavement. Paying a slaveholder or a middleman to free a slave is regrettable since freedom is a right and shouldn’t need to be purchased. Yet, when there are situations where the authorities will not take action, where the opportunity for freedom is available only through redemption, and where enslaved people are suffering and threatened, then redemption may be the only immediate answer. A good case can be made both for and against redemption; a key criterion has to be that it be used only when it won’t make things worse. It is a tactic in an overall movement against slavery, one to be used when other actions won’t serve.1 At a time when redemption was being presented as a strategy for antislavery work, our position was that it was only acceptable as a tactic of last resort. This “middle-of-the-road” position earned condemnation from both sides of the debate. How reassured I would have been to know William Lloyd Garrison’s view, as reported in Kellow’s chapter: that he had always maintained “the demand of the slaveholder for compensation was an unjust one. But I see no discrepancy in saying that a certain demand is unjust, and yet be willing to submit to it, in order to save a brother man, if this is clearly made to be the only alternative left to me.” Kellow’s assertion that the nineteenth-century abolitionists “were neither unanimous nor entirely consistent in their objections to slave redemption” would have helped me to understand the truly circular history of the issue. Learning from the past can be crucial when addressing the present, for there is a science as well as a history of redemption. The chapters on the economics of redemption that open this book represent an overture marking the transition from proto-science (where there is little agreed-on literature, debate, and peer review) to an emerging science. Within this new literature I would suggest that it is important to recognize the impact of the dramatic fall in the price of slaves over the past fifty years. After millennia during which slaves tended to be high-cost and capital purchase 1 Originally an internal memo for Free the Slaves staff, an expanded version was published as “Globalization and Redemption,” in Bales (2005), 112–125.

FOREWORD

ix

8

Price Index

7 6 5 4 3 2 1 0 BC BC C D 00 00 0B A 50 16 0 0 20 14 80 200 4 8 100 40 700 14 50 7 0 1 1 17 7 1 184 185 853 856 59 0 9 1 1 18 92 3 43 0 1 19 19 00 04 2 20 Year

Figure F.1 Price of slaves 2000 BC to 2004 AD. items, today their price has fallen to the same level as many goods seen to be “disposable.” Figure F.1 shows this collapse in the price of slaves. The price index is not based on monetary costs of slaves, but on the relationship of the price of slaves to the prices of oxen, productive land, and free agricultural labor over time. For example, if the cost of a slave was approximately equal to that of five oxen or the cost of hiring a free agricultural worker for fifteen months in the United States in 1856 (represented by a score of 4 on the price index in fig. F.1), the price of a slave has fallen to one-sixth of an ox in modern India, or 4 percent of the annual wages of a free farm worker in the Ivory Coast today. I point to this fall in price because of the importance it could have for economic analyses of all forms of enslavement, including the analysis of redemption. Because most slavery today is part of a hidden economy, and because in Sudan in particular buyers could be ignorant of price trends, models that assume a general level of shared knowledge need careful examination. This is especially the case for models comparing the relatively open economic system of the antebellum South with that of Sudan in the late twentieth century. Slave prices were general knowledge and public record in the United States in the 1850s. But what did redemptionists flying into southern Sudan in the 1990s know of the price of a slave in Khartoum? What did they know of historical pricing of slaves in money, goods, or livestock in the rounds of capture and redemption played out between the southern ethnic groups? Might there have been two parallel price systems in effect, one for “consumers” of slaves and one for “redeemers”? In the

x

KEVIN BALES

Sudanese market there was a profound knowledge differential: buyers knew very little and sellers knew a great deal. In much the way that an unscrupulous taxi driver may quote a jet-lagged Japanese businessman arriving at a New York airport an exorbitant rate for a drive to his hotel, sellers in Sudan knew that well-meaning Europeans and Americans arrived with suitcases full of money and heads empty of any useful market information. A lack of market information only exacerbated the fact that redemption was occurring within a context that often had little to do with logical economic assumptions. At different times slave raiding has been, as it can be today, an act of terror that happens to have an economic byproduct. As Jok Madut Jok makes clear in his chapter, for the government of Sudan slave-taking was primarily a weapon against a threatening and troublesome minority, and secondarily a way to reimburse its surrogate soldiers for neutralizing this threat. Had the economic logic of slaveholding driven the act of enslavement, then the individuals with the greatest productive capacity, adult men, would have been captured, not routinely killed. Building the role of terror into a model of a market for slaves is a real challenge. In this book several chapters on the Sudan make valuable theoretical contributions despite the fact that no reliable data are available on redemption programs themselves. At the time when the majority of redemptions were taking place, data collection was itself controversial. One part of the as yet unwritten history of the redemption movement in Sudan is the tension between different national groups of the Christian Solidarity International (CSI) coalition over the collection of uniform statistical information on redemption. The wise and patient insistence of the Canadian wing of CSI brought about the establishment of a database of redemptions, and promises that such information would be made more widely available. The subsequent refusal to share any of the collected information, even with other Christian Solidarity groups, was one of the causes of the collapse of the CSI coalition. As Karlan and Krueger note in their chapter, “the price of slaves traded will rise with a redemption program if the supply elasticity is finite. This can be monitored.” Indeed, it was monitored, but the results were kept secret. The excellent work of the Rift Valley Institute to build a database of “Abductees” has not yet been taken up for independent secondary analysis, and it is not yet clear if those data will have immediate application to the question of redemption. However, pointing to the lack of good aggregate information is not a counsel of despair, it simply underscores the need for some basic data before central questions about redemption can be fully examined empirically. With the reduction in levels of violence in southern Sudan and the

FOREWORD

xi

return of many of those abducted and enslaved, it becomes more possible to collect that information. As those that were enslaved return from northern Sudan, related questions of rehabilitation and reintegration arise. Engerman, in his chapter, provides an important discussion of the experience of American slaves after emancipation. The process of enslavement and liberation should not be characterized as events, but processes, often lifelong. Our understanding of the needs of ex-slaves is as sketchy today as our understanding of redemption has previously been. In the work of my own organization, with our partners overseas, we are engaged in the liberation of slaves in many ways, sometimes through force, sometimes through aided escape, and sometimes through fostering the social, economic, and political changes that enable slave communities to liberate themselves. A constant concern is to support ex-slaves in attaining economic autonomy and the rights of citizenship. The lack of research on the processes of rehabilitation and reintegration means that we often feel we are flying blind, but we are motivated by the knowledge that the botched emancipation of 1865 and its tragic aftermath carried the crippling legacy of slavery into subsequent generations. This is an outcome we are determined never to replicate. In some areas, particularly northern India, we have developed systems that seem to be generally successful at both liberation and reintegration, but with the numbers of freed slaves increasing we look to others to carry out the research that will illuminate this process and help academics and practitioners comprehend its complexities. Ongoing work against slavery around the world, including the tens of thousands of cases of enslavement in North America, means that the question of redemption is not one of simple historical or academic interest. While the scale of slave redemption in Sudan drew the attention of the public and academics alike, it was not an isolated instance. Today, in many countries and situations, the possibility exists to enter into an economic transaction that will lead to a radical transformation, for better or worse, of a person’s status as a slave, bonded worker, or forced laborer. The World Bank has supported a number of projects using “conditional cash transfers” to remove children from the workforce, and it would be a small step to consider using the same approach with children (or adults) in slavery. This volume considers areas of the world in addition to Sudan, such as McDougall’s work on French West Africa. We should build upon the work of this book to further expand our view of redemption to include its many possible forms and manifestations. Many of these forms will fit better to economic models in that they are not played out in a context of conflict.

xii

KEVIN BALES

It may seem a remarkable, even naïve, assertion, but I believe that for the first time in human history there is the possibility of bringing most slavery to an end. The main supporting factors for this assertion are that slaves are now at their smallest ever proportion of the global population, that they represent an extremely small capital investment within the broader economy, that the total number of slaveholders worldwide is probably no more than five to six million, that no major economic or political interests are reliant on slavery for survival, and that the international treaties and conventions banning slavery are now universally accepted. These are, of course, also reasons why the lives of the enslaved seem to drop from the view of policy makers. Whatever the case, redemption is one way to bring slaves out of bondage. It is probably not the best way to accomplish this, or even one of the best ways, but its logic, its mechanism of exchange, is both simple and effective in its simplicity. When faced with a person in slavery, we need a careful and dispassionate understanding of every tool that might lead to liberation. This volume makes a major contribution to that understanding.

Bibliography Bales, Kevin (2005), Understanding Global Slavery: A Reader. Berkeley: University of California Press.

Introduction KWAME ANTHONY APPIAH AND MARTIN BUNZL

This volume has its origins in an unsolicited telemarketing call. One of us (MB) was asked for money to (purportedly) free a slave in Sudan and was intrigued enough to solicit views from a variety of human rights and international relief organizations. Struck by the near unanimous condemnation of the practice, the editors began an extended conversation with others about just what (if only under idealized circumstances) would make such a practice morally wrong. We began with a small meeting at the Center for Human Values at Princeton, where we brought together a multidisciplinary group of philosophers, anthropologists, and economists, to begin thinking about these questions. Since then we have carried our discussions further and enlarged our numbers, continuing to exchange and develop our ideas. The results of the conversations prompted by the original practical question are represented in what follows. When we began this endeavor we thought of this as primarily a philosophical exercise. And indeed there are traditional arguments that can and have been marshaled against the practice of buying others even if it is to grant them their freedom. A UNICEF spokesperson captures the deontological cast of such arguments well in objecting that a buy-back program implicitly accepts that human beings may be bought and sold (Peter Crowley as reported by Lewis 1999). More broadly, it might be thought that, irrespective of the good consequences that might follow, buying human beings treats them as a commodity and therefore not as an end in themselves. Such arguments also figured large in some of the nineteenth-century debates (both in the African-American community and among Quakers) that Margaret Kellow takes up in chapter 9. Howard McGary assesses the philosophical merits of such arguments in chapter 11. That said, the vast majority of objections to buy-back programs have a decidedly more consequentialist cast. And consequentialist arguments often stand or fall on the underlying facts that they ride on. In this case, what they ride on is a matter of economics. Most people have a strong intuition that to enter the market for slaves will raise demand, thereby pushing up prices, which will draw more people into the slave trade, resulting in more people being enslaved. That looks like a classic case of helping the one but hurting the many (more). And such an intuition underlies the objections of a number of organizations (including UNICEF and Human Rights Watch, as Kreuger and Karlan point out in chapter 1).

2

APPIAH AND BUNZL

But is this intuition in fact correct? The first three papers in this volume examine this issue from different perspectives. In chapter 1, Dean Karlan and Alan Krueger use standard demand-supply market analysis to show why buying slaves would, at worst, result in their replacement on a onefor-one basis, but that with plausible assumptions about the elasticity of supply, the likely outcome would be a lower “rate” of replacement. In chapter 2, Carol Rogers and Kenneth Swinnerton argue that, to be complete, such an analysis needs to also take into account determinants of the rate of flow (or time) for people to move through the process of enslavement and redemption, as well as whether the suffering of enslavement is evenly spread over that process, as opposed to being more concentrated around the initial time of capture. In chapter 3, Arnab Basu and Nancy Chau use existing data on child trafficking and bondage to argue that market imperfections create impediments to easily achieving the simultaneous goals of granting human rights to the enslaved and decreasing the incidence of enslavement. Suppose that, intuitions notwithstanding, buying the freedom of some slaves does not hurt the majority. Still, does it help the few? How could it fail to do so? In chapter 4, Stanley Engerman makes the argument that it can indeed, in the context of economic history, in which he discusses the falls in output that accompanied the ending of slavery and the implications for the availability of foodstuffs. In chapter 5, Jonathan Conning and Michael Kevane take up the issue in a theoretical context, arguing that the collusion of landlords to restrict the freedom of ex-slaves to enter into certain kinds of contracts with landlords (combining their labor and land use) can result in lower levels of welfare for the emancipated than they experienced under slavery. If the philosophical argument about slave redemption rests on the facts of economics, it does not rest on these facts alone, except insofar as one poses that argument in purely hypothetical terms. To pose it with an interest in its practical import means looking at the circumstances of slavery itself, in places where redemption is an option. Here we do so from a variety of anthropological perspectives. In the case of Jok Madut Jok (chapter 6), the argument is that slave redemption by foreigners in the Sudan should be properly understood in both the context of the war and the long-standing policy of family-initiated redemption. In evaluating the economic consequences of the practice, he shows how there can be quite unanticipated consequences—in the case of Sudan, depending on who controlled the currency exchange from dollars to Sudanese dinars. Jok’s contribution is followed by an account (by Ann McDougall, chapter 7) that situates such questions in the complex anthropological geography of Mauritania. McDougall’s contribution shows that one cannot evaluate

INTRODUCTION

3

the significance of attempts to end slavery from the “outside” without understanding its local cultural history. In chapter 8, Lisa Cook examines whether or not there are lessons to be learned from the emancipation of serfs in Russia that might be applied to modern cases of slavery. Notwithstanding these complexities, historically, opposition to slave redemption, even when held as a matter of principle, runs up against the contingencies of practice. This tension is discussed in chapter 9 by Margaret Kellow in her account of debates about redemption in both the nineteenth-century African-American and Quaker communities. Even where redemption was practiced, just under what description it was to be understood turned out to be crucial, as John Stauffer points out (in chapter 10) in his treatment of Frederick Douglass’s view of the purchase of his freedom by British Quakers. Suppose redeeming a slave helps the one and does not hurt the many. And suppose we answer objections to the implicit commodification of human life involved in such transactions. Still, ought we not to consider if we could do greater good by helping others instead—say the starving? But as Martin Bunzl argues (in chapter 12), such interests ought to be tempered by attention to just how likely it is that such prescriptive advice will actually be followed, even if doing so would produce a greater good than freeing a slave. The last chapter contains reflections by Anthony Appiah on the moral significance of slavery. He argues that once we understand what is wrong with slavery we will see that legal emancipation is only the beginning of a process of freeing the enslaved from the consequences of the combination of low status and minimal autonomy that are at the heart of the evil of enslavement. Despite its breadth of scope, this volume was conceived during and prompted by consideration of the conflict in southern Sudan. Whether the recent end of this conflict represents a permanent cessation of hostilities remains to be seen. But as Jok Madut Jok points out in his poignant summaries of his interviews with Sudanese abductees in the appendix, even if the conflict is over (although perhaps just relocated to the Darfur region), the formerly enslaved will continue to pay the cost of slavery for the remainder of their lives. In this volume, we take ourselves to be addressing the question of the ethics of buying the freedom of an enslaved person. Just how many slaves there are in the world today is a function of how narrowly or broadly one construes the term. We tend to think of slavery as a relationship in which ownership is involved because of the centrality of ownership to the history of slavery in the United States. But as Kevin Bales points out, historically, many forms of slavery lacked such a component. Bales

4

APPIAH AND BUNZL

estimates there are 27 million slaves in the world today, using a conception of slavery that makes the control of a person through violence central.1 The numbers go up (to perhaps 200 million) on more expansive definitions.2 To treat this as a book about slavery is to construe the term more widely than the ownership conception would allow, since some of the topics (like child trafficking taken up in chapter 3) don’t fit that narrow view. That said, we are largely indifferent to this (semantic) issue— what concerns us here are questions about people in the world that lack freedom. Despite the fact that their lack of freedom may be different in origin, circumstance, or degree, what they have in common is that in each case we can ask the same set of questions: What can one do about it? What ought one to do about? In helping the one does one hurt the many? In helping the one, does one actually render the one better off? One thing we hope this book demonstrates is that, in thinking about the moral issues we face when we consider intervening in other societies through the market, we must rely on many kinds of expertise. Faced with a straightforward question—should I send money to redeem a slave?— it turns out we must seek guidance from many fields. Making a final 1 “Slavery is a social and economic relationship in which a person is controlled through violence or its threat, paid nothing, and economically exploited” (Bales 2000, 3). Such a conception is designed to encompass chattel slavery, debt bondage, and contract slavery (in which workers are tricked by the promise of a contract and then enslaved) (Bales 2000, 12–13). Of these, only the first involves an assertion of ownership. 2 For example, the 1956 United Nations Declaration on Slavery asserts a much wider view:

(a) Debt bondage, that is to say, the status or condition arising from a pledge by a debtor of his personal services or of those of a person under his control as security for a debt, if the value of those services as reasonably assessed is not applied towards the liquidation of the debt or the length and nature of those services are not respectively limited and defined; (b) Serfdom, that is to say, the condition or status of a tenant who is by law, custom or agreement bound to live and labour on land belonging to another person and to render some determinate service to such other person, whether for reward or not, and is not free to change his status; (c) Any institution or practice whereby: (i) A woman, without the right to refuse, is promised or given in marriage on payment of a consideration in money or in kind to her parents, guardian, family or any other person or group; or (ii) The husband of a woman, his family, or his clan, has the right to transfer her to another person for value received or otherwise; or (iii) A woman on the death of her husband is liable to be inherited by another person; (d) Any institution or practice whereby a child or young person under the age of 18 years, is delivered by either or both of his natural parents or by his guardian to another person, whether for reward or not, with a view to the exploitation of the child or young person or of his labour. [Supplementary Convention on the Abolition of Slavery, the Slave Trade, and Institutions and Practices Similar to Slavery, 226 U.N.T.S. 3, entered into force April 30, 1957, cited in http://www1 .umn.edu/humanrts/instree/f3scas.htm.] The 200 million figure is from Bales 1999, 3.

INTRODUCTION

5

decision, once we have the advice in hand, involves balancing a wide variety of considerations. And in the real world that balancing can be very difficult.

Bibliography Anti-Slavery Society (no date), http://www.anti-slaverysociety.addr.com/slavery3 .htm. Bales, Kevin (1999), Disposable People. Berkeley: University of California Press. Bales, Kevin (2000), The New Slavery. Santa Barbara: ABC-CLIO. Lewis, Paul (1999), “U.N. Criticism Angers Charities Buying Sudan Slaves’ Release,” New York Times, March 12.

Chapter One Some Simple Analytics of Slave Redemption DEAN S. KARLAN AND ALAN B. KRUEGER

The idea of purchasing the freedom of slaves is not new. Alexander Hamilton, for example, helped found the Society for Promoting the Manumission of Slaves in New York in 1785, and raised money to buy and free slaves (see Randall 2003). The Pennsylvania Abolition Society and Quaker groups likewise purchased the freedom of a significant number of slaves. Even earlier, in the sixteenth century, the church issued orders to raise a considerable sum of money to redeem the freedom of Spanish and Portuguese captives held as slaves in North Africa (Eltis 1993). Despite the long historical precedent, slave redemption remains a controversial response to the horrible practice of slavery. Two types of slave redemption programs exist, ones that merely attempt to free a certain number of slaves, and ones that are part of a concerted effort to free all slaves and forever stop future slavery in the area. In this chapter, we primarily discuss the former. The United Nations Children’s Fund (UNICEF) considers slave redemption efforts “absolutely intolerable” and has condemned the practice in the Sudan (Lewis 1999). UNICEF makes two claims, that purchasing slaves’ freedom for money (1) exacerbates the slave trade, and (2) does not address the root of the problem. Human Rights Watch (2002) also has denounced slave redemptions in the Sudan, citing concerns that monetary incentives could lead to more raiding of free Dinkas, although the organization acknowledges, “It has not been possible to date to ascertain whether the monetary incentive produced more raiding in practice.” The group also raises the valid concern that some of those who have been redeemed were not actually slaves. There seems to us little doubt that redemption efforts do not address the root of the problem unless accompanied by more systematic emancipation and enforcement policies. This acknowledgment is not necessarily dispositive vis-à-vis slave redemption programs, however. From a practical standpoint one could still be interested in whether a redemption program reduces the number of slaves in captivity. This is the primary question analyzed here. Here we apply conventional economic tools to analyze the likely effects of a slave redemption program on the quantity of slaves in captivity.

10

KARLAN AND KRUEGER

Although our simple approach undoubtedly misses important features of the slave trade, we suspect it provides a good starting place for discussions of slave redemption programs. Our model applies most directly to chattel slavery, but even in situations where explicit markets for slaves do not exist, we suspect that our model provides insights into the latent costs of slave acquisition and the determinants of the number of slaves in captivity. If nothing else, our simple application of the supply-and-demand model to slave redemption highlights fundamental forces at work in a slave redemption program in a situation in which only pure economic interests are at work. In the economic framework we describe in the next section, “The Basic Model,” slave redemption programs are closely analogous to programs to repurchase high-emission cars. States such as Arizona and California have initiated programs to remove older vehicles with high rates of pollution emission from the road. Similar programs have been used in countries such as Argentina, Canada, Denmark, Norway, Scotland, and Hungary (Dill 2001). In some cases owners are compensated with cash, in others with transportation vouchers or loans for new cars. These programs are designed to use the market system to reduce air pollution, and are generally viewed as having been successful (Dill 2001). In both cases, slavery and old cars, the owner of the “property” is posing a harm to society (one a moral or ethical outrage, the other a polluting externality), and in both cases the proposed policy is to provide a financial incentive to the owner to induce him to cease inflicting this harm on society. If we raise the demand for older cars and then scrap them, the price of old cars rises and the number of old cars being used is reduced. A typical critique, just as in the slavery redemption case, is that these programs provide incentives to “create” clunkers (by giving people an incentive to hold onto cars longer than they would otherwise or to import old cars from out of state). Likewise, slave redemption programs create an incentive for slave dealers to bring more slaves to market. The number of slaves demanded by slave holders will be diminished under a redemption program because the price of slaves is likely to rise. In the next section, “Extension: Scarring Effects,” we discuss an important respect in which a slave redemption program differs from marketbased pollution abatement programs: the freed slave may be scarred for life. The “Cost Function” section discusses cost functions and estimates of supply elasticity. The Basic Model A standard economic approach to modeling the slave trade and slave redemption is to consider the supply of slaves to the “market” and the

SIMPLE ANALYTICS OF SLAVE REDEMPTION

11

demand for slave services. The supply of slaves in a specified period of time will depend on the cost of capturing slaves and delivering them to market, which would include any social stigma associated with being a slave supplier. The demand for slaves by potential slaveowners will depend on the value of the output that slaves produce net of monitoring and enforcement costs, boarding costs, and any stigma associated with owning slaves. The interaction of the supply and demand determines the price at which slaves are traded and the number of slaves that are bought and sold. In a competitive market, the supply curve is determined by the marginal cost of supplying a good to the market. The supply curve for slaves likely slopes upward, as it is likely to be more costly to provide another slave to the market than it was to capture and deliver the previous slave to the market. That is, the easiest to capture, most productive slaves are captured first. On the margin, the next 100 slaves supplied to the market are likely to be more costly to provide than the previous 100. If it were costless to capture slaves—and there were enough slave providers to generate competition—the price of a slave in equilibrium would be zero because the market would be flooded with slaves. Figure 1.1 illustrates the equilibrium in the slave market. The supply curve is upward sloping because, as just discussed, the marginal cost of capturing and delivering slaves is likely to be rising with the number of slaves in captivity. We will return to the polar case of an infinitely elastic (or horizontal) supply curve for slaves shortly. The demand curve is downward sloping because the demand for slaves is a derived demand, dependent on the amount of output produced by slaves. As is standard in economics—and particularly likely to hold in the agricultural markets where slaves are often forced to work—we assume that there is declining marginal productivity of labor, which necessitates that the demand curve is downward sloping. The equilibrium (meaning the price and quantity of slaves traded that the market tends toward) is determined by the intersection of the supply and demand curves. The equilibrium price is P* and the equilibrium quantity is Q* slaves in captivity. This figure assumes an implicit set of institutions similar to those that were in place in the United States in the antebellum period—there was an active slave market in which slaves were brought to market, and slaves were exchanged at a market-determined price. To focus on a redemption program, the figure abstracts from differences in prices for slaves with different attributes, and treats slaves as homogeneous as far as slaveowners are concerned. Next we introduce a redemption program. We interpret a redemption program as follows: An outside organization steps in and purchases the freedom of X slaves, regardless of their price. After they are freed, the

12

KARLAN AND KRUEGER

Price of Slaves

Supply

P*

Demand Q*

Quantity of Slaves Figure 1.1 Demand and supply for slaves absent a slave redemption program.

redeemed slaves may be instantly recaptured, or they may work as free men and women. If they are instantly recaptured and brought to market at zero cost, then there is an infinitely elastic supply of slaves (which is inconsistent with figure 1.1). If, however, some of the liberated slaves retain their freedom, and the cost of acquiring new slaves and bringing them to market is higher than the cost of supplying the last slave to the market prior to the redemption program, then the equilibrium price will rise and the number of slaves in captivity will decline. The effect of a redemption program is illustrated in figure 1.2. The slave redemption program purchases X slaves, regardless of their price, and sets them free. This causes the demand curve to shift horizontally to the right by X slaves, to D
. The supply curve reflects the marginal cost of providing slaves to market, including any freed slaves who are recaptured. The equilibrium number of slaves purchased (including both slaves in captivity and those redeemed) is Q
. Notice that Q
is less than Q*  X, where Q* is the number of slaves in captivity absent the redemption program. This implies that the number of slaves in captivity has declined. The number has not declined by X, however. The reason is that the quantity supplied increases. The number of slaves in captivity declines by   Q
 Q*  X. The key assumptions for this result involve the slope of the supply and demand curves. If the supply curve were flat, instead of upward sloping, the number of slaves in captivity would be unchanged, despite the redemption program. Economists measure the responsiveness of supply by the elasticity of supply, or percentage change in supply for a percentage change

13

SIMPLE ANALYTICS OF SLAVE REDEMPTION

Price of Slaves

Supply

P’ D’ = New Demand

P*

D = Demand Q*

Quantity of Slaves

Q’ Q* +X X

Figure 1.2 Demand and supply for slaves with a slave redemption program.

in price. A flat supply curve corresponds to an infinite elasticity of supply. So a key concept is the supply elasticity. A redemption program will be more successful in situations where the supply curve is less elastic. It will be completely ineffective if the elasticity of supply is infinite. In the section “Cost Function” we will discuss how to think about determining whether the elasticity of supply is infinite.1 Notice another implication of the model: the price of slaves traded will rise with a redemption program if the supply elasticity is finite. This can be monitored. Thus, a more successful program will raise the price of slaves. This should be viewed as a positive result, as a higher price will discourage current slaveowners from owning slaves; they would rather sell them to the redeemers. Ironically, the consequence of “raising the price of slaves” is often cited as a reason to oppose slave redemption, yet a higher price is exactly the sign one would expect if the intervention has reduced the number of slaves held in captivity; indeed, it is the mechanism by which slave ownership is reduced. The slope of the demand curve is also relevant. The slope of the demand curve reflects the responsiveness of slaveowners to a change in the cost of acquiring slaves. If the demand curve of slaveowners were a vertical line, 1 If the demand curve were vertical, or perfectly inelastic, then the redemption program would not affect the number of slaves in captivity either. The reason for this is that demand is not sensitive to price. We suspect an inelastic demand curve is unlikely, however, in the sectors where slaves are commonly forced to work.

14

KARLAN AND KRUEGER

instead of a downward-sloping line as shown in figures 1.1 and 1.2, then the number of slaves purchased by slaveowners would be independent of the price, and the slave redemption program would not reduce the number of slaves in captivity. Of course, it is possible that the organization that is redeeming slaves might be swindled in that the person who is released was not actually a forced worker. Human Rights Watch reports on anecdotal evidence of children being rented or borrowed to be exchanged for cash by unscrupulous middlemen. The number of freed slaves, X, may be smaller than the number of people that a slave redeemer believes has been set freed. Certainly such fraud is costly from the slave redeemer’s perspective, but it does not alter the analysis. It simply means that the net effect, in terms of slaves freed, is less per dollar spent. As long as everyone freed under the redemption program is not a “rented” individual who was not a genuine slave, and supply and demand are not perfectly inelastic, the net effect will still be a reduction in slaves in captivity. Another way of thinking about the “fake slave” problem is to consider it a pro-rata increase in the price of freeing a true slave. If 50 percent of slaves redeemed are “fake slaves,” then instead of paying $500 per freed slave, the effective price is $1000 per freed slave. This model undoubtedly misses much, but it provides a benchmark economic model against which to evaluate slave redemption plans. We suspect a number of relevant institutional features can be interpreted in the context of the basic supply-and-demand model outlined here. For example, institutional, geographic, and political aspects of the environment may raise or lower the elasticity of supply of slaves. In addition, the model helps one to evaluate claims about the success or failure of slave redemption programs. For example, Human Rights Watch’s concern that slave redemption may cause increased raiding can be interpreted as a movement up the supply curve, from Q* to Q
. It would not be surprising if slave redemption caused additional people to be captured. But the purchasing and liberation of slaves could still reduce the amount of slaves in captivity, even if raiding increases as a result. The two key variables to monitor to judge the effectiveness of slave redemption are the total number of slaves in captivity and the price of slaves.

Extension: Scarring Effects The analogy of redeeming slaves to repurchasing high-emission cars fails when one considers the longer term damage from ever having been a slave. Anecdotal evidence, which is all that is available, suggests that the scarring from some forms of slavery can adversely affect former slaves

SIMPLE ANALYTICS OF SLAVE REDEMPTION

15

long after they are liberated. For example, Nicholas Kristof (2005) provides a wrenching report on two forced child prostitutes whose freedom he purchased, one of whom was unable to readjust to life in her original village and voluntarily returned to prostitution. Bales (2004) provides much anecdotal evidence suggesting that many ex-slaves lead shattered lives, and often return to bondage. The fraction of freed slaves who are unable to adjust to their freedom, or whose lives are reduced to a shadow of what they otherwise would have been had they not been forced into slavery, is unknown but probably substantial. The well-being of freed slaves is relevant because an unambiguous result of the theoretical model outlined in the previous section is that the number of people rotating through the state of slavery will increase under a slave redemption program. Thus, the number of people who were ever forced into slavery would be increased in a slave redemption program. If being forced to work as a slave, even briefly, leaves a permanent scar on individuals, then the purchase of slaves’ freedom can lower welfare.

Cost Function: The Origin of the Supply Curve The prior analysis suggests that an estimate of the elasticity of supply is essential for estimating the likely impact of a slave redemption program on the number of slaves held in captivity. The supply elasticity depends critically on the institutional setting and the particulars of slavery in a given society. In a competitive market, the supply curve is simply the marginal cost of capturing and delivering slaves to the market. The cost structure of supplying adult slaves in the Sudan for agricultural or household labor clearly is different than that of supplying child prostitute slaves in Southeast Asia, and different than that of supplying indentured servitudes “voluntarily.” We attempt here to provide some examples and structure to organize how one might think about estimating the relevant cost curve. We will discuss costs for two types of suppliers, households (e.g., indentured servitude, where a household sells a family member in order to satisfy a debt) and slave entrepreneurs (e.g., a slave trader who captures individuals by force and sells them to others for farm, domestic, factory, or prostitution labor). We start with the case of the household. The cost of supplying an additional household member to the slave market is the sum of the lost stream of revenue (through either household production, wages, or self-employment income) from that individual, plus the emotional cost of selling that individual and losing his or her company. The moral, emotional, and/or social stigma costs clearly are harder if not impossible to quantify, and for the sake of analysis we assume this to be

16

KARLAN AND KRUEGER

linear in the quantity supplied (i.e., the same emotional cost to the families per slave provided, regardless of the total number of slaves supplied). This implies the marginal costs do not increase as supply increases. Regarding lost income, however, at a constant price it is likely that the least productive family member would be sold first, then the next most productive, and onward. The same holds when thinking about this across households: families with the least productive children (responding merely to economic incentives) will sell their children first, then those with the next most productive, and onward. This implies that the marginal cost of supplying a slave increases with the number of slaves supplied. To estimate the slope of the supply curve, one could estimate the person-by-person productivity of individuals within a household, or the productivity of individuals across households. For household production, (i.e., a farm or small enterprise), the marginal productivity of each additional worker likely decreases as we hold the physical capital in the household fixed. Next, consider the cost function of an entrepreneur who supplies slaves. What are the costs of capturing a slave? First, if the entrepreneur buys slaves from households, then the analysis for the household applies here as well: firms will have to compensate households, and households will charge more for each person, assuming the marginal productivity of labor in the household increases with fewer members in the household (holding physical capital fixed). The slave supplier’s costs also might include bribes to local officials, wages to thugs hired to capture the slaves, guns and ammunition to enforce the capture, and so on. An analysis of the cost structure of such a firm would require detailed information about the constraints to growing the operation. Given the lack of institutions to enforce illegal contracts, it is likely that such operations can be only so large before it is impossible for a single owner to avoid being robbed by his own employees. So as illegal firms grow, it is plausible that the costs of supplying the next slave rise. These considerations suggest that the supply curve in many settings is unlikely to be perfectly elastic, but the precise elasticity of supply will depend on the particular circumstances.

Conclusion A slave redemption program cannot abolish slavery by itself, but it can cause a reduction in the number of slaves in captivity. If the supply elasticity of slaves to the market is finite, a slave redemption program would be expected to reduce the number of slaves in captivity and raise the price of slave ownership. By artificially increasing demand for slaves, and then

SIMPLE ANALYTICS OF SLAVE REDEMPTION

17

setting a portion of them free, a slave redemption program will drive up the price of slaves supplied to the market and discourage some slaveowners from holding slaves as long as the elasticity of supply is finite. To assess the merits of these arguments, one would need two types of data: the price at which slaves are exchanged (ideally before and after redemption programs) and the number of slaves in captivity (again, ideally before and after a redemption program). Additionally, data allowing estimation of the shape of the cost function for supplying slaves to the market would provide useful information for determining whether a slave redemption program is likely to be efficacious by enabling an estimate of the elasticity of supply. Monitoring prices of slaves before and after a slave redemption program is implemented provides an indirect indication of the effectiveness of the program. If the price is unchanged, then either the program was too small to have any detectable effect or the supply curve was very elastic. Unless the slave market is organized and open, as it was in antebellum America, for instance, monitoring prices may be very difficult. Still, even if data on prices are unavailable, it may be possible to collect data on costs of capturing and delivering slaves to slaveholders. As discussed above, the relevant costs depend entirely on the institutional setting of the slave market in the country of interest. For those settings in which households decide, due to abject poverty or other reasons, to sell their kin into slavery, detailed data on household production might provide some insight into the shape of the supply curve, and the likely effectiveness of a redemption program. Of course, slave redemption could be considered morally repugnant because it requires the purchase of slaves, which could be viewed as condoning the slave trade, and it transfers resources to those who capture and sell slaves. But, just as paying cash for high-emission cars does not indicate approval of air pollution, we think a case could be made that purchasing the freedom of slaves is a noble act. However, one would want to be sure that the net effects of the program are positive, and that the limited resources that are available for slave purchases are used in the situations where they have the greatest positive impact for welfare. The simple model we provided gives a framework for considering the effects and effectiveness of slave redemption initiatives. Another consideration is whether slave redemption diverts attention, energy, and resources of the international community from alternative means of reducing or abolishing forced labor. Slave redemption is clearly a short-term solution, where the ultimate goal is the complete abolition of slavery. Slave redemption programs can help to make slavery uneconomical for slaveowners by driving up the cost of slaveownership, however, and therefore help to hasten its end.

18

KARLAN AND KRUEGER

Slave redemption programs do not solve the roots of the problem, which often include both poverty and poor institutional protection of human rights. Slave redemption programs could be combined with implementation of legal and political institutions that guarantee continued emancipation. This would be a solution similar to that proposed by several founding fathers of the United States, such as James Madison, who proposed: “A general emancipation of slaves ought to be—1. Gradual. 2. Equitable, and satisfactory to the individuals immediately concerned. 3. Consistent with the existing and durable prejudices of the nation” (Madison 1819). Madison’s proposal (obviously not adopted), which may appear morally objectionable to many in hindsight, was potentially appealing to slaveowners because they would have received some compensation for their slaves. It is impossible to know whether slavery might have ended sooner, and without incurring the tremendous costs of the American Civil War, had a Madisonian approach been pursued. Few if any slave markets today resemble the organized slave market that occurred in the United States before the Civil War, so it is questionable whether a transition plan enacted through legal channels could work in a setting such as the Sudan, for example. Nevertheless, in some settings redemption programs might reduce the net number of slaves in captivity, even after accounting for a likely movement up the supply curve, because an increase in price resulting from an external increase in demand (due to the slave redemption program) will discourage some slaveowners from purchasing or holding slaves. Nevertheless, from the standpoint of the international community, a redemption program creates a trade-off between the flow and the stock of slaves: a redemption program probably reduces the stock of slaves in captivity at any point in time, but increases the flow of slaves in and out of slavery. Any scarring effect of time spent in the state of slavery upon freed slaves should be weighed against the salutary effect of a reduction in the number of slaves in captivity. Bibliography Bales, Kevin (2004), Disposable People: New Slavery in the Global Economy, Berkeley: University of California Press. Dill, Jennifer Lynn (2001), “Travel Behavior and Older Vehicles: Implications for Air Quality and Voluntary Accelerated Vehicle Retirement Programs,” Ph.D. dissertation, University of California, Berkeley. Eltis, David (1993), “Europeans and the Rise and Fall of African Slavery in the Americas: An Interpretation,” American Historical Review 98: 1399–1423. Human Rights Watch (2002), “Slavery and Slave Redemption in the Sudan,” Human Rights Watch Backgrounder, March 2002, www.hrw.org/backgrounder/ africa/sudanupdate.htm.

SIMPLE ANALYTICS OF SLAVE REDEMPTION

19

Kristof, Nicholas (2005), “Back to the Brothel,” New York Times, 22 January 2005, A15. Lewis, Paul (1999), “U.N. Criticism Angers Charities Buying Sudan Slaves’ Release,” New York Times, 12 March 1999, A7. Madison, James (1819), “Letter to Robert J. Evans, author of the pieces published under the name of Benjamin Rush,” 15 June 1819. Randall, Williard Sterne (2003), Alexander Hamilton: A Life, New York: HarperCollins.

Chapter Two Slave Redemption When It Takes Time to Redeem Slaves CAROL ANN ROGERS AND KENNETH A. SWINNERTON

The main purpose of this chapter is to analyze the impact of slave redemption programs in the context of an economic model in which it takes time to find slaves to redeem. The model discussed in the previous chapter by Dean Karlan and Alan Krueger overlooks this kind of market friction. While many of the insights of the model they present do not change in our model, some do, particularly those concerning what happens to the price at which slaves are exchanged. For example, in the Karlan and Krueger framework, a redemption program causes the price of slaves to rise at the same time that it reduces the number of slaves. A rise in the price of slaves necessarily indicates that the number of slaves has fallen. In the model we discuss, a fall in the number of slaves, ceteris paribus, can be accompanied by an increase, decrease, or no change in the price of slaves. Therefore, the behavior of the price of slaves is not a sufficient indicator of what has happened to the number of slaves, nor need it convey information about other factors that might affect the welfare of slaves. This chapter unfolds in four sections. First we briefly review the supply-and-demand model, highlighting the features that will be compared and contrasted with results of our subsequent analysis. Next we present our analysis of slave redemption in a “matching” model that allows sellers and buyers of slaves to take time to find each other. Both the supplyand-demand and matching models suggest reason to be concerned that a redemption-induced reduction in the number of people remaining in slavery comes at the expense of causing more people to experience some slavery, at least temporarily, and so the costs of experiencing some slavery are important to a welfare evaluation of redemption programs. Our third section discusses this issue at some length. The final section offers concluding remarks. A Brief Review of the Supply-and-Demand Framework Figure 2.1 is essentially the diagram that Karlan and Krueger use to motivate their discussion. The supply curve shows how the number of episodes

21

IT TAKES TIME TO REDEEM SLAVES

Price of Slaves

Supply of episodes of slavery

PB

PA

Demand for slaves (including demand from redeemers)

Demand for slaves by non-redeemers C

A

B

Quantity of Slaves Figure 2.1 Redemption in a supply-and-demand framework.

of slavery that occur depends on the market price of slaves. The innermost downward-sloping curve is the demand curve from nonredeemers. This curve shows how the number of slaves held by those who demand slaves for use as slave labor depends on the market price of slaves. Before the institution of a slave redemption program, the equilibrium price of slaves is PA, and the number of episodes of slavery is the same as the number of people demanded and enslaved, i.e., the quantity at point A. The redemption program demands BC slaves, so that the total demand for slaves becomes the demand generated by those who buy slaves with the intention of using them as such, plus the demand of the redeemers who intend to free the slaves they buy. The market demand curve becomes the outermost downward-sloping curve. The supply of slaves must be at least equal to the number of slaves actually bought, and so even though the redemption program leads to a decline in the number of people held in slavery to point C, it also must lead to an increase in the number, to point B, of people who suffer an episode of slavery. In this framework, it must also lead to an increase in the price at which slaves are exchanged, i.e., from PA to PB. A Matching Model of Slave Redemption The supply-and-demand model does not explicitly discuss the factors that affect the time it will take to achieve the predicted reduction in the

22

ROGERS AND SWINNERTON

number of slaves or the time spent in slavery by the additional people who endure an episode of slavery due to the redemption program. For instances where it takes time to complete a transaction, economists turn to “search” or “matching” theory (see, e.g., Pissarides 1990). This theory has been used, for example, to analyze labor markets—how long it takes someone to find a job; marriage matches—how long it takes someone to find a spouse; and real estate markets—how long it takes someone to find a house. It has also been used to explain how time affects the precise terms on which these deals are consummated, e.g., a worker’s wage, a marriage dowry, or a home price. Since it likely does take time for sellers and buyers of slaves to find each other, we think it is natural to analyze the market for slaves in a matching model. As we shall see, some of the predictions of the supply-anddemand model continue to be true in such a framework. However, our simple matching model produces two new insights. The first is that the increase in the number of episodes of slavery can occur even if there is no increase in the price of slaves. The second is that unless there is an increase in the price of slaves, redemption programs do not secure the freedom of any person enslaved when the program began. Absent a price increase, redemption programs can at best slow the flow of new slaves into slavery. There are two key features to a matching model. One determines the likelihood that a buyer meets a seller in a fixed period of time. The other addresses how they settle on the price at which they will exchange a slave when they do meet. In building our model, we need to allow for two types of buyers. One type buys slaves for the purpose of using them to produce some good or to provide some service. We refer to these as buyers. The others are redeemers who buy slaves to give them their freedom. We also allow for two potential types of sellers: slaveholders who use slaves for the production of some good, or the provision of some service; and, slave raiders who capture slaves with the express intention of selling them to someone else. Let us suppose that the value that a slave can provide for anyone who uses the slave to produce a good or service is z. Now consider what happens if a buyer and a slaveholder come together to negotiate an exchange of a slave. The slave may be exchanged if and only if the price is z. The slave produces z for the slaveholder, so the slaveholder will not let the slave go for any less. The slave would only produce z for the buyer, so the buyer would not pay any more. We note that at the price of z the slave need not be exchanged: both the buyer and seller are strictly indifferent about making the exchange. Nevertheless, unless the price is z, at least one or the other would not even consider making the exchange. Let us now turn to exchanges that involve slave raiders. We will assume that each slaveraider deals with only one slave at a time, and that

IT TAKES TIME TO REDEEM SLAVES

23

each buyer and redeemer seeks to buy just one slave. These assumptions simplify the exposition of the matching model, but are not essential to our conclusions, which remain valid if more than one slave can be exchanged at a time. The first thing a slave raider needs to do is to capture a slave to sell. Adopting the assumption that generated the upward-sloping supply curve in the supply-and-demand model, we assume that the cost of capturing a slave is a function of the aggregate number of slaves offered by slave raiders, and that this cost increases with the number offered. Call this function C(S). Note that this cost is “sunk” at the time of capture. After securing a slave to offer, a slave raider must find someone to buy the slave. For the moment, we assume that there are no redeemers and just concentrate on how a slave raider achieves an exchange with a buyer. We avoid a discussion of the many ways that a slave raider may search for a buyer and instead follow the convention in matching models of positing that in any given period of time, a slave raider has some probability of success, q, of finding a buyer. This probability is related to the level of aggregate activity in the market for slaves. To capture this, we posit the existence of a function, q  B  where B stands for the total  S number of buyers searching for one slave each and S for the number of slave raiders looking for buyers. The probability that an individual slave raider is matched with a buyer is thus a function of the ratio of the aggregate number of buyers to slave raiders in the market. It is natural to assume that as there are more buyers per slave raider, the probability that a B increases. slave raider will find a buyer goes up, so that q increases as S Nevertheless, a slave raider is never absolutely certain to find a buyer in a given period of time, so q is always strictly less than one. The average time it takes a slave raider to sell a slave is related to the probability, q, that a slave raider finds a buyer in a given period of time. To see how, suppose that q is the probability of finding a buyer in one month, and that this probability is one in 12. With the monthly probability of meeting a buyer at 1 , the seller expects to wait 12 months to find a buyer. 12 As this example implies, generally, the expected time to sale of a slave is the inverse of the probability of sale; therefore, from the time that a slave 1 raider secures a slave to sell, he expects it to take periods to find a buyer q for that slave. When the raider and buyer meet, they must negotiate a price at which the slave will be exchanged. Assuming the slave raider incurs no perperiod costs to support a slave, so that there are no costs to be saved by unloading the slave, the payoff to the slave raider from selling a slave is

24

ROGERS AND SWINNERTON

the price, p, at which the exchange occurs. We could include a cost savings from unloading the slave, but doing that would only make our mathematical expressions more complex without changing the nature of our analysis. Note as well that the cost of capture, C(S), does not figure into the slave raider’s payoff. Since that cost was incurred at the time of capture, it cannot be saved. For the buyer, the payoff from making the exchange is the value earned from using the slave to produce a good or service less the price paid to purchase the slave, i.e., z
p. Clearly, the buyer will not deal if the price of the slave exceeds z, and so the price bargained must be between zero and z. The precise price bargained depends on the relative bargaining strength of the two parties, and on assumptions about the bargaining process. In economics, it is typical to assume that bargaining power is set and is known to the negotiating parties, and that the price ultimately bargained is Pareto-efficient in the eyes of the two parties to the bargain. This means that there is no other price that would be preferred by both parties. The well-known Nash bargaining problem leads to Pareto-efficient bargains. In our setting, the Nash bargaining problem may be stated as: Choose p to maximize [ z − p] [ p]1− where the parameter
indexes the relative bargaining strength of the buyer and 1

is the strength of the seller. We assume that all buyer and seller pairs have exactly the same distribution of bargaining power within their negotiation, although this assumption can be relaxed without changing the nature of our discussion. Note as well that the function maximized is factored into (functions of) the buyer’s and the slave raider’s payoffs. Letting p* be the solution to our Nash bargaining problem, it is straightforward to show that p* = (1 − )z. To see how intuitive this solution is, notice that p* is lower the higher is
, the relative bargaining power of the buyer. In particular, note that if the buyer has all the bargaining power (
 1), p*  0; while if the slaveraider has all the power (
 0), p*  z. A critical feature of the solution for our later discussion is that so long as the buyer exercises some bargaining power (
 0), p* is less than z. We can now discuss how the number of slaves captured is determined. When deciding whether to capture a slave, a slave raider expects 1 to receive a payoff of p* at periods in the future. Suppose an earlier q

25

IT TAKES TIME TO REDEEM SLAVES

pay off is preferred to a later one: payoffs received in the future are discounted at rate i. This means that a payoff of p received one period in the  1  future would have a present value of p   , a payoff of p received two 1+ i 2  1  p periods in the future would have a present value of  1 + i  , and so forth. Since the expected amount of time until a slave raider is matched with a buyer and receives payoff p* is equal to 1 , the expected value  B q   S of this payoff at the time the raider captures the slave is: 1

 1  q B  Expected present value of selling the slave = p *   S .  1 + i  Let us simplify this expression by using the notation F(B/S) to stand for 1

 B the discount factor  1  q S  , so that the expected present value of selling  1+ i the slave can be written as p*F(B/S). We note that this present value increases when the ratio of buyers to sellers increases. That is because the larger is this ratio, the higher is the probability that a seller finds a buyer in a given period of time, and the shorter is the length of time to a sale. The expected present value is the same as the expected benefit, as of the time of capture, from capturing the slave. So long as this expected benefit exceeds the cost of capture, C(S), raiding is expected to be a profitable endeavor, and more slave raiders will want to undertake the activity. In fact, absent binding restrictions on raiding activities, additional raiding will occur until the expected benefit of selling a slave is equal to the cost of capture—that is, until

 B p * F   = C(S).  S

(Equilibrium condition, no redeemers)

Now let us assume that R redeemers enter this market. For the moment, let us also assume that they bargain in precisely the same way as other buyers, so that there is no change in the bargained price, i.e., it

26

ROGERS AND SWINNERTON

stays at the same value p*. With redeemers, the equilibrium condition becomes:  B + R = C(S) p*F   S 

(Equilibrium condition witth redeemers)

Using the two alternative equilibrium conditions, we can produce a diagram that looks somewhat like figure 2.1, but with a different interpretation. In figure 2.2, we have graphed the equilibrium conditions as functions of S. The upward-sloping line is the function C(S). The innermost downward-sloping line is the expected value to a slave raider of selling a slave  B when there are no redeemers in the market, p * F   . The expected  S value line slopes downward because for a given number of buyers, an increase in slave raiders selling slaves lowers the ratio of buyers to slaves to be sold, thereby lowering the probability that a seller finds a buyer in a fixed period of time, and increasing the length of time to sale. Since sales farther off in the future are worth less today, the expected value of making a sale falls when there are more slave raiders trying to sell in the market. When there are no redeemers looking for slaves to redeem, equilibrium occurs at point A. If redeemers enter the market with the intention of redeeming a positive number, R, of slaves, the expected value of selling a slave becomes  B + R . This expected value is represented in figure 2.2 by the outp*F  S  ermost downward-sloping curve. Redeemers essentially shift the expected value curve upwards. When redeemers enter the market, there are more buyers per seller, making it easier to sell a slave in a given period of time. So the duration to sale falls, and the expected value of a slave rises. Unlike in the supply-and-demand framework, the two expected value curves are not parallel, and the horizontal distance between them is not equal to R. This is because the shift in figure 2.2 does not represent a direct increase in the demand for slaves, but rather the effect of the redeemers 1 on the length of time it takes to sell a slave, i.e., on . Redemption activity q reduces the duration of time to a sale, but this fall does not affect the expected value of sale by the same amount at each level of S. This means that we will not be able to read the number of slaves that remain in slavery after redemption programs directly off the figure. What we can read from figure 2.2 is the effect of redemption on the quantity of slaves captured who are either destined for buyers or for redeemers. A redemption program causes an increase from A to B in the

IT TAKES TIME TO REDEEM SLAVES

27

Figure 2.2 Redemption in a matching model.

number of people who suffer slavery, similar to the effect seen in figure 2.1. The move from A to B represents the increased capture of slaves due to redemption, and it represents an increase in the total number of slaves available for sale. We can also see from figure 2.2 that redemption programs cause the expected value from the sale of a slave to increase from EBA to EBB. Because we have assumed the price of slaves to be fixed in all instances at p*, this is of a similar spirit to, but not the same as, the price effect shown in figure 2.1. In this example, the expected value goes up because redemption programs make the sale of slaves happen more quickly, not because of changes in the price of a slave. In addition to the obvious contrast with the prediction about price from figure 2.1, this point may be of some practical relevance because one way in which redemption programs have been defended against concerns that they are having some negative impact is to point out that their activities have had no upward effect on the price of slaves exchanged. Jok Madut Jok (2001, 175) notes from his experience observing slavery in Sudan that the price at which slaves are exchanged remained at around $50 after the entry of international redeemers into the market. He also (176) cites one redeemer working in Sudan on behalf of Christian Solidarity International (CSI) who says he will not pay more than $50 for a slave, and will stop redeeming if the practice evolves into a “free market.” Finally, Jok reviews arguments that suggest that local redeemers

28

ROGERS AND SWINNERTON

(e.g., families, friends, and neighbors) of those captured would resist the activities of the outside redeemers because it would raise the price they must pay. He finds that instead the local redeemers welcome the help from outsiders (178). These arguments all turn implicitly on the notion that absence of an increase in the price of slaves is a good indicator that redemption programs are having no unintended negative impacts. But in our model, negative impacts have no necessary implication for price, and so price is not a relevant indicator for assessing whether slave redemption programs are having some negative impact. In the context of our model, we can make an even stronger point. Suppose redeemers determine the price they will pay for slaves by observing the average selling price in the market, and decide to make take-it-orleave-it offers of just that price. Before redeemers enter the market, slaves can be exchanged at two prices, z, and p*, which is strictly less than z. The average of these two prices must be strictly less than z. So the price a redeemer will pay must necessarily be less than z. If this is how redeemers operate, they will never be able to buy back slaves actually being used in production. Only a slave raider would sell a slave at a price below z. In order to buy from slaveholders, the redeemer therefore needs to be prepared to raise the price that he will pay to z or higher. Certainly, the assumption that we made that each slave produces z for a slaveholder drives the result that no slave in production will be redeemed unless redeemers are prepared to pay higher than the observed market price; however, the more general point is that some slaves will be worth more than the observed price to their slaveholders, and redeemers will never be of any help to these slaves unless they are willing to pay higher than that price. It is also worth noting that were redeemers to offer a price of slightly more than z, the effect might not be to induce more slave raiding. A price of slightly more than z would induce slaveholders to join slave raiders as willing sellers in the market. This would have the effect of lowering the ratio of buyers and redeemers to sellers and therefore of lowering the probability that a seller makes a sale. Unless the higher price paid by redeemers is high enough, the expected value of a sale could fall, relative to that in our prior equilibrium with redeemers, and even with respect to that observed in the equilibrium without redeemers. In the latter instance, redemption rescues existing slaves while creating an incentive to capture fewer new slaves. An increase in price in this instance implies that redemption does not cause harm, and in fact is a positive indicator of successful slave redemption programs. Returning to the assumption that redeemers pay p*, we see that the only effect that redeemers have is on the flow of new slaves into slavery. With figure 2.2 we have already shown that redeemers’ presence increases the capturing activities of slave raiders. We have yet to determine how many of the extra captured slaves go to buyers, and thus remain in slavery,

IT TAKES TIME TO REDEEM SLAVES

29

and how many are redeemed. While figure 2.2 does not aid in making this determination, it is possible to use the equilibrium equation on which this figure is based to determine the elasticity of sales by slave raiders with respect to the number of redemptions. An elasticity is the percentage change in one variable (sales by slave raiders) with respect to a change of one percent in another (number of redemptions). If the elasticity of sales by slave raiders with respect to redemptions is greater than one, then more slaves are sold than are redeemed, implying that some of the slaves who are captured because of the redemption program must remain in slavery. If the elasticity is less than one, then more slaves are redeemed than were captured, implying that redemption ends more episodes of slavery than it creates, and the number of people left to flow into slavery falls. The equilibrium equation (with redeemers) can be shown to imply that the elasticity of slave raiding with respect to redemptions is less than one. In terms of figure 2.2, this means that while redemption programs induce AB more episodes of slavery, they redeem more than AB slaves. Intuitively, this is for two reasons. The first is that since capturing an additional slave raises the cost of capturing still more, there is a natural brake on the number of new episodes of slavery. The second reason is that the redemption programs raise the opportunity cost of holding onto a slave to look for a buyer because the expected value of selling a slave decreases as the date to sale is put off. Because rising costs of capture tend to discourage more capture, and because rising opportunity costs tend to discourage waiting for buyers, the net effect of slave redemption programs is to cause the number of people sold to buyers to fall, even though the programs cause more people to be captured by slave raiders. From the perspective of the people who suffer through slavery, the important social welfare trade-off to evaluate is whether reducing the number of people who enter into longer-term slavery is worth exposing more people to the capture that creates the supply of potential slaves. Two questions seem relevant to the evaluation of this trade-off. The first is, how long are the redemption-induced captives held by slave raiders? The answer to this is that the time to redemption for any slave redeemed should be, on average, 1 periods. The second is, what is the nature of the experience of an q eventually redeemed captive while waiting to be redeemed? We discuss this question in the next section. Costs Associated with Redemption-Induced Episodes of Slavery Unless the price at which slaves are exchanged rises so high that all slaveholders would rather sell their slaves than hold on to them, both the supply-and-demand and the matching models unambiguously predict

30

ROGERS AND SWINNERTON

that the benefit of slave redemption in terms of lowering the number or stock of slaves comes with a cost of exposing some people to slavery who otherwise would not have been exposed. In the matching model, we saw explicitly that these redemption-induced slaves will suffer some time in slavery. So their experiences during this time, that is, the costs imposed upon them, are relevant in evaluating the benefits against the costs of slave redemption. Much of the balance in this evaluation should turn on those costs that are imposed on everyone who experiences some time as a slave, at or near the time they enter slavery. If these “up-front” costs are particularly high, redeemers should be especially cautious about the risk of inflicting them on more people. However, from our reading of a variety of stories about how people come to be slaves, it is clear that there is no universal argument to be made that the risk of these up-front costs should be of overriding concern. They should probably be given very high weight in some contexts but little in others. We discuss three examples that illustrate this point. The first example comes from the Sudan, in the mid-to-late 1990s. A number of sources (e.g., United Nations General Assembly 1995 and 1999; Jok 2001; Report of the International Eminent Persons Group 2002) established the existence of slavery and a slave trade in that country and at that time. All deal to some extent with the way slaves enter slavery, but Jok (2001, 33–40) paints the most vivid picture. Women and children were taken in raids on Dinka (a southern Sudanese tribe) villages. They may have been raped, beaten, or both. Often the men in the villages—the fathers, sons, and brothers—were killed. Livestock, the main productive asset, was stolen. Villages were burned to the ground. The women and children endured an arduous trip to the place where they took up their slave duties. More beatings and rape occurred along the way. All of this harm occurred just at the very beginning of someone’s time as a slave: harm to their own physical and mental welfare, loss of family members, and destruction of livelihood and homes. If just becoming a slave involved these harms, then slave redemption, if it did increase the number of episodes of slavery, was responsible for the imposition of these costs on some people. In fact, in the case of modern-day Sudanese slavery the physical harm done at the beginning may be the bulk of the physical harm inflicted (Jok 2001, 63): It is easy for an Arab slave owner to avoid criticism as well as . . . free his conscience if he can create a situation for the Southerners displaced to the North where they have no choice but to accept the exploitation of their labor. The slaveholders have attempted to justify keeping

IT TAKES TIME TO REDEEM SLAVES

31

slaves by claiming it to be an act of mercy. . . . The slavers are often convinced that they are doing a favor for the captives; they regard the Dinka culture as inferior and believe that these Southerners are fortunate to have been incorporated into a superior culture. This is the kind of explanation that has caused some historians to describe early slavery in the Muslim world as benign. Some writers are doing the same regarding present-day slavery in Sudan. A courteous treatment of slaves, undoubtedly, makes slavery more acceptable to the northern society and sometimes to the slaves who were caught at a very young age and incorporated into northern culture and religion. But this does not make it less than slavery. A slave is a slave. Moreover, the current Sudanese slavery is less an economic practice than a cultural project, because there are many poor Baggara who hold slaves that often live no worse than their master. Inasmuch as this quotation suggests that the daily life of being a slave once one has been a slave for awhile may be less harsh than what one suffers early in one’s tenure of being a slave, it suggests a reason to be more worried about the costs that may be inflicted on redemption-induced slaves, than the benefits produced by securing the saving of other slaves. Slave redemption did occur during this time in Sudan, and it spurred great controversy. Some of the controversy was specifically about whether or not redemption programs encouraged more raiding for slaves (see, e.g., Miniter 1999, and Human Rights Watch 2002). In response to concerns that they did or could, one group of redeemers, Christian Solidarity International (2002) issued a statement that said, in part: The fact is that since 1998, while CSI has redeemed increasing numbers of slaves, the number of slave raids and the number of women and children taken into bondage have diminished. This trend has been confirmed by the independent UN Special Rapporteur for Sudan, Gerhart Baum[,] in his most recent report to the UN Commission for Human Rights. What the CSI statement does not acknowledge, but what is clear from reading many of the other already cited sources, including those from the United Nations, is that there were a variety of other initiatives taking place in the Sudan at the time aimed at reducing raiding, and it is not at all clear that CSI is justified in the causal implication this passage suggest. Returning to the view from the model of the previous section, the fact that redeemers apparently did not cause an increase in the price of slaves suggests another reason to be skeptical that redemption led to a decrease in raiding activities. We tend to share the view of Jok (2001, 175) that there is no empirical evidence to evaluate the true impact of slave redemption

32

ROGERS AND SWINNERTON

on the incidence of Sudanese slave raids. In light of this, because the raids in general are so violent, and because economic theory suggests that redemption should lead to more of them, we would counsel extreme caution to redeemers in situations such as the one discussed in Sudan. People who become slaves do not always enter slavery through violent raids or abductions. Our second example concerns trafficking in human beings. Trafficking, particularly in children and women, is a well-known and global phenomenon that often leads to slavery. But during the time of traffic, it may not be particularly harmful. This is because victims of trafficking are often tricked into willingly going with their trafficker (see, e.g., Bales 1999; Rogers and Swinnerton 2004). The victim leaves her home on the promise of a better life—a good job or better schooling—in some far-off destination. But when she arrives to her ultimate destination and is enslaved, she finds her ability to leave severely restricted: sometimes violently, but perhaps also or instead through psychological manipulation or the theft of necessary travel documentation. It is easy enough to recast our model as one of trafficking if we relabel our model’s “slave raiders” as “traffickers.” Then we see that, at worst, redemption increases the flow of people trafficked, but stems the flow of people who actually make it to the final destination of slave. In most of the stories we have read, the victim does not realize she has been tricked until she reaches the final destination, that is to say, after she has left her trafficker, and the horrors recounted occur from this point on. Rarely have we seen an explicit or detailed account of treatment of a victim while being trafficked. But precisely because the conditions endured during trafficking are not addressed by those concerned with the phenomenon, it seems possible that, at least compared to the way people become slaves in Sudan, the treatment is relatively benign, and that taking the risk of encouraging trafficking to stem the flow into slavery may be more palatable in the context of trafficking than in a situation like that in Sudan. Our final example concerns instances when a person enters into a situation voluntarily that either appears to be slavery to others, or may not become “forced” slavery until later in the person’s tenure in that situation. The chapter by Stanley Engerman in this volume gives a number of historical examples of “voluntary slavery.” A recent study of girl and boy victims of sexual exploitation in Mexico (Azaola 2000) gives some examples of children who knowingly, voluntarily, and without any coercion, enter prostitution. A few—willingly, knowingly, and without coercion— choose to do that permanently. Others decide they would like to leave and can do so without being prevented from doing so. Still other are, at the time they want to leave, subjected to a variety of coercive mechanisms that prevent them from doing so. The analysis relevant for the last group

IT TAKES TIME TO REDEEM SLAVES

33

is essentially the same as for the trafficking example discussed earlier. Our concern now is with the case when entry and exit are purely voluntary.1 Again we turn to our model to provide some insights on the risk of redemption programs in these situations. This time, there are no slave raiders, but for theoretical purposes, we can treat a voluntary slave as a self raider. When the perceived benefit they expect to enjoy from entering prostitution exceeds the perceived cost of doing so, they enter. Note that the upward-sloping C(S) curve in figure 2.2 suggests that as more and more victims make themselves available, the cost of purchasing them rises. This is because the latest victims to enter will likely require a higher payoff to do so than the earlier ones. A redemption program increases the number of victims who make themselves available for, but decreases the number who actually enter, prostitution. In this instance, a redemption program provides the victims who participate with a better option than prostitution. Since the offering of this option causes more people to choose to leave than choose to enter, and since “victims” in all instances make choices, we may be least concerned about “up-front” costs associated with redemption programs in these instances.

Concluding Remarks In this chapter we analyzed slave redemption programs in a simple matching model. Unlike in a simple supply-and-demand framework, where sufficiently large and effective redemption programs must lead to an increase in the price at which slaves are exchanged, we find that such programs have no necessary impact on price. We also demonstrate that a slave redemption program can slow the flow of people into the actual state of slavery, but at the same time can increase the number of people captured to be slaves. We discussed redemption programs in the context of three contemporary examples of slavery. This discussion suggests that the risk of harming someone in the name of helping someone else appears 1 We take a moment to emphasize our intended meaning of the word “voluntary” in this context, and the implication of our definition. By voluntary, we mean that no one has kidnapped or imprisoned the child who becomes or remains a prostitute, and that the child could leave at any time. Some may object to this definition by arguing that the child is “forced” by circumstance: she may have no other choice but to become a prostitute, if she wishes to survive. But then the tragedy is not so much that she has chosen prostitution, as it is that the only choice available to her is between prostitution and death. Whereas legal prohibitions all by themselves can improve the lot of children who are stolen into and imprisoned in prostitution, they will harm the child who is prostituted because she has no better choice, unless better choices are also provided to her. Eliminating voluntary prostitution therefore requires that attention be paid not only to the fact that the child is a prostitute, but also to making sure that she has better choices available.

34

ROGERS AND SWINNERTON

highest in situations where violence or coercion occurs immediately, lower in situations where the individual is not harmed until some later time, and lowest when the individual enters the situation of his or her own accord. The welfare comparisons discussed so far in this paper have all been focused on slaves and potential slaves. There is another welfare issue to consider: redeemers and society at large may be better off if some program can be devised to bring about a reduction in slavery without first, or in part, encouraging more of it to occur. There is likely an endless list of alternative policy options and actors to consider, and all probably deserve at least the same analytical attention we pay here to redemption. We conclude by just listing a few, in the context of our examples from the previous section. We also limit ourselves to suggestions that could conceivably be funded in the same manner as redemption programs have been recently, namely, by appealing to individuals to make voluntary donations in the hopes of helping other individuals. In the case of Sudan, not only is the vulnerable population—the Dinka—identifiable by ethnicity, according to Jok (2001), but so are the slave raiders. They are the Baggara. We suggest two alternative options, one targeted at each group. To target the Dinka, it may be worth considering refugee sponsorship programs that offer them the opportunity to move from the place where slave raiding is likely to occur and resettles them in other countries. To target the Baggara, perhaps a scheme whereby they are paid not to raid could be considered. In instances of trafficking (via trickery) and “voluntary slavery,” a productive solution may lie in finding ways to give vulnerable people better choices so that they do not make themselves available for these situations. For example, if women or children are vulnerable to entering the sex trade because their other economic opportunities are dismal, can something be done to better the other opportunities so that at least one is better than the sex trade? This is in fact similar to a question that is receiving much attention in the area of child labor policy (see, e.g., United States Department of Labor, 2000 and 2002), where, for example, strategies such as subsidies to families who send their children to school and the provision of income-generating assets to families vulnerable to participation in child labor have been devised to try to expand the set of choices from which families can choose, and hopefully, offer one that is better than child labor. In fact, in some instances the question is the same, as some forms of child labor that concern policymakers are in fact practices that many would label slavery (see, e.g., International Labor Organization 1999). The point of this concluding discussion is not to advocate for another option over redemption so much as to point out that there are other

IT TAKES TIME TO REDEEM SLAVES

35

feasible policy options. In some full social welfare accounting, informed not only by economics but also by ethics, the benefits of slave redemption may outweigh the costs. But it certainly is also wise to consider, and more extensively than we have the opportunity to do here, the possibility that some other program might generate equal benefit at lower cost.2

Bibliography Azaola, Elena (2000), Stolen Childhood: Girl and Boy Victims of Sexual Exploitation in Mexico. Mexico City: DIF/UNICEF/CIESAS. Bales, Kevin (1999), Disposable People. Berkeley: University of California Press. Christian Solidarity International (2002), “Christian Solidarity International Statement on Slave Redemption Work in Sudan,” http://iabolish.com/ redemption/CSI-60.htm. Human Rights Watch (2002), “Slavery and Slave Redemption in the Sudan, Human Rights Watch Backgrounder,” http://www.hrw.org/backgrounder/ africa/sudanupdate.htm. International Labor Organization (1999), Convention Concerning the Prohibition and Immediate Action for the Elimination of the Worst Forms of Child Labor (No. 182), 1999, adopted by the International Labor Conference at its 87th session, 17 June 1999, Geneva. Jok, Madut Jok (2001), War and Slavery in Sudan, Philadelphia: University of Pennsylvania Press. Miniter, Richard (1999), “The False Promise of Slave Redemption,” Atlantic Monthly, 284(1), July, pp. 63–70. Pissarides, Christopher A. (1990), Equilibrium Unemployment Theory, Oxford: Basil Blackwell Ltd. Report of the International Eminent Persons Group (2002), “Slavery, Abduction and Forced Servitude in Sudan,” http://www.state.gov/p/af/rls/rpt/10445.htm. Rogers, Carol Ann, and Kenneth A. Swinnerton (2004), “A Theory of Exploitative Child Labor,” forthcoming in Oxford Economic Papers. United Nations Commission on Human Rights (1994), Situation of Human Rights in the Sudan: Report of the Special Rapporteur, Mr. Gáspár Bíró, submitted in Accordance with Commission on Human Rights Resolution 1993/60, Geneva. United Nations Commission on Human Rights (2004), Report of the United Nations High Commissioner for Human Rights and Follow-Up to the World Conference on Human Rights: Situation of Human Rights in the Darfur Region of the Sudan, Geneva. United Nations General Assembly (1995), Interim Report on the Situation of Human Rights in the Sudan prepared by Mr. Gáspár Bíró, Special Rapporteur 2 We thank Martin Bunzl for providing very helpful comments. The views presented in this paper are the personal views of the authors, and do not represent the official views of the U.S. Department of Labor.

36

ROGERS AND SWINNERTON

of the Commission on Human Rights, in Accordance with Commission 1995/77 of 8 March 1995, New York. United Nations General Assembly (1999), Report on the Situation of Human Rights in the Sudan prepared by Leonardo Franco, Special Rapporteur, in Accordance with Commission on Human Rights Resolution 1999/15 and Economic and Social Council Decision 1999/230, New York. United States Department of Labor (2000), By the Sweat and Toil of Children, vol. VI: An Economic Consideration of Child Labor, Washington, DC. United States Department of Labor (2002), Advancing the Campaign against Child Labor, vol. II: Addressing the Worst Forms of Child Labor. Washington, DC.

Chapter Three An Exploration of the Worst Forms of Child Labor: Is Redemption a Viable Option? ARNAB K. BASU AND NANCY H. CHAU

Introduction According to recent International Labor Organization (ILO) estimates, 211 million children between the ages of 5 and 14 are economically active, and of those, 186.3 million are child laborers.1 In addition, 5.7 million children are in forced or bonded labor, 1.8 million in prostitution and 0.3 million in armed conflict. Table 3.1 provides a summary of the regional distribution of child labor and underscores the magnitude of the problem. The extent of child labor varies regionally. Africa and the Middle East host approximately 30 percent of all economically active children worldwide. Meanwhile, the figure stands at around 60 percent in Asia (excluding Japan) and 8 percent in Latin America (including the Caribbean). Thus, in absolute terms, the incidence of child labor is highest in Asia (excluding Japan), as the most densely populated region of the world, with 127 million economically active children, compared to about 61 million in Africa and the Middle East and 17 million in Latin America (including the Caribbean). In addition, child labor can also be found in southern Europe and, increasingly, in the transitional economies of central and eastern Europe (ILO-IPEC 1997). However, in relative terms, sub-Saharan Africa ranks first when child labor prevalence is measured in terms of participation rates. In particular, approximately one out of every three children (or 29 percent) between 5 and 14 years of age are engaged in economic activities. In Asia about one in five children or 19 percent are part of the workforce, while in Latin America, one in six (or 16 percent). 1 According to ILO terminology, a child who worked for one hour or more the previous week is economically active. A child laborer is defined as a child between the ages of 5 and 11 who is economically active, or one aged 12–14 who does 14 or more hours of nonhazardous work per week or 1 hour of hazardous work per week.

38

BASU AND CHAU

TABLE 3.1 Estimates of economically active children (aged 5–14) in 2000

Region Developed economies Transition economies Asia and the Pacific (excluding Japan) Latin America and the Caribbean Sub-Saharan Africa Middle East and North Africa Total

% global total by group

Participation Rate (economically active children/ total child population) (%)

2.5 2.4

1 1

2 4

127.3

60

19

17.4 48

8 23

16 29

13.4 211

6 —

15 16

Number of economically active children (millions)

Source: ILO 2002.

The majority of child labor (70 percent) can be found in the agricultural sector (table 3.2). In urban centers, children are mostly employed in the informal sector, especially as domestic helpers, or in sweatshops, and only 5 percent are employed in the export sector. Despite the focus on the export sectors of developing countries as primary employers of child labor, the share of employed children is well below 1 percent in the export-processing zones.

TABLE 3.2 Distribution of Child Labor (5–14 years of age) by Type of Activity, 1997 Type of Activity Agriculture, hunting, forestry, and fishing Mining and quarrying Manufacturing Construction Wholesale and retail trade, restaurants, and hotels Transport, storage, and communication Community, social, and personal services Source: ILO Bureau of Statistics, 1996, 1997.

Both sexes (%)

Boys (%)

Girls (%)

70.4 0.9 8.3 1.9

68.9 1.0 9.4 2.0

75.3 0.9 7.9 1.9

8.3 3.8 6.5

10.4 3.8 4.7

5.0 — 8.9

THE WORST FORMS OF CHILD LABOR

39

Child labor is not a new phenomenon. As Humphries (2003) points out, child labor was more prevalent in the nineteenth century in the newly industrializing countries of Britain, France, Belgium, western parts of Prussia, and the United States than in today’s developing economies. Whereas Robert Peel’s Factories Act of 1802 in Great Britain was widely recognized as the start of the decline in child labor, similar efforts by developing countries in the twenty-first century seemed to have had relatively little impact in eradicating the problem. These efforts include the ratification and adoption of a range of international legislation (such as the ILO Convention 138 concerning Minimum Age for Admission to Employment [1973], the UN Convention on the Rights of the Child [1989], and the ILO Convention 182 on the Elimination of the Worst Forms of Child Labour [1999]), in addition to domestic legislation by developing countries against child labor employment. It has been frequently alleged that the lack of effectiveness of legislation to mitigate the child labor problem in developing countries has to do, at least in part, with lax enforcement of these laws, due either to a lack of resources or to corruption. But more importantly, it has also been argued that legislation per se, without an accompanying increase in the economic well-being of poor households (as was the case for much of newly industrialized western Europe and the United States), is unlikely to have any significant impact on the incidence of child labor. Indeed, the basic economic theory of child labor identifies poverty as the main reason altruistic households choose to send their children to work. Basu and Van (1998) posit that if adult wages fall below a subsistence level, children will be sent to work to supplement household income in order to safeguard a minimum consumption level. Thus, it is possible for otherwise identical economies to exhibit drastically different child labor employment patterns, with high adult wages and no incidence of child labor in some countries, low adult wages and a high incidence of child labor in others.2 In addition, the prevalence of low adult wages and high levels of child labor, if unchecked and allowed to impact the decision-making of individual households with regard to education, can also translate into low investment in human capital for current and future generations. A vicious cycle of poverty can thus ensue, with child labor serving as both the cause and the effect of the phenomenon. As the 2 In recent years, economists have identified a host of market imperfections, for example lack of alternatives to child labor, i.e., education (Ranjan 2001; Grote, Basu, and Weinhold 1999), credit market failures that prevent poor households from borrowing to finance either education or consumption (Baland and Robinson 2000; Ranjan 1999; Basu and Chau 2003, 2004) and coordination failures on the part of households to bargain for higher wages (Basu and Chau 2003) as important factors in addition to poverty that lead to the emergence and persistence of child labor.

40

BASU AND CHAU

developed world and international organizations target poverty amelioration as a major challenge in the twenty-first century (United Nations Millennium Declaration 2001), it is perhaps not all that surprising that there has been a concurrent resurgence of interest in child labor research. While the nexus between child labor and poverty naturally engages policymakers to focus on the economic aspects of child labor, there is a parallel interest in the ethical perspective. Kanbur (2003) and Satz (2003a, 2003b) identify the market for child labor as a “noxious” market. The term “noxious” is introduced depending on whether one or more of the following four criteria are present. The first is concerned with ‘weak agency.’ In the context of the market for child labor, the criterion of weak agency is particularly appropriate, since children lack the power within a family to make decisions regarding how to allocate their time. Whereas in ideal labor markets individuals make informed decisions regarding their time allocation, parents make market decisions on behalf of their children. Even when parents behave altruistically and their children’s well-being comes first, the notion of weak agency may continue to apply, particularly when parents are equipped with less than perfect information regarding the true effects of working on their children or about the real nature of work. The second criterion has to do with distributive inequality. Again, markets for child labor are often “noxious” based on this criterion. Examples abound, and include societies where poor children work for rich multinational firms (as with the case of sweatshops), and where relatively poor children work as bonded laborers to pay off their parents’ debt, while the children of rich families are well educated. The third criterion is concerned with harmful outcomes. Indeed, the recent wave of attention focusing on the worst forms of child labor belongs to this category. At stake are a child’s cognitive ability and basic human rights, for example, which may be severely violated for a period of time over which the child has little or no control. Well-documented cases in which child labor unwittingly becomes involved in the parallel markets, such as child trafficking and prostitution, is yet another example in this regard. These examples also fit well into the final criterion that Kanbur (2003) and Satz (2003a, 2003b) single out, one having to do with negative externalities. In particular, the question of whether decentralized and voluntary decisionmaking on the part of households, or the pursuit of profit-maximization, on the part of employers, serve societal goals and values is not clear-cut when children are involved in physically and mentally harmful activities such as trafficking and prostitution. Perhaps in light of these ethical dimensions of child labor, particularly as they relate to distributive inequality, harmful outcomes, and negative

THE WORST FORMS OF CHILD LABOR

41

externalities, two specific types of the “worst forms of child labor,” namely, child trafficking and child labor in debt bondage, have frequently been referred to as a form of forced labor or slavery, even though a child’s ultimate involvement with these activities may have come from an exante “voluntary” agreement with a recruiter, employer, or landlord. We will now turn to a more in-depth look at these two forms of child labor. In consonance with the theme of this volume, our main focus will be on the efficacy of legislation that grants human rights (to those entrapped in these forms of slavery) on the incidence of the problem itself. Worst Forms of Child Labor As a matter of definition and identification, the ILO Convention on the Worst Forms of Child Labour (182) calls for the immediate suppression of all extreme forms of child labor including: (a) all forms of slavery or practices similar to slavery, sale, trafficking of children, forced or compulsory labor including debt bondage and serfdom. (b) the use, engagement, or offering of a child for the purposes of prostitution, production of pornography or pornographic performances, production of or trafficking in drugs or other illegal activities. (c) the use or engagement of children in any type of work, which by its nature or the circumstances in which it is carried out, is likely to jeopardize their health, safety, or morals.3 Although the ILO’s Convention on the Worst Forms of Child Labor has received the most international attention, there were two other earlier ILO conventions that specifically address the issue of slavery: 29, the Forced Labour Convention (1930), and 105, the Abolition of Forced Labour Convention (1957), which defines forced or compulsory labor as “all work or service which is extracted from any person under the menace of any penalty and for which the said person has not offered himself voluntarily.” In addition to these ILO conventions, the United Nations specifically targets trafficking through its International Agreement for the Suppression of White Slave Traffic (1904),4 while the United Nations Supplementary Convention on the Abolition of Slavery (1956) defines debt bondage as “the status or condition arising from a pledge by a debtor of his personal services or of those of a person under his control See Diller and Levy (1997) for an excellent survey of international conventions and laws governing child labor. 4 See Dennis (2000) for a comprehensive treatment of UN and other international legislation regarding child labor. 3

42

BASU AND CHAU

as security for a debt, if the value of those services as reasonably assessed is not applied towards the liquidation of the debt or the length and nature of those services are not respectively limited and defined.” Despite international attention and calls for the elimination of the worst forms of child labor, formal economic analysis on the subject remains scant. Rogers and Swinnerton (2002) consider a general theory of the worst forms in which parents decide to send their children to work based on incomplete information regarding the nature of the work. They show that a ban on the worst forms of child labor leads to an increase in child wages in an economy and increases the incidence of child labor. Thus, the prevalence of the worst forms of child labor in the presence of imperfect information impinges on the supply of children to the labor market. In a more specific context, Friebel and Guriev (2004) consider the economic incentives of a trafficker of illegal immigrants. They posit a theory that accounts for circumstances in which trafficked or illegal immigrants have no recourse to formal credit markets. As such, victims of trafficking have no choice but to borrow from the trafficker to finance their migration and repay in the form of labor services. Thus, when moving from the illegal to the legal sector becomes costly due to stricter deportation laws, fewer immigrants default on their debt obligation. This serves to strengthen the incentives of traffickers and leads to an increase in the incidence of trafficking. The phenomenon of debt bondage is explicitly modeled in a sequence of papers by Basu and Chau (2003, 2004). Our model considers the situation where poor agrarian households have no access to formal credit markets and borrow from landlords-cum-moneylenders during the lean season to finance their subsistence consumption in return for a pledge of adult and child labor services during the peak or harvest season. The basic model is also extended to an overlapping generation framework, in which it is shown that the effects of bondage in one generation can spill over to adversely affect future generations’ welfare, and their ability to break free from bondage. In this situation merely freeing households from debt bondage by paying off their debts (without improving their ability to finance subsistence consumption in the lean season) results in the same households falling back into debt bondage the very next year. These theoretical papers notwithstanding, there is little empirical work on a cross-national basis that attempts to unravel the determinants of these two types of the worst forms of child labor, with the exception of Basu and Chau (2003) for debt bondage. Thus, the primary goal of this chapter is to explore the determinants of trafficking on a cross-national basis, and to underscore our findings on debt bondage from Basu and Chau (2003, 2004). Even though the institutional features of these two

THE WORST FORMS OF CHILD LABOR

43

types of child labor are distinct, they nevertheless share two common economic features that precipitate the problem—market imperfections and differential bargaining power amongst the concerned parties. In the case of trafficking, market imperfection stems from incomplete information on the part of buyers of services from trafficked persons, who do not know how much to pay these workers. Meanwhile, potential sellers of services (the trafficked individuals) are likely imperfectly informed about the sort of work they will be engaged in and accordingly what sort of pay to expect. In the case of debt bondage, the market imperfection stems from the absence of a formal credit market that can insure poor households against income uncertainty. In terms of bargaining power, the problem of trafficking is made worse as a larger number of uncoordinated and anonymous buyers and sellers are pitted against a small number of coordinated traffickers. Buyers lack bargaining power, since the consumption of services offered by trafficked individuals is illegal. The same lack of bargaining power also plagues potential sellers, who are unable to coordinate with each other due to their unmet income needs. Likewise for debt bondage, a large number of poor households who are unable to unionize and bargain for higher wages are typically faced with a small number of landlords-cum-moneylenders. In the case of trafficking, unequal bargaining power implies that traffickers reap most if not all of the rent generated from the buyer-seller match. For the case of debt bondage, unequal bargaining power implies that the structure of the debt service contract makes it particularly difficult for households to subsequently break free of bondage. The economics of the worst forms of child labor, as seen through the lens of these two forms of market imperfections, underscores an additional issue that lies at the heart of the question of whether redemption can work: with multiple market imperfections, piecemeal policy reforms may well falter (Lipsey and Lancaster 1956). Redemption is of course an example of piecemeal reform, since it does not address the underlying sources of the child labor phenomenon. Thus, there is little guarantee that the granting of asylum to trafficked victims will not, for example, instigate a renewed surge in demand and supply of trafficked children. Likewise, simply paying off the debts of existing bonded child laborers does not release these children from the root cause of the phenomenon, ultimately having to do with the lack of access to credit and the lack of bargaining power in employment relations. Put another way, the central question here is whether there are justifiable reasons to believe that the following two social objectives are competing rather than reinforcing: (i) freedom from slavery and serfdom as a human right for those who have been enslaved, and (ii) a sustainable decrease in the observed incidence of these two forms of

44

BASU AND CHAU

child labor. Our examination in what follows can be understood as a step in this direction, drawing on existing data on trafficking and debt bondage.

Child Trafficking and Prostitution The term trafficking refers to a set of interrelated activities that encompass migration, prostitution, and acts that violate human and, especially, children’s rights. The term is synonymous with illicit trade in human beings across international borders or within the same country. Trafficking of children is often discussed together with the trafficking of women, the main reasons being that (i) available data on trafficking of women are not disaggregated by age and (ii) there is considerable debate regarding the age at which a child should be considered an adult. For example, the majority of women coerced into prostitution are between 16 and 24 years of age (International Labour Organization 2002). However, 16- and 17-year-old girls are children according to the 1989 UN Convention on the Rights of the Child (CRC) and the ILO Worst Forms of Child Labor Convention (182). The issue of age is further complicated by considerable variations in national laws regarding the age until which an individual is considered a “child.” As Dennis (2000) points out, Ireland protects those under 17 as children, while Australia, Belgium, the Netherlands, New Zealand, Switzerland, and the United Kingdom accord protection to those under 16. The age for protection is 15 in France, Sweden, and Denmark, while it is 14 in Austria and Germany. The age of consent for sexual matters is 15 in Denmark; thus, the word “child” in the Danish provision on child pornography is only applicable to individuals below the age of 15. The above variation regarding the legal definition of a child notwithstanding, there is a broad consensus regarding activities that fall under the definition of trafficking. Trafficking is said to occur if (i) a child is misled with false reports or promises, coerced, or otherwise forcibly recruited/handed over to transporters; (ii) a child is lied to about the destination; and (iii) a child is lied to about either the nature of work (i.e., recruited as a dancer but forced into prostitution) or the wages and methods of payment. Trafficking may also take the form of physical or mental abuse, confinement, inadequate or nonexistent health care, poor accommodation, and hazardous work (International Labour Organization 2002). The determinants of trafficking can be broken up into push (supplyside) and pull (demand-side) factors. Some of the supply-side factors that lead to trafficking and prostitution are the same that lead to the emergence and persistence of child labor in the first place: poverty and lack of

THE WORST FORMS OF CHILD LABOR

45

educational opportunities. In addition, armed conflict in some African countries such as Sierra Leone and Sudan gives rise to social fragmentation that makes it easier for children to be forcibly removed and trafficked by various factions. On the demand side, the largest demand for child trafficking and prostitution can be linked to the growth in income in both developed and developing countries. For developed countries, two factors are frequently alleged to be in play: (i) the rise of tourism from developed to developing countries and the subsequent rise in the demand for “sex tourism” as developed countries increasingly strengthen laws to protect minors and increase enforcement related to prostitution, and (ii) economic growth in developed countries, low fertility, and the subsequent increase in demand for cheap migrant labor. The latter effect is also evident in some developing countries that have witnessed relative prosperity over the last decade, and where the native population has gradually moved away from low-skilled, low-wage employment sectors. Consequently, migrants, legal and illegal, are now filling employment in these low-wage sectors. As an example, an increased number of children have migrated and/or have been trafficked into Thailand from Myanmar, Laos, and Cambodia to work in exploitative jobs previously done by Thai children (International Labour Organization 2002). The pattern of trafficking and the types of work that trafficked individuals are engaged in vary across countries and continents. For instance, in Africa the most common source countries of trafficking are Sierra Leone, Malawi, Mozambique, Nigeria, and Somalia, from where children and women are trafficked to South Africa, Gabon, Gambia, and Western European countries primarily to work in the sex industry. There is also a steady supply of trafficked children from Mali to Côte d’Ivoire who end up working in cocoa plantations (UNICEF 2003). In Asia, the most common source countries of trafficking seem to be Bangladesh, Nepal, Vietnam, Bhutan, Laos, and Cambodia, while the host countries are India, Thailand, Sri Lanka, Saudi Arabia, United Arab Emirates (UAE), and Australia. Trafficked children and women are engaged in a variety of work. The primary activity remains prostitution, but an increasing number also work as domestic helpers in host countries, while young boys are smuggled from Bangladesh and India to work as camel jockeys in Saudi Arabia and the UAE (International Labour Organization 2002). In Central and South America, the source countries of trafficking are Bolivia, Colombia, Ecuador, and Honduras, and the host countries are Argentina, Belize, Mexico, El Salvador, and, in some cases, Guatemala, where trafficked women and children are usually coerced into prostitution. Similarly, trafficked women and children from the former republics of the Soviet Union (Estonia, Georgia, Ukraine, Uzbekistan, Kazakhstan and Kyrgyzstan) are usually forced

46

BASU AND CHAU

into prostitution in Western European countries (especially, Germany, Italy, and Greece). It bears emphasis that a distinguishing feature of international trafficking is that a middleman is involved. Such a middleman may be an individual “recruiter” with an aim to serve the receiving end at the lowest cost, a smuggler of illegal migrants specializing in evading border controls, or a criminal network operating on both the supply and the final demand sides (U.S. Department of State 2003). Thus, in addition to the aforementioned push and pull factors, it is important to understand how these different types of traffickers operate, and more importantly, how legislation granting human rights to trafficked individuals in the host and source countries of trafficking affects the incentives of traffickers. Unfortunately, there are no data on how traffickers operate, not to mention their earnings from trafficking. As a result, we proxy the incentives of traffickers by analyzing their locational choice as captured by host-source matches amongst countries, as delineated below.

Cross-National Evidence on Trafficking of Women and Children To the best of our knowledge, there has been no systematic cross-national empirical analysis of the features characterizing countries that are subject to the forces of international trafficking. Given the problems associated with segregating data on women from that on children due to the various interpretations of age, we look at the joint incidence of trafficking in women and children. As an initial step towards an understanding of the push and pull factors of international trafficking of women and children, we make use of the country-by-country reports in the Trafficking in Persons Report (U.S. Department of State 2003) and the Protection Project (2002). We are interested in obtaining two sets of information for each country. First, we classify countries based on four mutually exclusive groupings: a host country, a source country, a trafficking hub (both a host and a source country), or one with no reported incidences of trafficking. Second, and specifically with an eye towards examining the push and pull forces that dictate the direction of international trafficking, we construct a binary variable traffickhs. The unit of analysis here is a potential host-source country pair, country h and country s; the variable takes on a value of 1 if trafficking from country s to country h has been reported, and 0 otherwise. Table 3.3 lists the 187 countries included in our data and the nature and prevalence of trafficking that have reportedly taken place in these countries. It should be noted that countries are sorted based on reported incidence only, and should be interpreted as such.

47

THE WORST FORMS OF CHILD LABOR

TABLE 3.3 List of Countries and Status of Trafficking None Andorra Bahamas Barbados Burundi Comoros Croatia Djibouti Egypt Eritrea Fiji Iceland Jamaica Lesotho Liechtenstein Luxembourg Maldives Malta Marshall Islands Micronesia Monaco Namibia Nauru New Zealand Niue Oman Palau Palestine Papua New Guinea Paraguay Saint Kitts and Nevis Saint Lucia Saint Vincent and the Grenadines Samoa San Marino

Host Antigua Australia Austria Belgium Belize Bosnia and Herzegovina Botswana Canada Central African Republic Chile Côte d’Ivoire Denmark Finland France Gabon Gambia Germany Greece Hong Kong (SAR)

Hub

Source

Afghanistan Albania Argentina Bahrain Bangladesh Benin

Algeria Angola Armenia Azerbaijan Belarus Bhutan

Brazil Brunei Bulgaria

Bolivia Cape Verde Colombia Cuba Ecuador Estonia Ethiopia Georgia Guyana Honduras Iraq Kenya Latvia

Israel Italy Japan Kuwait Lebanon Libya Macau (SAR) Mauritius Netherlands

Burkina Faso Cambodia Cameroon Chad China Congo, Dem. Rep. Costa Rica Cyprus Czech Republic Dominican Republic El Salvador Equatorial Guinea Ghana Guatemala Haiti Hungary India Indonesia Iran

Madagascar Malawi Mauritania Moldova Morocco Mozambique Nepal Nicaragua Sierra Leone

Norway Portugal

Kazakhstan Kosovo

Slovenia Somalia

Qatar Rwanda

Kyrgyzstan Laos

Tajikistan Zambia

Saudi Arabia Singapore

Liberia Lithuania Continued

48

BASU AND CHAU

TABLE 3.3 (Continued) None Sao Tome and Principe Seychelles Solomon Islands Tonga Trinidad and Tobago Tunisia Turkmenistan Tuvalu Uruguay Vanatu

Host

Hub

Spain

Malaysia

Suriname Swaziland Sweden Switzerland

Mali Mexico Mongolia Myanmar

Syria United Arab Emirates United Kingdom United States Yemen Yugoslavia

Niger Nigeria

Source

Pakistan Panama Peru Philippines Poland Romania Russian Federation Senegal Slovakia South Africa South Korea Sri Lanka Sudan Taiwan Tanzania TFYR Macedonia Thailand Togo Turkey Uganda Ukraine Uzbekistan Venezuela Vietnam Zimbabwe

Despite this obvious drawback, a number of useful insights may be gleaned by comparing the economic, demographic, legislative, and law enforcement characteristics of these countries stratified according to our classification. All economic and demographic variables are taken from World Bank (2004) for the year 2000.

THE WORST FORMS OF CHILD LABOR

49

These economic, demographic, and labor market characteristics of host and source countries put international trafficking squarely in the context of the push and pull factors enumerated above. Doing so does not, however, allow one to distinguish the forces that govern international voluntary migration, for example, from the specific circumstances under which the trafficking of individuals occurs, including coercion, deceit, kidnapping, and forced or slave labor. To this end, we collected data on legislative and law enforcement variables in order to examine the extent to which international trafficking may also be looked at as a locational choice problem for criminal activities. These legislative and law enforcement variables are taken from the Protection Project (2002), the latest available years of the UN’s International Crime Victim Survey and the Seventh United Nations Survey of Crime Trends of the United Nations (2004) for 2000. Macroeconomic, Labor Market, and Demographic Factors First, while systematic estimates of the size and scope of international trafficking are unavailable, our four-tier classification allows for a raw gauge of the share of the world’s children who may be at risk of being trafficked. In particular, countries that are trafficking hubs host close to 77 percent of all children (ages 0–14) captured in our sample of 187 countries, yet these same countries account for only 18 percent of the total gross domestic product of all countries combined in 2000. An additional 9 percent of all children live in source countries of international trafficking, but these source countries produce 1 percent of the total gross domestic product of the 187 countries. The remaining 14 percent of children live in host countries (11.4 percent) and countries with no reported incidences of international trafficking (2.5 percent); combined, these two groups of countries account for over 81 percent of the total gross domestic product of the countries in our sample. In terms of the labor force and demographic characteristics of these countries, a typical worker in a source country of international trafficking or a trafficking hub is more likely to be employed in agriculture as compared to a worker in a host country. We also observe a higher dependency ratio (ratio of children ages 0–14 to total population of a country) due primarily to the larger size of the child population in source countries and trafficking hubs. There is likewise a correspondingly higher incidence of child labor in source countries and trafficking hubs, with shares of economically active children (ages 10–14) 13.7 percent in source countries and 12.9 percent in trafficking hubs, compared to around 3.8 percent in host countries. Employment of adults, in contrast, exhibits the opposite pattern, with adult unemployment rates for both

50

BASU AND CHAU

female and male nearly twice as high in source countries and trafficking hubs, compared to host countries of trafficking. Meanwhile, dependence on income through workers’ remittance from employment or other income sources abroad as a percentage of gross national product is higher in source countries and trafficking hubs (2.6 percent) as compared to host countries of trafficking (1.6 percent). Taken together, and as table 3.4 summarizes, these observations are in concertwiththepresumptionthatsourceandhostcountriesofinternational trafficking stand on opposite ends of the spectrum of countries ranked based on income, dependency ratio, the degree of industrialization, and domestic employment rates. Turning now to global links, a different picture emerges. The average host country’s dependence on trade, as measured by the trade share of GDP, is in fact quite similar to that of the average source country. Indeed, we cannot reject the hypothesis that the trade share of all source countries (including trafficking hubs) is the same as the rest of the countries in the sample. As table 3.5 summaries, the same is true when the comparison is made between source countries alone (not including trafficking hubs) and the rest of the countries in the sample, or host countries alone and the rest of the countries in the sample. In addition to merchandise trade, one may also take international tourism as another measure of the strength of global links. In this respect, we observe no statistically significant difference between international tourism expenditure, either as a fraction of total exports or gross national product, between source and host countries of trafficking. More precisely, the notion that countries like Thailand which rely heavily on tourist revenues (which includes revenues from alleged sex tourism) are also host countries of trafficked women and children for prostitution-related activities is rejected by our findings. TABLE 3.4 Macroeconomic, Labor Market and Demographic Factors

Employment in Ag (% of total) Share of Minors (ages 0–14) in Total Population (%) Share of Seniors (ages 65
) in Total Population (%) Labor Force Characteristics Child Labor (% 10–14 age) Female Unemployment Rate Male Unemployment Rate Worker Remittance (% of GDP)

Host

Hub

Source

All

5.53

26.90

25.10

17.90

26.37

34.13

35.69

32.37

9.48

5.73

5.85

6.83

3.84 7.55 5.95 1.64

12.90 11.15 9.43 2.40

13.67 11.93 11.25 2.65

10.18 10.24 8.38 2.17

51

THE WORST FORMS OF CHILD LABOR

TABLE 3.5 Global and Informational Links

Trade Share of GDP (%) International Tourism Expenditure (% imports) International Tourism Receipt (% exports) International Tourism Expenditure (% GDP) International Tourism Receipt (% GDP) Official Development Assistance (% GDP)

Host

Hub

Source

All

86.32

77.11

87.35

87.06

6.42

4.91

6.02

5.77

10.21

10.44

12.13

14.01

2.84

1.94

2.79

2.66

5.55

3.69

3.65

6.57

2.88

4.30

9.60

6.67

Indeed, international tourism receipts (whether as a fraction of export revenues or of gross national product) are on average smaller in countries that host trafficked victims in our sample.5 Legislations and Law Enforcement There are a number of international conventions that are related to the condemnation of the trafficking of women and children. We have looked at the patterns of ratification of the ILO conventions on the Abolition of Forced Labour (105) and the Worst Forms of Child Labour (182), along with three other United Nations protocols: (i) calls for an end to the sale of children for the purposes of prostitution and pornography (OPSC), (ii) calls for an end to the trafficking of persons, especially women and children (PPSPT), and (iii) the Migrant Workers’ Convention (MWC), which calls for the protection of the rights of migrant workers and members of their families. As may be expected, the patterns of the ratification of these conventions differ widely (table 3.6). These patterns range from almost universal commitment to abolish forced labor, to the relative popularity, among host countries of trafficked individuals, of a commitment to eliminate the worst forms of child labor, which includes the sale of children and international trafficking. By contrast, ratification of the Migrant Workers Convention is more prevalent among source countries of international trafficking. 5 Though beyond the scope of this chapter, global capital linkages between these four classes and the rest of the world exhibit a rather more systematic pattern. For example, source countries of international trafficking and trafficking hubs are more dependent on official development assistance than are host countries of international trafficking.

52

BASU AND CHAU

TABLE 3.6 International and National Legislations Host

Hub

Source

All

Ratification of International Conventions ILO Convention 105 ILO Convention 182 PPSPT OPSC MWC

88.1% 61.9% 47.6% 50.0% 4.8%

82.5% 54.0% 50.8% 33.3% 14.3%

89.7% 37.9% 37.9% 34.5% 20.7%

84.3% 53.0% 47.0% 39.8% 13.3%

National Legislations Grant Legal Status to Trafficked Victims Prohibit Prostitution Prohibit Pimping Prohibit Brothels

22.2% 35.6% 86.7% 55.6%

6.2% 41.5% 75.4% 41.5%

0.0% 34.4% 75.0% 50.0%

7.6% 38.3% 79.2% 47.0%

Turning now to national statutes with specific reference to trafficking and prostitution, we see a relatively high average percentage of host countries enacting laws that protect the rights of trafficked victims. Around 22 percent of host countries grant legal status to trafficked victims, whereas no source countries do so. With respect to legal restrictions on prostitution and other related activities, we find that laws banning prostitution are in fact most common among trafficking hubs, followed by host and source countries. Meanwhile, laws banning activities surrounding and promoting prostitution, such as pimping, pandering, and brothels, are more common in host countries. With the exception of the migrant worker convention, which covers voluntary migrants as well, the preponderance of these observations would seem to indicate that a larger average share of host countries enact legislation answering to the international call to end trafficking and to protect trafficked victims.6 Host countries are likewise more likely to grant legal status to trafficked victims and to enact legislation that legalizes prostitution. 6 It is important to note that why and when countries adopt international conventions is an area of inquiry that is still in its infancy. Chau and Kanbur (2002) show, with specific reference to the four core labor standards of the ILO, that the ratification pattern is highly convention-specific. What is important in our context is that whereas countries with higher income per capita (the host countries) appear to be more likely to participate in international conventions and domestic legislation protective of victims’ rights, the same is not true for many other core labor standard conventions. In fact, the stage of development of an economy is not always a good predictor of ratification.

53

THE WORST FORMS OF CHILD LABOR

TABLE 3.7 Crime and Law Enforcement Host

Hub

Source

All

Police Personnel (per 100K) 289.65 288.78 272.04 286.38 Total Recorded Crimes (per 100K) 4918.92 1825.94 1387.27 2991.97 Total Convicted Persons (per 100K) 809.62 735.15 337.01 692.38 Convicted Persons per Recorded Crime 0.17 0.30 0.41 0.27 Rule of Law Index 0.73 –0.30 –0.52 0.01

In addition to the enactment of national laws and the ratification of international conventions, one might argue that of even more importance is the extent to which these laws are enforced. To this end, table 3.7 summarizes data taken from the UN Survey of Crime Trends (Seventh Survey) for 2000. The capacity of policy enforcement is expressed in two ways. The variable police denotes the number of police personnel per 100 thousand persons. The variable convicted persons per crime gives the number of convicted persons per reported crime in a country. The variable police allows us to pin down the physical capacity of the police force, while, the variable convicted persons per crime is concerned with the efficiency of the police force. These two variables give two very different pictures of the capability of police enforcements. In particular, host countries have, on average, a higher police force per capita than do source countries and trafficking hubs. Nevertheless, the number of convicted persons per recorded crime is also the lowest there. These conflicting observations may have to do with underreporting in official police data. There are two reasons why underreporting in source countries is of interest in the context of trafficking. First, for traffickers operating in potential source countries, underreporting is of course advantageous, since the likelihood of getting caught is accordingly lower. Second, underreporting may also be a signal of the public’s distrust of the police force, due for instance to corruption among public officials. Both of these factors concern the degree of access to effective law enforcement, and separate the (economic) push and pull factors of international migration, as distinct from the criminal activities associated with international trafficking.7 As a partial remedy to this issue of access, we will further make use of the “rule of law” governance indicator8 from Kaufmann, Kraay, and Zoido-Lobaton (1999a, 1999b). 7 Exactly why report rates are lower in source countries is of course of independent interest, although it is certainly beyond the scope of this chapter. 8 The “rule of law” indicator is a composite index of (i) voice (e.g., freedom of press and freedom to associate) and accountability; (ii) political and stability/lack of violence;

54

BASU AND CHAU

Empirical Findings and the Economic Consequences of Redeeming Trafficked Individuals In what follows, we report the results of the logit estimation9 that accounts for the push and pull factors of trafficking, in which the dependent variable of interest is a host-source country match. Three different sets of results are presented in tables 3.8–3.10. The first regression employs basic macro variables in source and host countries, along with regional, distance, geographic, and political variables. The variable comborder is an indicator variable, taking on a value of 1 if two countries share a common border. The variable comreg indicates common region, the regions being defined as East Asia and the Pacific, Western Europe, North America, Eastern and Central Asia, Sub-Saharan Africa, South Asia, and Latin American Countries. The variable landlock is a dummy variable for landlocked countries, and the variable transition takes on a value of 1 for countries in transition (for example, former Soviet republics) and 0 otherwise. The second regression introduces legislative variables and international conventions related to international trafficking of children and women. The third regression introduces law enforcement variables. From these results, a number of general remarks can be made. First, as table 3.8 shows, the stage of economic development of the countries in question plays an important role throughout. In particular, a higher real per capita income (the variable lrgdppc denoting real gross domestic product per capita) in a potential destination country, and a lower real per capita income in a potential country of origin, raises the likelihood of trafficking between these two countries. This suggests that international inequality in the distribution of income is systematically correlated with international trafficking. Introducing the Gini coefficient10 as a measure of intra-country inequality in the distribution of income to (iii) government effectiveness; (iv) regulatory framework; (v) rule of law, and (vi) control of corruption. The index ranges from –3 (worst) to
3 (best). 9 In our logit regression, the binary variable, traffick , takes on the value of 1 when inhs cidences of trafficking going from the source country, s, to the host country, h, have been observed. Otherwise, traffickhs is set to zero. The logit regression model assumes that the probability that traffickhs  1 takes on the following form: Prob (traffickhs  1/xs, xh,yhs)  exp(axs
bxh
cyhs)/(1
exp(axs
bxh
cyhs)) where xs and xh are respectively source and host country-specific variables, and yhs pertains to country pair variables, such as when two countries share a common border. The maximum likelihood coefficients accordingly indicate whether an increase in xs, xh, and yhs are associated with a higher likelihood that two countries s and h are indeed a source and host country pair of international trafficking. 10 The Gini coefficient measures the degree of inequality in income distribution within a country. A value of 0 indicates a perfectly equitable income distribution (such that the poorest

55

THE WORST FORMS OF CHILD LABOR

TABLE 3.8 Gravity Model of the Push and Pull Factors of Trafficking Basic Economic and Geographic Variables Dependent Variable: Incidence of Trafficking of Women and Children

Host Country

lrgdppc transition

I

II

III

0.625*** 0.030 –1.286*** 0.208

0.703*** 0.145 –2.176*** 0.901 –0.008 0.020

0.625*** 0.031 –1.201*** 0.210

gini landlock Source Country

lrgdppc transition

–0.443*** 0.140 –0.431*** 0.028 1.756*** 0.112

gini

–0.953*** 0.201 3.474*** 0.603 0.107*** 0.025

landlock Common

–0.649*** 0.109

constant

–6.943*** 0.321 Region (comreg) yes*** Border (comborder) yes***

N Wald Chi2 Prob  Chi2 Pseudo R2 Log Likelihood

–0.511*** 0.032 1.984*** 0.116

28056 964.680 0.000 0.206 –2154.613

–7.673*** 2.393 yes*** yes***

–6.235*** 0.351 yes*** yes***

870 81.830 0.000 0.339 –111.172

28056 950.780 0.000 0.215 –2131.545

* Significant at 1% level. ** Significant at 5% level. *** Significant at 10% level.

check whether income inequality within countries (as opposed to income inequality across countries) has a role to play in raising the likelihood of trafficking, reduces the number of available observations by a great deal. However, the results are in support of Rogers and Swinnerton (2001), and 10 percent of the population has access to 10 percent of national income), while a value of 1 indicates perfect inequality in the distribution of income (for example, the top 1 percent of the population has access to all or 100 percent of national income).

Source Country

Host Country

prostitution

grant legal

lrgdppc

transition landlock

opsc

ppspt

iloc182

prostitution

grant legal

lrgdppc

–0.525*** 0.038 –0.379 0.256 0.319** 0.102

0.508*** 0.039 1.130*** 0.135 0.224** 0.106 –0.196* 0.102 –0.007 0.112 –0.205 0.180 yes*** yes***

I

–0.534*** 0.038 –0.374 0.259

–0.162 0.103 –0.065 0.114 –0.233 0.179 yes*** yes***

0.503*** 0.039 1.099*** 0.131

II

–0.524*** 0.033

yes*** yes***

–0.372*** 0.098

0.645*** 0.033

III

–0.499*** 0.032

yes*** yes***

0.152 0.101

0.601*** 0.034

IV

Dependent Variable: Incidence of Trafficking of Women and Children

Gravity Model of the Push and Pull Factors of Trafficking Legislations and International Conventions

TABLE 3.9

–0.486*** 0.032

–0.343* 0.174 yes*** yes***

0.598*** 0.032

V

Region Border

constant

* Significant at 1% level. ** Significant at 5% level. *** Significant at 10% level.

N Wald Chi2 Prob  Chi2 Pseudo R2 Log Likelihood

Common

transition landlock

opsc

ppspt

iloc182

23256 1014.860 0.000 0.238 1875.809

–5.640*** 0.398 yes*** yes***

0.269*** 0.104 0.377*** 0.099 –0.045 0.142 yes*** yes***

23562 993.920 0.000 0.235 1886.411

–5.332*** 0.392 yes*** yes***

0.347*** 0.101 0.370*** 0.098 –0.099 0.139 yes*** yes***

23562 877.580 0.000 0.216 1933.292

–6.162*** 0.356 yes*** yes***

yes*** yes***

0.353*** 0.098

23562 904.350 0.000 0.215 –1937.685

–6.237*** 0.357 yes*** yes***

yes*** yes***

0.374*** 0.098

23562 904.350 0.000 0.212 1944.462

–6.011*** 0.360 yes*** yes***

–0.109 0.138 yes*** yes***

Source Country

Host Country

grant legal

lrgdppc

conviction rate transition landlock

police

report rate

rule of law

prostitution

grant legal

lrgdppc

–0.419*** 0.050 –0.315

yes*** yes***

0.011 0.006

0.421*** 0.050 1.145*** 0.119 0.129 0.105

I

–0.406*** 0.052 –0.323

yes*** yes***

0.493*** 0.064 1.149*** 0.120 0.120 0.104 –0.025 0.098

II

–0.218*** 0.127 –0.301

–1.247** 0.627 yes*** yes*

0.043* 0.022

0.172*** 0.202 1.352*** 0.228 0.002 0.291

III

–0.580*** 0.074 –0.458

yes*** yes

0.001 0.001

0.542*** 0.110 1.037*** 0.216 –0.199 0.270

IV

Dependent Variable: Incidence of Trafficking of Women and Children

Gravity Model of the Push and Pull Factors of Trafficking Law Enforcement

TABLE 3.10

–0.486*** 0.032 –0.232

–1.055* 0.567 yes*** yes

0.333 0.209 1.383*** 0.236 0.065 0.290 0.211 0.262

V

Region Border

constant

* Significant at 1% level. ** Significant at 5% level. *** Significant at 10% level.

N Wald Chi2 Prob  Chi2 Pseudo R2 Log Likelihood

Common

conviction rate transition landlock

police

report rate

rule of law

prostitution

22052 920.410 0.000 0.232 1949.345

–4.919*** 0.385 yes*** yes***

yes*** yes***

–0.023*** 0.008

0.245 0.303*** 0.092

22052 909.790 0.000 0.231 1950.047

–6.066*** 0.645 yes*** yes***

yes*** yes***

0.245 0.284*** 0.091 –0.282*** 0.084

2352 199.200 0.000 0.310 375.502

–1.903*** 1.445 yes*** yes***

–1.049*** 0.447 yes*** yes*

–0.097*** 0.021

0.352 0.380 0.247

2756 236.980 0.000 0.257 387.105

–4.960*** 1.282 yes*** yes***

yes*** yes

0.001 0.001

0.306 0.507** 0.218

2352 203.330 0.000 0.316 372.689

–6.067*** 2.115 yes*** yes***

–1.213*** 0.441 yes*** yes*

0.364 0.237 0.251 –1.410*** 0.254

60

BASU AND CHAU

suggest in particular that income inequality in source countries has a place among the push factors of trafficking. Second, the role of ratifying international conventions is much less clear, as shown in table 3.9. In particular, for a potential source country, having ratified the Worst Forms of Child Labor convention (the variable iloc182), and the two other United Nations protocols that respectively call for an end to the sale of children for the purposes of prostitution and pornography, and to the trafficking of persons, especially women and children (the variables opsc and ppspt respectively in table 3.9), is in fact associated with a higher likelihood of trafficking. A number of possible interpretations apply here. For example, ratifying may only signify the commitment to combat the problem of trafficking, and says little about the effectiveness of the measures implemented. Perhaps even more straightforwardly, there is little need to ratify a convention if the problem is nonexistent to begin with. Third, and in terms of law enforcement, we note that the coefficient associated with physical size of the police force is insignificant for host countries and of the wrong sign for source countries, indicating in particular that a larger police force is in fact associated with a higher likelihood of the outflow of trafficked victims (table 3.10). We attribute this to the endogeneity of the police variable (in the sense that the stage of development for a country and the effectiveness of the police force are positively correlated), an issue highlighted by Levitt (1997) and Soares (2004).11 In other words, poorer host countries also tend to have a more ineffective and corrupt police force. In contrast, the estimated coefficient for the rule of law variable is negative and significant for source countries. Controlling for this, and using the “conviction per reported crime” variable as a measure of police effectiveness, we find that effective enforcement is associated with a reduced likelihood of trafficking both for host and for source countries. Fourth, in terms of legislative variables, two interesting observations are worth noting from table 3.10. One would expect that either granting of legal status to victims in source countries or banning prostitution in host countries should ameliorate the problem of trafficking. Paradoxically, granting legal status to trafficked victims in host countries is associated with an increased likelihood of trafficking. Moreover, a legal ban on prostitution in source countries is also associated with a higher likelihood of trafficking. In sharp contrast to the findings in Friebel and Guriev 11 Becker (1968), Ehrlich (1973), and Stigler (1970) take observed crime rates as outcomes of the cost-benefit assessments, and demonstrate the endogeneity of the extent of criminal activities depending on the stage of development of an economy, inequality in the distribution of income, and enforcement capabilities.

THE WORST FORMS OF CHILD LABOR

61

(2004), our results indicate that the provision of amnesty and the inflow of trafficked victims appear to go hand in hand. Meanwhile, the banning of prostitution and the outflow of trafficked victims also appear to be positively correlated. Although these latter empirical findings might seem counterintuitive, there is a logical explanation to these observed outcomes. In a series of theoretical papers (Basu and Chau 2005; Basu and Seiberg 2005), we find that these paradoxical results may be due to the impact that legislations have on the push and pull factors of trafficking, and on the incentives for traffickers. First, the granting of asylum in host countries reduces the number of trafficked prostitutes. However, if the demand for services rendered by prostitutes remains unchanged, then the returns to prostitution will rise with the granting of asylum in the countries. Subsequently, it becomes more lucrative for traffickers to target countries that grant asylum to trafficked individuals as potential destinations. On the other hand, a ban on prostitution in the source countries either reduces the supply of prostitutes or raises the price commanded by those involved in the sex trade. Both these factors raise the opportunity income of traffickers. Instead of taking an individual out of a country, traffickers have the alternative to engage the individual in illegal prostitution within the source country. The increased profits from this illegal prostitution strengthen the bargaining power of traffickers abroad, as they can now negotiate a better price for those taken to the host country. In sum, the twin objectives of granting legal rights to trafficked individuals and reducing the incidence of trafficking conflict, unless (i) adequate attention is paid to the perverse incentives of traffickers and (ii) greater efficiency is attained in the apprehension of traffickers in host and source countries.

Child Labor in Debt Bondage Debt bondage is generally defined as the pledge by a debtor of his personal services and/or the services of people under his control as repayment of monetary loans. The adoption of the ILO convention on the Worst Forms of Child Labour represented one among myriad international actions that target child labor resulting from debt bondage. Notable amongst these are (i) law enforcement efforts such as the training of labor inspectors to enforce child labor laws (ILO-IPEC 1997); (ii) direct action assisting and providing funding for governments and local NGOs to liberate children from debt bondage and to provide education, small business loans, and other forms of assistance (InFocus Programme 2002); and (iii) extranational initiatives that condition international trade benefits upon the extent of child labor and bonded labor practices.

62

BASU AND CHAU

Among the extranational initiatives, in November 2000 a $3.5 million program was launched by the ILO to achieve the sustainable liberation of an estimated 75,000 men, women, and children in Nepal from bonded labor through education. The United States instituted a partial suspension of General System of Preference (GSP) benefits to Pakistan in October 1996 specifically targeted at child labor and bonded labor problems.12 Further, alleged violations of labor rights in countries including Myanmar, Guatemala, Liberia, and Maldives have led to a partial or total suspension of GSP privileges. Along similar lines, the European Union also conditions GSP benefits, and, in 1998, began offering additional benefits to countries that apply laws consistent with ILO conventions. The EU further provides for the suspension of GSP benefits in cases where any form of slavery is practiced or where slave labor is utilized, as defined by the ILO and Geneva conventions (Brenton 2000). In the context of debt bondage, what individuals cannot foresee ex ante is the length of time it would take them to break free of the vicious cycle of engaging in loan contracts. From the documented cases of debt bondage in India, Nepal, and Pakistan, where the institution of debt bondage is widely prevalent in rural areas, it is easy to understand why this is the case. Agricultural activity in much of the Indian subcontinent is seasonal in nature, comprising a lean season where little work is available for landless individuals in rural households, and a peak or harvest season where these individuals find work at the ongoing market wage. The market wage in the peak season depends on the tightness of the labor market, which in turn depends on weather conditions that determine how high the demand for labor is, as well as the number of individuals seeking employment. In addition, landless individuals have little or no access to formal credit markets due to a lack of collateral (and hence a high risk of default). Consequently, poor individuals have to meet their consumption needs in the lean season either from their own savings or by borrowing from wealthy landlordscum-moneylenders by pledging their labor services in the peak season. Naturally, the amount of labor services demanded by a landlord depends on the amount that the debtor needs to borrow. Thus, for the very poor with a large gap between their savings and consumption needs, child labor services are pledged along with adult labor services to cover the debt repayments.13 Unfortunately, this gives rise to a cyclical problem. The greater the number of poor individuals in an economy, the higher the amount of 12 See Office of the United States Trade Representative (1997) for a list of GSP suspensions, partial suspensions, and reviews. 13 Basu and Chau (2004) explore the link between debt bondage and fertility decisions. In certain situations, it can be shown that poor households have a greater than optimal number of children with the intention of paying off their debts, which, in turn, depresses the harvest season wage further and makes the households worse off over the long run.

THE WORST FORMS OF CHILD LABOR

63

labor services (adult and child) pledged to the landlord in the peak season. This leads to a low labor demand in the peak season, which keeps the market wage down, preventing individuals from saving and forcing them to engage in debt bondage the next lean season. A careful look at the forces that perpetuate debt bondage underscores an important institutional failure—the lack of savings for poor individuals, which increases the likelihood of exposure to consumption shortfalls in the lean season, when income is uncertain. In this respect, if individuals could either bargain for a better credit-labor contract or negotiate through a labor union higher wages in the peak season, then debt bondage would be a far less pervasive phenomenon. Therefore, we start by looking at the national legislations currently in place in India, Nepal, and Pakistan that provides individuals in rural households the right to negotiate for better wages. India The Bonded Labor System (Abolition) Act of 1976 prohibits all forms of bonded labor. However, debt bondage continues to be prevalent in rural India, particularly in the states of Bihar, Orissa, and Andhra Pradesh (U.S. Department of Labor 1995). Bonded child laborers are reportedly pledged to employers by parents typically in exchange for a loan (U.S. Department of Labor 1995). In a 1979 survey carried out in ten states of India, about 30 percent of families under debt bondage put one or more children into bondage, mostly to cover debt repayment. The situation is compounded by the lack of bargaining power on the part of agricultural workers in the rural areas of India. Indeed, legal protections of worker rights are effective only for the 28 million workers in the organized industrial sector, out of a total workforce of more than 397.2 million (Bureau of Democracy, Human Rights and Labor 1999; Tucker 1996). Nepal Bonded labor is a continuing problem in agriculture in Nepal, even though the Nepal Labor Act specifically prohibits forced or bonded child labor (Bureau of Democracy, Human Rights and Labor 1999). An estimated 100,000 individuals work under the bonded labor system in the southern Terai region. In many cases, the entire family, including children, is involved in performing the tasks required, and bondage with the landlord is expected to terminate only when debt is paid off in full. Like India, legal protection that provides for the freedom of association and the right to unionize is, in practice, limited to those employed in the formal sector of the economy. Collective bargaining agreements cover an estimated 20 percent of wage earners in the organized sector (Bureau of Democracy, Human Rights and Labor 1999). Pakistan The use of slaves and forced labor is outlawed in Pakistan by the Constitution of 1973 (article 11). Meanwhile, the Bonded Labor System (Abolition) Act adopted in 1992 outlawed bonded labor, canceled existing bonded debts, and forbade lawsuits for the recovery of existing debts (Bureau of Democracy, Human Rights and Labor 1999). Despite

64

BASU AND CHAU

these legislations, bonded child laborers are extensively used in laborintensive sugar cane and cotton farms. A possible reason behind the persistence of the bonded child labor phenomenon has to do with farmers routinely incurring debt due to poor harvests, the need to finance agricultural production and meet minimum consumption needs (U.S. Department of Labor 1995), which in turn necessitates the pledge of household labor services as a method of payment. Industrial workers in Pakistan are legally permitted to form trade unions subject to major restrictions in some employment areas. However, there is no provision in the law granting the right of association for agricultural workers. Apart from these national legislations, India and Nepal have also ratified the UN Slavery Convention (1926) and Supplementary Convention on Slavery (1957). Estimates from official government sources and nongovernmental organizations (NGOs) of the number of bonded workers, children and adults combined, in these three countries differ widely, as shown in table 3.11 on the following page. In spite of this wave of campaigns and policy actions, little information on a cross-national basis is available on the empirical correlations between legislative, economic, and financial indicators and the phenomenon of child labor in debt bondage. As a consequence, the impact of liberating households from bondage on the ability of these households to remain free subsequently, remains an open question. The findings follow closely from Basu and Chau (2003). Cross-National Evidence on Debt Bondage We constructed a measure of the reported incidence of child labor in debt bondage, bondchild, based on entries in the Human Rights Report (U.S. Department of State 1999). A value of bondchild  1 is assigned for countries with reported incidences of child labor in debt bondage (table 3.12), and bondchild  0 otherwise.14 In addition, countries are assigned a value of bondchild  0 whenever the incidence of child labor (between the ages of 10 to 14) is reported as zero in World Bank (2001). This is done in order to address a source of sample bias that may be inherent in descriptive reports, involving missing observations or reports, particularly in countries where the problem of bonded labor is nonexistent. 14 The interested reader is referred to Basu and Chau (2003) for the details in the construction of the binary variable bondchild. We do not use observations of bonded child labor that involves the smuggling of children from another country to focus on the source country of debt bondage. The United States is a case in point, where bondchild takes on the value of 0.

26.5% (1.55 million)

14.7% (3.3 million)

Nepal

Pakistan

75%

80.50%

71.50%

% of Rural Population

Yes

Yes

Yes

Legislations Outlawing Bondage

2 million*** Industrial workers only

Kamiya System Human Rights Commission of Pakistan

28 million out of estimated 397.2 million 20% of organized sector

2.6–15 million 100,000**

Industrial

Collective Bargaining

353,000*

Incidence of Bondage

ISSC (1992)

Government of India (1979) Tucker (1996)

Survey

* 30% of families put one or more children to work in order to repay debt. ** Incidence of bonded child labor is higher for landless households. In 46% of the cases, there is more than one child working for a wage. *** Bonded child labor is extensively used in sugar cane and cotton farms.

5.57% (11.5 million)

India

Country

Participation Rates (ages 5–14)

TABLE 3.11 Government and NGO Statistics on Debt Bondage in India, Nepal and Pakistan

66

BASU AND CHAU

TABLE 3.12 List of Countries with Bondchild  1 Albania Bangladesh Benin Bhutan Bolivia Brazil Burkina Faso Burma (Myanmar) Cambodia Cameroon Chad

Colombia Congo Côte d’Ivoire Dominican Republic Egypt Ethiopia Ghana Guinea Haiti India Indonesia

Kenya Laos Liberia Libya Malawi Mali Mauritania Morocco Mozambique Namibia Nepal Nigeria

Pakistan Peru Sierra Leone Somalia Sri Lanka Sudan Thailand Togo Uganda

Observance of Core Labor Rights To address the question of whether a country adequately observes core labor rights, we take into account three distinct indicators. We adopt first a legislative measure for employment relations in agriculture from the International Labour Office (2001). The indicator variable (legexag) takes on a value of 1 whenever child labor in agriculture is exempt from national minimum age legislations. A second measure (enforce) deals with enforcement of labor standards and labor rights. Here, we simply use the four-point score in OECD (2000), A country is assigned a value of 4 if the enforcement of freedom of association and rights to organize is deemed adequate. Countries in which severe violations have been reported have the lowest score, with enforce  1. The final indicator is an outcome measure that gives the share of economically active children between the ages of 10 and 14 (World Bank 2001).

Stage of Financial Development and Credit Market Imperfection In order to measure the degree of access to credit for rural households, and to capture the ease of consumption smoothing, we made use of a number of indicators. Some, such as the variable intspread, which denotes the average interest rate gap and captures the gap between the official lending interest rate and the deposit rate (1994–98) (World Bank 2001), and the variable priv, which denotes the share of private credit (by deposit money banks and other financial institutions) to GDP (Beck, Demirguc-Kunt, and Levine 2000), are taken from existing data sources.

THE WORST FORMS OF CHILD LABOR

67

These capture, though imperfectly, the observed “price” and “quantity” of access to credit market. An additional variable bankfin is taken from the Heritage index of banking and finance (1995–99 average). This is a five-point score indicating decreasing degrees of freedom of the private sector to access banking and financial services. In order to measure the degree of access to consumption-smoothing loans, whether from formal or informal sources, we constructed an indicator riskshare. This is obtained by estimating the extent to which variability in gross domestic product per capita ( log yit) translates to a corresponding variability of household consumption per capita ( log cit) (1970–98, constant 1995 prices). Data on gross domestic product and household consumption per capita are taken from World Bank (2001). For each country i, the variable riskshare is taken to be the estimated least squares regression coefficient
i of the following regression equation: log cit  i

ilog yit
it If
i  0, the household consumption is fully insured from per capita income shocks. At the other extreme, with
i  1, there is perfect passthrough of income variability to household consumption variability. In all, child labor in debt bondage is reported in 43 of the 134 countries for which data are available. As may be expected, child labor in debt bondage is a developing country phenomenon, although the converse is not true (table 3.13). Approximately 60 percent of countries free from observed incidence of debt bondage are in fact low- and middle-income countries (World Bank 2001). In addition, the per capita GDP (1994–98) level of $4489 perfectly separates the two groups of countries, those with and without bondage. As in Krueger (1997), there is likewise a threshold income level dividing countries with and without child labor. Indeed, over 40 percent of countries with reported incidence of children in debt bondage are exporters of non-fuel primary products, with agricultural exports contributing to more than 50 percent of total export revenue. Despite this apparent dependence on agriculture, the value added of agricultural workers is significantly lower in countries where debt bondage exists. In terms of the adoption of core labor standards in the two groups of countries, the average percentage of economically active children (10–14 years of age) is about 23 times higher in countries where children in debt bondage is reported (table 3.14). This is despite the fact that almost all of the countries included in the dataset have adopted international conventions or national legislation on minimum worker age in one form or another. The rights for workers to negotiate wages and form unions are much more popularly observed in countries in which children in debt bondage are not at issue. Surprisingly, legislative exceptions for child labor

GDP Per Capita (constant 1995 US$) 1994–98 (rgdpch) Low and Middle Income Countries Exporter of Manufactures Exporter of Non-Fuel Primary Products Exporter of Fuel Exporter of Services Diversified Exporters Value Added per Worker in Agriculture (US$, 1998) (agavemp) Share of Labor in Agriculture

Variable

12283 0.492 0.459 0.285 0.285 0.383 0.437

16475 12.1

0.604

0.297

0.088 0.088 0.176 0.253

13391

12.06

Std. Dev.

10488

Mean

bondchild  0 Min

0.2

186.3

0 0 0 0

0

0

182.3

TABLE 3.13 Macro-Economic Characteristics and Export Orientation

48.7

53097

1 1 1 1

1

1

47988

Max

42

64

91 91 91 91

91

91

84

Obs

35.76

844.1

0.419 0.07 0.186 0.326

0

1

830.6

Mean

19.92

907.6

0.499 0.258 0.394 0.474

0

0

917.3

Std. Dev.

bondchild  1

1

136.1

0 0 0 0

0

1

109

Min

51.3

4165

1 1 1 1

0

1

4489

Max

5

38

43 43 43 43

43

43

38

Obs

1.465 0.66 3.405

8.962 0.823 2.819 59.47

Observance of Labor Rights Child Labor (% of children 10–14) legexag enforce

Financial Sector Development Indicators Intspread (%) riskshare bankfin priv (%)

Mean

12.12 0.605 1.018 33.6

5.782 0.479 0.964

Std. Dev.

bondchild  0

37.93 1 4

Max

1.203 75.16 –1.039 3.422 1 5 9.438 166.2

0 0 1

Min

TABLE 3.14 Observance of Core Labor Standards and Financial Indicators

53 67 78 35

80 50 42

Obs

10.29 0.926 3.364 21.62

23.22 0.733 2.214

Mean

7.392 0.283 0.838 16.7

13.82 0.45 0.579

Std. Dev.

bondchild  1

3.026 0.36 2 2.766

0.1 0 1

Min

30.78 1.768 5 74.4

52.67 1 3

Max

21 32 39 18

43 30 14

Obs

70

BASU AND CHAU

in agriculture is in fact a common phenomenon regardless of whether debt bondage is reported or not. In terms of the stage of development of financial markets, countries with bonded child labor have on average a larger interest rate gap, a lower GDP share of private credit issued by deposit money banks and other financial institutions, and a lower degree of freedom of private access to credit markets, according to the Heritage index of banking and finance. In addition, countries where no children are reportedly engaged in debt bondage also tend to perform better in terms of the ability of the average household to smooth consumption in the face of per capita income shocks.

Empirical Findings and the Economics of the Redemption of Bonded Laborers We performed a sequence of logit estimations, in order to check for the marginal impacts of economic performance, observation of labor rights, and financial market development on the likelihood of reported incidences of children in debt bondage. Tables 3.15 and 3.16 reproduce the findings of Basu and Chau (2003). Although the degree of statistical significance varies, there are three factors that appear to be associated with an increase in the likelihood of child labor in debt bondage. Specifically, child labor in debt bondage is more likely in countries where real per capita income is low, the rights of workers to negotiate wages and form unions are not enforced, and financial markets are underdeveloped.15 In light of these findings, Basu and Chau (2004) evaluate the economic consequences of granting freedom to those engaged in debt bondage. It is noted that the key reason individuals engage in debt bondage is low income, which in turn is due to their inability to bargain for higher wages or better terms for their debt service contracts. This translates into low savings in any given year, and thus a shortfall of savings to cover minimum consumption needs in the next lean season. Simply put, granting freedom to individuals (children and adults) by paying off their outstanding debts to the landlord does little to affect their income, savings, and vulnerability to debt bondage in subsequent years. Redemption of households generates a temporary positive impact but leaves the long-run incidence of debt bondage unaffected. 15 Dehejia and Ghatti (2002) similarly show a positive correlation between cross-country financial development and the incidence of child labor. In addition, Guarcello et al. (2002) show that credit rationing and income shocks are important determinants of child labor in Guatemala.

71

THE WORST FORMS OF CHILD LABOR

TABLE 3.15 Logit Regression (Debt Bondage) Stage of Development, Core Labor Rights and Financial Development Dependent Variable: Incidence of Child Labor Due to Debt Bondage I lrgdppc intspread

–1.9388*** (–2.46) 0.1663 (0.275)

bankfin

II –2.0658*** (–3.46)

III –2.2702*** (–2.19)

2.3703 (0.55)

riskshare

constant

–2.0426*** (–3.84)

–1.1137 (–0.116)

priv

enforce

IV

–1.7213*** (–2.48) 17.5038*** (2.54)

N 40 Wald Chi2 12.1900 Prob  Chi2 0.0068 Pseudo R2 0.7045 Log Likelihood –6.3026

–0.6750 (–1.43) 20.1275*** (3.08) 55 16.6900 0.0008 0.5937 –12.6779

–0.6355 (–0.79) 17.1691*** (2.98) 41 16.5900 0.0009 0.6408 –7.7504

3.3389* (1.79) –0.0543 (–0.07) 12.8437*** (3.98) 50 18.9200 0.0003 0.6049 –11.7130

* Significant at 1% level. ** Significant at 5% level. *** Significant at 10% level.

Moreover, a stricter enforcement of legislation concerning debt bondage can simply encourage poor individuals to send their children into other (possibly worse) forms of work to generate income for the household. Similarly, cheaper access to formal credit for households reduces the incidence of debt bondage. However, easier access to credit also leaves open the possibility of households overborrowing and then sending their children to work in order to repay their loans. Both these forms of intervention reduce the vulnerability of households to engage in debt bondage but may be unsuccessful in preventing the incidence of child labor per se. Two policies that seem to have a positive effect on eliminating both child labor in debt bondage, and child labor in general, are (i) the enforcement of the right to unionize and collectively bargain for higher wages and (ii) the institution of rural public works programs. For example, the rural public

72

BASU AND CHAU

TABLE 3.16 Logit Regression (Debt Bondage) Stage of Development, Financial Development and Minimum Age Legislation in Agriculture Dependent Variable: Incidence of Child Labor Due to Debt Bondage I Lrgdppc Intspread

–1.8315*** (–4.39) –0.0496 (–0.69)

Bankfin

II

III

–2.6266*** (–4.43)

–2.5824*** (–3.36)

–0.7469 (–0.18)

Riskshare

Constant N Wald Chi2 Prob  Chi2 Pseudo R2 Log Likelihood

–1.7577*** (–4.39)

–1.1930 (–1.59)

Priv

Legexag

IV

–0.2610 (–0.25) 13.7117*** (4.06) 47 19.9700 0.0002 0.5843 –12.5300

–0.9192 (–0.94) 22.5799*** (3.77) 71 22.0600 0.0001 0.6202 –17.7119

–1.8623 (–1.20) 19.9649*** (3.47) 44 14.7200 0.0021 0.7278 –7.6836

–0.6343 (–0.85) –0.2819 (–0.35) 13.1054*** (4.30) 65 19.6900 0.0002 0.5107 –20.9459

* Significant at 1% level. ** Significant at 5% level. *** Significant at 10% level.

works program in India guarantees employment at a minimum wage for individuals unable to find work in the lean season.16 Conclusion In this chapter we have considered two forms of child slavery: child trafficking and prostitution, and child labor in debt bondage. In spite of widespread recognition of the parameters involved in the persistence of these two forms of slavery, policy intervention has so far been piecemeal 16 See Basu (2002) for a discussion on the economics of public works programs in rural India.

THE WORST FORMS OF CHILD LABOR

73

and targeted towards accomplishing the twin objectives of granting human rights to those enslaved and reducing the incidence of these two problems. However, the issue of the worst forms of child labor, particularly child slavery in the two above forms, has received very little attention from economic theorists and empiricists. As we have argued, systematic cross-country analysis underscores the problems inherent in achieving the twin objectives stated above, simply because there are multiple market imperfections associated with the emergence of these two forms of slavery. Instead of granting asylum to trafficked victims, effective policy intervention entails stricter enforcement of trafficking laws and heavier punishments for those involved in trafficking, simultaneously in the host and source countries. In the case of debt bondage, instead of redeeming children by repaying their outstanding debts, a far more effective policy would be to institute rural public works programs during the lean season and enforce the right to unionize and collectively bargain for wages.

Bibliography Baland, J. M., and James Robinson (2000), “Is Child Labor Inefficient?” Journal of Political Economy 108: 663–679. Basu, Arnab K. (2002), “Oligopsonistic Landlords, Segmented Labor Markets and the Persistence of Tied-Labor Contracts,” American Journal of Agricultural Economics 84: 438–53. Basu, Arnab K., and Nancy H. Chau (2003), “Targeting Child Labor in Debt Bondage: Evidence Theory and Policy Prescriptions,” World Bank Economic Review 17: 255–281. Basu, Arnab K., and Nancy H. Chau (2004), “Exploitation of Child Labor and the Dynamics of Debt Bondage”; Journal of Economic Growth 9: 209–238. Basu, Arnab K., and Nancy H. Chau (2005), “Trafficking in Women and Children as a Locational Choice: Theory and Evidence.” Unpublished paper, Department of Economics, College of William and Mary, Williamsburg, VA. Basu, Arnab K., and Katri Seiberg (2005), “A Simple Model of Child Prostitution and Trafficking.” Unpublished paper, Department of Economics, College of William and Mary, Williamsburg, VA. Basu, Kaushik, and Pham Hoang Van (1998), “The Economics of Child Labor,” American Economic Review 88: 412–427. Beck, Thorsten, Asli Demirguc-Kunt, and Ross Levine (2000), “A New Database on Financial Development and Structure,” World Bank Economic Review, September: 597–605. Becker, Gary (1968), “Crime and Punishment: An Economic Approach,” Journal of Political Economy 76: 169–217. Brenton, Paul (2000), “Globalisation and Social Exclusion in the EU: Policy Implications.” Center for European Policy Studies, Brussels, working document no. 159.

74

BASU AND CHAU

Bureau of Democracy, Human Rights and Labor (1999), Country Report on Human Rights Practices for 1998. Washington, DC: U.S Department of State. Chau, Nancy H., and Ravi Kanbur (2002), “The Adoption of International Labor Standards: Who, When and Why,” Brookings Trade Forum, 2001. Brookings Institution, Washington, DC. Dehejia, Rajeev., and Roberta Gatti (2002), “Child Labor: The Role of Income Variability and Access to Credit Across Countries,” NBER working paper no 9018. Cambridge, MA: Bureau of Economic Research. Dennis, Michael J. (2000), “Newly Adopted Protocols to the Convention on the Rights of the Child,” American Journal of International Law 94: 789–796. Diller, Janelle., and David M. Levy (1997), “Child Labor, Trade and Investment: Toward a Harmonization of International Law,” American Journal of International Law 91: 663–696. Ehrlich, I. (1973), “Participation in Illegitimate Activities: A Theoretical and Empirical Investigation,” Journal of Political Economy 81: 521–565. Friebel, Guido, and Sergei Guriev (2004), “Smuggling Humans: A Theory of Debt-Financed Migration,” discussion paper no 1025, Institute for Labor Studies (IZA), Bonn. Grote, Ulrike, Arnab K. Basu., and Diana Weinhold (1999), “Child Labor and the International Policy Debate—The Education/Child Labor Trade-Off and the Consequences of Trade Sanctions,” ZEF-Discussion Papers on Development Policy, no. 1: Bonn: Center for Development Research. Guarcello, Lorenzo, Fabrizia Mealli., and Furio Rosati (2002), “Household Vulnerability and Child Labor: The Effect of Shocks, Credit Rationing, and Insurance,” Report for the Understanding Children’s and Its Impact Project, ILO, World Bank, and the United Nation’s Children’s Fund, Florence, Italy. Humphries, Jane (2003), “Child Labor: Lessons from Historical Experience of Today’s Industrial Economies,” The World Bank Economic Review 17: 175–196. InFocus Programme (2002), Global Report under the Follow-up to the ILO Declaration on Fundamental Principles and Rights at Work Stopping Forced Labour, Geneva: International Labor Organization, http://www.ilo.org/public/ english/standards/decl/publ/reports/report2-html/part1ch6-9.htm. Informal Sector Service Center (ISSC) (1992), Bonded Labor in Nepal under the Kamaiya System. Kathmandu: Informal Sector Service Center. International Labour Office (1998), World Labour Report, International Labor Office, Geneva. International Labour Office (2001), World Labour Report, International Labor Office, Geneva. International Labour Organization (ILO) (2002), “Unbearable to the Human Heart: Child Trafficking and Action to Eliminate It,” International Programme on the Elimination of Child Labour, Geneva: International Labour Organization. International Programme on the Elimination of Child Labour (ILO-IPEC) (1992), Children in Bondage: A Call for Action. Geneva: ILO/UN Centre for Human Rights. International Programme on the Elimination of Child Labour (ILO-IPEC) (1997), Action against Child Labour: Lessons and Strategic Priorities for the

THE WORST FORMS OF CHILD LABOR

75

Future—A Synthesis Report. Geneva: International Labor Organization, http://www.ilo.org/public/english/standards/ipec/publ/policy/synrep97/. Kanbur, Ravi (2003), “On Noxious Markets,” in P. Pattanaik and S. Cullenberg (eds.), Globalization, Culture and the Limits of the Market: New Essays in Economics and Philosophy, (New Delhi: Oxford University Press). Karim, Farhad (1995), Contemporary Forms of Slavery in Pakistan. New York: Human Rights Watch. Kaufmann, Daniel, Aart Kraay., and Pablo Zoido-Lobaton (1999a), “Aggregating Governance Indicators,” World Bank Policy Research Department Working Paper no. 2195. Kaufmann, Daniel, Aart Kraay., and Pablo Zoido-Lobaton (1999b), “Governance Matters,” World Bank Policy Research Department working paper no. 2196. Krueger, Alan (1997), “International Labor Standards and Trade,” in Michael Bruno and Boris Pleskovic (eds.), Annual World Bank Conference on Development Economics, 1996 (Washington, DC: The World Bank). Levitt, S. D. (1997), “Using Electoral Cycles in Police Hiring to Estimate the Effect of Police on Crime,” American Economic Review 87: 270–290. Lipsey R. G., and Lancaster K. (1956), “The General Theory of the Second-Best,” Review of Economic Studies 24: 11–32. Office of the United States Trade Representative (1997), “Monitoring and Enforcing Trade Laws and Agreements—Fact Sheet.” Washington, DC: Office of the United States Trade Representative, Executive Office of the President, September 30. Organization for Economic Cooperation and Development (OECD) (2000), “International Trade and Core Labor Standards.” Paris: Directorate for Employment Labor and Social Affairs, OECD. Protection Project (2002), Human Rights Report on Trafficking in Persons, Especially Women and Children. Baltimore: John Hopkins University. Ranjan, P. (1999), “An Economic Analysis of Child Labor,” Economics Letters 64: 99–105. Ranjan, P. (2001), “Credit Constraints and the Phenomenon of Child Labor,” Journal of Development Economics 64: 81–102. Rogers, C., and Kenneth Swinnerton (2001), “Inequality, Productivity, and Child Labor: Theory and Evidence.” Working Paper, Georgetown University, Department of Economics, Washington, DC. Rogers, C., and Kenneth Swinnerton (2002), “A Theory of Exploitative Child Labor,” Working Paper, Washington, DC: United States Department of Labor. Satz, Debra (2003a), “Noxious Markets: Why Some Things Should Not Be for Sale,” in P. Pattanaik and S. Cullenberg (eds.), Globalization, Culture and the Limits of the Market: New Essays in Economics and Philosophy (New Delhi: Oxford University Press). Satz, Debra (2003b), “Child Labor: A Normative Perspective,” World Bank Economic Review 17: 297–309. Soares, Rodrigo (2004), “Development, Crime and Punishment: Accounting for the International Differences in Crime Rates,” Journal of Development Economics 73: 155–184. Stigler, George (1970), “The Optimum Enforcement of Laws,” Journal of Political Economy 78: 526–536.

76

BASU AND CHAU

Tucker, Lee (1996), The Small Hands of Slavery: Bonded Child Labor In India. New York: Human Rights Watch. UNICEF (2003), “Trafficking in Human Beings Especially Women and Children in Africa,” Florence, Italy: Innocenti Research Center. United Nations (2004), The Seventh United Nations Survey of Crime Trends and Operations of Criminal Justice Systems. Crime Prevention and Criminal Justice Division, United Nations Office. United Nations Interregional Crime and Justice Research Institute (2004), International Crime Victim Survey. http://www.unicri.it/icvs/. United Nations Millennium Development Goals (2000), http://www.un.org/ millenniumgoals/. U.S. Department of Labor (1995), By the Sweat and Toil of Children, vol II. Washington DC: Bureau of Labor Affairs. U.S. Department of Labor (1998), By the Sweat and Toil of Children, vol. V. Washington DC: Bureau of Labor Affairs. U.S. Department of Labor (1999), By the Sweat and Toil of Children, vol VI. Washington DC: Bureau of Labor Affairs. U.S. Department of State (1999), Country Report on Human Rights Practices for 1998. Washington, DC: Bureau of Democracy, Human Rights and Labor. U.S. Department of State (2003), Trafficking in Persons Report. Office of the Undersecretary for Global Affairs. U.S. Department of State Publication 11057. World Bank (2001), World Development Indicators. CD-ROM. Washington DC: The World Bank. World Bank (2004), World Development Indicators. CD-ROM. Washington DC: The World Bank.

Chapter Four Slavery, Freedom, and Sen STANLEY ENGERMAN

Introduction Economic historians have long wrestled with the tensions between useful and measurable economic perspectives on changes in welfare associated with development, such as that denoted by the standard of living, and a broader, less tangible, approach summarized as the quality of life.1 In Development as Freedom, Amartya Sen summarizes his recent thinking on the meaning of development, returning with new insight to this tension in evaluating economic changes. Sen has long rejected more conventional economic interpretations of the standard of living in terms of opulence based solely on material conditions and suggested instead an interpretation in terms of people’s capabilities and functionings. Functionings are the various things that a person may value doing or being. For example, not being enslaved is a valuable functioning, just as is living a life of normal length (or longer) or being healthy. A person’s capability refers to the feasible set or sets of functionings that circumstances allow him or her to achieve.2 As Amartya Sen (1999, 75) says, “capability is thus a kind of freedom: the substantive freedom to achieve alternative functioning combinations (or less formally put, the freedom to achieve various lifestyles).”3 When the Chapter 4 (excluding the addendum) originally appeared (in a longer form) in Feminist Economics 9 (2–3), 2003: 185–211 (http://www.journalsonline.tandf.co.uk/openurl.asp? genre
article&issn
1354-5701&volume
9&issue
2&spage
185) and is reprinted here with permission of Taylor & Francis Ltd and the Journals Association (The International Association for Feminist Economics). 1 For a survey of the earlier debate on the British standard of living in the Industrial Revolution and some proposed expansions of the concept, see Engerman (1994). 2 See, for earlier discussions of these concepts, Amartya Sen (1980, 1993). 3 Sen (1999, 87–110) points to certain important deprivations of individual capacities related to, but not conceptually the same as, low incomes, premature mortality, undernourishment, persistent morbidity, widespread illiteracy, and missing women. All but the last

78

STANLEY ENGERMAN

concern is with capabilities that allow a person to achieve a minimum level of well-being (i.e., to escape poverty), Sen discusses basic capabilities. Sen does not discuss trade-offs among these basic capabilities, although he does give one reference as that between the level of living under slavery and freedom in the antebellum U.S. South. Other writers in this tradition follow Sen’s lead. Martha Nussbaum, for example, after presenting her list of Basic Human Functional Capabilities, says that “the list is, emphatically, a list of separate components. We cannot satisfy the need for one of them by giving a larger amount of another. All are of central importance and all are distinct in quality.”4 The issue of possible trade-offs between economic need and political or social freedoms is, however, of critical importance.5 Sen holds not only that political and economic freedoms are both important but that political freedom is often necessary for satisfying economic needs. He focuses mainly on the complementarity rather than the trade-offs between freedoms, trade-offs that exist not only at low levels of development but also at higher levels of income. The need to deal with trade-offs among basic rights and the importance of ordering the “good things” cannot be ignored. Culturally and temporally specific standards may mean that we disapprove of choices that have been made, but, nonetheless, they may have seemed necessary to those involved. Slavery, voluntary and otherwise, has been a frequent have been considered to be major problems by more conventional measures of welfare, and presumably similar issues of desired lifetime vs. desired healthy lifetime (not just a recent problem as seen in the Greek myth of the difficulties of Tithonus, promised by the gods an eternal life but not an eternal youth); of choices of consumption patterns that yield utility to individuals but at costs in terms of health and life expectation; of group vs. individual decision-making; and of allocations of resources within the family, remain to be analyzed under any of the welfare criteria noted. 4 Martha Nussbaum (1995, 85). Why this “listing limits the trade-offs it will be reasonable to make, and thus limits the applicability of quantitative cost-benefit analysis” is not clear (pp. 85–86). 5 The relationship between political freedom and economic growth has been frequently discussed. Is political freedom a necessary prerequisite for economic development, or, alternatively, does economic growth lead to political freedom? Henry Sumner Maine (1885, 112) poses the problem of whether, with an extended franchise of working-class voters, this would have permitted the introduction of labor-saving innovations in England. There are more general problems of majority rule that he points to, which can lead to loss of minority rights and the refusal to provide rights, including that of immigration, to noncitizens. Another type of trade-off was described by Orson Welles in the classic movie, The Third Man (1949). To justify his own illegal activities in post–World War II Vienna, he points out that the bloodshed, murder, etc., of the Italian states under the Borgias led to Michelangelo, Leonardo, and the Renaissance, while all the Swiss had to show for 500 years of democracy was the cuckoo clock. Welles claims that this elicited a letter from a Swiss official claiming that, contrary to general opinion, the Swiss have never produced cuckoo clocks. See Frank Brady (1990, 450–451).

SLAVERY, FREEDOM, AND SEN

79

occurrence in human societies, generally reflecting dire circumstances faced by persons at the lowest income levels. Even in discussing Sen’s “goods,” trade-offs exist regarding costs, resources, and expenditure allocations. This paper shows that for people living at the level of subsistence, important trade-offs are made between different basic capabilities. The paper is located empirically in the context of slavery, a context which Sen himself has explored. The literature on slavery shows that, at times, people have been forced to make trade-offs between basic capabilities. Slavery, therefore, is sometimes the outcome of people having to make choices between different aspects of freedom, such as between the freedom to be liberated and live free and the freedom to survive and be healthy. Sen’s conceptualization of freedom in terms of expanding capabilities has to confront the issue that people, past and present, have had to make trade-offs between different basic freedoms or capabilities. Examining slavery can give us some insights into the dynamics of those trade-offs. Finally, and again consistent with Sen’s interests, this paper discusses a gender perspective on the nature of the trade-offs explored. The freedoms of male and female slaves were violated and curbed in different ways and to different extents, and the nature of new freedoms and the choices that followed emancipation were likewise gendered. This paper emphasizes the harsh trade-offs between freedoms that slaves faced and suggests that they are echoed in the bitter choices that continue to confront many disadvantaged peoples today, especially women and children. Slavery and Freedom Slavery has taken many different forms; it has been among the most frequent of human institutions, existing in almost all societies in the past and in most parts of the world. Slaves have experienced different work regimes and differences in physical and material treatment, depending on various economic, political, cultural, and ideological circumstances. Nevertheless, the coercion permitted slaveowners has been almost universal, with the limitations on choices permitted to slaves who have always been subject to the master’s control. Lost liberties have often been the outcome of involuntary acts; however, in societies with low and/or highly variable levels of income, people have been willing to sacrifice their liberties and those of their family members in exchange for the ability to survive. At the lowest levels of income, where slavery became a preferred alternative to weakness or death, the conditions of the free were often similarly dire, and moving out of slavery did not mean any material benefits to the newly freed. “Voluntary” slavery has a long and geographically dispersed history, but there is only very limited literature on the topic. Societies with voluntary slavery, as all poor societies, seem to have lacked

80

STANLEY ENGERMAN

the dynamic and wealth characteristics of some of the large societies based on involuntary slavery (see, e.g., Nieboer 1910, 428–30, 437–40). Voluntary slavery is based on an agreement between purchasers and sellers in which both agree to specific terms. While the existence of severe constraints may mean that the “voluntary” choice reflects an absence of opportunities and might be regarded as nonvoluntary due to the limited choices available, there are similar difficulties in describing arrangements made between legally free people where differences in wealth and opportunities exist. (I take these issues up further up in the addendum to this chapter.) One important aspect of slavery, a word which seems without any favorable connotations, is that societies reserve the slave status for outsiders, not members of their own society (see, e.g., Patterson 1982, 7). What societies have considered the definition of the outsider has varied over time; religion, nationality, ethnicity, and race have been utilized. In some cases the outsider characteristic has been socially fabricated, and the discrimination underlying the definition of an enslavable outsider persists even when slavery has ended. The fact that the modern New World slavery was based on race has meant a continuation of racial beliefs after emancipation, serving to limit the gains from the transition from slavery to freedom. While the preexistence of slavery is not necessary for racial, religious, or ethnic prejudice, the contemporary role of racism has certainly meant a significant difference for ex-slaves and their progeny between legal freedom and the achievement of equality. A frequent source of contention is whether it is possible to compare slave and free societies, or if these statuses are so distinct that any attempts to look at them together can only be misleading. In order to determine how feasible such comparisons might be, it is best to focus upon specific questions. To a fundamentalist Marxist, slavery represents a separate mode of production, different from what comes before and after. Forms of institutions and organizations, the nature of decision-making, and the role of technology differ among modes of production, and for some purposes it may be desirable to treat each mode separately. For other questions, however, certain commonalities can make comparisons between modes of production very important. Thus, based on their common humanity, comparing living standards and aspects of the quality of life between slave and free should provide useful information. Descriptions of relative life expectation, health, consumption, and related aspects of the quality of life may serve to cast light on modes of production based on free or slave labor and may help answer questions as to why individuals might choose to move between these situations. In explaining his preference for “a freedom-based perspective on development” (Sen 1999, 28) in Development as Freedom, Amartya Sen raises the example

SLAVERY, FREEDOM, AND SEN

81

of the high material standard of living of U.S. slaves to contrast the value of freedom with the judgments made from “an evaluative system that focuses only on culmination outcomes” (Sen 1999, 28, citing Fogel and Engerman 1974). Freedom implies the ability to make choices, which may include earning less in return for more leisure, less intensity of work, more time with family, more desired geographic location, and so on.6 From this perspective it appears clear that measured income is not to be regarded as an accurate measure of welfare if the achievement of high levels of material consumption is at the cost of actions that can limit individual and family choice. Several questions may be raised about Sen’s argument on this point. Freedom is not easy to define, since it may relate to individual rights or to group freedoms. Even what many consider to be freedom (in contrast with slavery) includes a number of legal and social constraints which, while perhaps less limiting than systems based on absolute government fiat, mean that individuals do not have unlimited choice or are not always treated equally.

Choices and Freedom All people make choices from among available opportunities but under constraints, whether imposed by nature, by other people, or by the self. This choice process is examined in the basic economic model of consumer behavior. Individuals do not all choose the same alternative, reflecting differences in tastes, differences in incomes, and differences in the nature of the constraints faced. Changing constraints with unchanged tastes will generally lead to the selection of different alternatives, as will changing tastes with unchanged constraints. The constraints may be natural, as in the Malthusian limit on the capacity of land to provide adequate food, or they may be social, either deriving from the power of certain groups or individuals or else by some apparently agreed-upon set of enforced codes, legal or otherwise. The market imposes constraints even though people may be legally free to make choices, as long as prices and incomes limit opportunities (Hale 1952; Steinfeld 2001).7 The more limiting the constraints and the fewer the alternatives available, the less free 6 The dilemma was, of course, well known, and goes back a very long time. The Greek atomist Democritus wrote, apparently some time in the middle of the fifth century bc, “poverty in a democracy is preferable to so-called prosperity among dictators to the same extent as freedom is to slavery.” On the gender issue, Democritus claimed (as have many others in later years) that “some men rule cities but are slaves to women.” See Paul Cartledge (1998, 35, 38). 7 For a comparison of labor coercion under slavery and freedom, see O. Nigel Bolland (2002).

82

STANLEY ENGERMAN

we consider individuals, as Sen’s discussion of capabilities would suggest. The fewer the constraints and the greater the choices (even if limits and deprivation remain), the more free we generally regard people. Since all people face constraints regarding choices, we cannot describe individuals as entirely autonomous. And freedom does not always protect individuals, since it may mean freedom to starve, freedom to be beaten by someone else, or freedom to enter uneven exchange. Freedom to choose, with the ability to exchange and transact, and the capacity to behave in accord with one’s chosen values seems preferable to situations devoid of the ability to make choices. Even if freedom is not legally limited, it may be limited based upon restricted alternatives. In Development as Freedom Sen has emphasized the opportunities to choose and to exchange and transact with others as essential aspects of human freedom. The broader the choices that can be made and the less that individuals are coerced and controlled by the state or by other parties, the more freedom exists. The enhancement of choice may reflect either broadened political rights or economic improvement, yielding a larger potential flow of goods and services or of capabilities from which to choose. It has long been argued that not only is freedom a good in itself, but that its occurrence would lead to other benefits, such as greater life expectation, more and better food, and a higher rate of economic growth with benefits for all, than could be achieved by societies with coercion and limited choices. Much of the recent scholarly work on slavery and serfdom has questioned this contention that “all good things come together.”8 This scholarship poses some difficult problems for economic and political theory taken up in the next section. Most economic examinations of the nature of freedom and its benefits presume that incomes are above subsistence (however defined, whether based on physiological or cultural factors) and therefore no consideration need be given to the problem of survival. Even if incomes were above subsistence, the possibility of nonfreedom providing a higher living standard than freedom could present individuals with difficult choices. Yet in many societies, past and present, some individuals have had very low incomes, either permanently or for shorter periods due to famines or natural disasters. In the long discussions concerning the relations between persons, three major types have been discussed: master-servant; parent-child; and manwoman.9 The first two clearly represent dependent relations resting on 8 See, e.g., Sen (1999, ch. 6), and numerous recent World Bank publications. Sen does point to some “dissonances” in outcome, but more frequently argues that there is a “remarkable empirical connection that links freedoms of different kinds with one another” (Sen 1999, 11). 9 See, for this breakdown, Hugo Grotius ([1646] 1925, 231–259), Samuel Pufendorf ([1688] 1934, 839–946), and William Blackstone ([1765] 1979, 410–54).

SLAVERY, FREEDOM, AND SEN

83

coercive relationships which are intended to restrict choice, if presumably for the benefit of the coerced. The distinctions drawn between the choices open for men and for women suggest a similar lack of equality and the relative absence of a decision-making capability granted women. The difficulties of defining freedoms within a social group or a family group still remain, as illustrated by recent debates on such issues as the acceptability of the long-time practice of female genital mutilation in some cultures, defining of the rights of children, and the ability of women to redefine the terms of the marriage contract. Gender, age, and legal status define some of the groups that have confronted limitations on their freedoms. Children are traditionally treated as incapable of making rational choices up to some specified age (the age itself being a major source of disagreement) at which time they become adults able to make their own choices. Women were long regarded as not fully capable of making choices, which accounted for their special treatment in legislation, a condition that still exits in many places. This sometimes meant their being given more favorable treatment than men, although this often meant limitations on their rights to freely choose living and working conditions. Similar types of controls, limiting freedom of action, have been applied to the aged, the mentally disabled, the physically disabled, and convicts and felons, among other groups.10 The most extreme case of constraints on personal freedom is slavery.

Slavery, Freedom, and Living Conditions The complexities of contrasting slavery and freedom often arise from the expectation that “all good things go together,” and that freedom necessarily entails better living conditions and more rapid economic growth than does slavery. Sen (1999, 29) notes that while it is possible that slaves in the United State did have higher material living standards than did free workers, nevertheless “slaves did run away, and there were excellent reasons for presuming that the interests of the slaves were not well served by the system of slavery.” Nevertheless, the low rate of runaways in most slave societies suggests that various means of maintaining the system were used, including force as well as rewards, in the short and long run. The desire to avoid enslavement seems obvious, and the flight of southern U.S. slaves when northern armies moved into the South was not surprising, but under customary conditions in the U.S. South, as in 10 Debates, similar for all these groups, relate to the questions of whether appropriate policies to offset inequality are affirmative action programs, cash grants, or provision of more education, and also concerning the trade-off between short-run and long-run benefits.

84

STANLEY ENGERMAN

most slave societies, the system of controls and compromises, of power and accumulation, was such that runaways as a share of the total slave population were small in number. Unbalanced powers meant a curtailment of the slaves’ ability to leave the plantation and the system. In some societies, although the slaves were freed by law, they had to leave the slaveowner on their own volition. The long periods in which people chose not to leave, and to remain slaves, indicates the complexities of low-income societies. Runaways in the United States and elsewhere were predominantly male, few leaving with women and young children (Heuman 1985; Franklin and Schweninger 1999, 210–13). The low runaway rate for women no doubt reflected physical and cultural factors, particularly the difficulties of leaving with children or an unwillingness to leave children behind, and was not due to any difference in women’s willingness to accept enslavement.11 There were some southern petitions to courts by free blacks to be returned to slavery, usually petitions made by older free blacks, and by women with young children, but these were few in number. An interesting case was recorded in Virginia regarding the petition of a woman to be re-enslaved to the master of her slave husband, “from whom the benefits and privileges of freedom, dear and flattering as they are, could not induce her to be separated” (Phillips 1918, 446). Such requests existed also in other slave societies. In some cases, such as colonial Africa, legal abolition meant only the ending of the state’s willingness to enforce the return of runaways. Since many slaves did not leave their places of residence, de facto slavery persisted. The desire to be re-enslaved or the refusal to accept freedom, while rare in the American South, was generally based on considerations of the benefits and costs of freedom relative to those of remaining in the slave status. Relevant factors included the desire to remain with family members (since the freed often had to move to another state); the belief that better physical treatment remained possible as a slave, particularly for the elderly or mothers with children; and the feeling that, given the conditions in the South, the position of a free person of color was “more degrading, and involves more suffering in this State, than that of a slave who is under the care, protection and ownership of a kind and good master.” The freed black, on the other hand, “lives a thousand times harder, and in more destitution, than the slaves of many of the planters.” In addition, 11 In most New World slave societies, as in much of the world today, slave men apparently had a different trade-off between individual freedom and maintenance of the family than did women. Men have apparently been more willing to leave families behind than have women.

SLAVERY, FREEDOM, AND SEN

85

without the protection of a “trusted white man” the ex-slaves were fearful of kidnapping or violence.12 Recent studies of slave societies have shown that slave economies were capable of experiencing rapid economic growth using a production system of gang labor and had an ability to adjust to changes in world demand for export commodities by changing crop patterns and geographic locations to achieve growth (Fogel 1989, 17–113). It has also been shown that some slaves were provided with consumption levels in excess of those of some members of the free populations in the same country and, as was the case for U.S. slaves, of most populations elsewhere in the world (Fogel 1989, 132–47). Such higher standards of living may have reflected the master’s perceptions of what was needed to secure a greater intensity of work. That skilled slaves were granted higher material compensation while their prices were above those of other slaves indicates that the returns to skilled human capital were divided in some uncertain proportion between slaves and masters. In many cases there was a surplus above subsistence to be fought over, and negotiated divisions between masters and slaves were the outcome. There is a difference, in some important regards, between legal status and economic status. Slaveowners did not always do what they were legally permitted to do (which included rather complete control over the slave’s life and body), although at times they exceeded their rights.

Manumissions and Emancipations The process of manumission provided individual slaves with freedom and gave them certain other rights (but not always all those granted citizens), yet without affecting the status of those still enslaved. Newcomers to a nation or newly freed ex-slaves would not be given the full set of rights that belong to the “established” members of society. Those considered, for whatever reasons, to be outsiders lack key rights and therefore suffer limitations in their freedom. In some cases, freedom can be regarded as a zero-sum game, gains for some coming only at the expense of others, so that measures to increase the freedom of everyone are not possible. While freedom refers to self-ownership by individuals, legal and other constraints may limit the choices open to nominally free individuals. 12 See Ulrich Phillips (1909, II, 161–164; 1918, 446–447); Kenneth Stampp (1956, 92–93); Eugene Genovese (1974, 399–401); Deborah White (1985, 117–118); and Orlando Patterson (1982, 27).

86

STANLEY ENGERMAN

Slaves were always given quite limited legal rights relative to the rights that slave ownership gave the masters or that non-slave, non-slaveholders had, and these slave rights were often not enforceable by law and were subject to the master’s tolerance. Slaves, in most places, did not, however, lack all ability to bargain and negotiate terms of living and working, and thus were able to influence their living and working conditions, either within or outside of the legal framework. Their bargaining power, however, was quite limited relative to that of most free workers. Serfs also had limited rights, but there were more limitations on the lords’ controls over serfs than of the slaveowners’ control over slaves. In most societies, even today, there is a period during which rights of newcomers (immigrants) are limited, particularly those regarding suffrage. Such restrictions can impact on income levels, job opportunities, and the choices available to newcomers in the period before they can become citizens with identical rights to the native born. In the major slave societies of the New World, freedom for all slaves did not occur until the nineteenth century, decades later than the first emancipations in the northern states of the United States. The key dates of the abolition of slavery in the Americas were 1834 for the British colonies; 1848 for the French (besides Haiti) and Danish colonies; 1863 for the Dutch; 1865 for the United States; 1873 for Puerto Rico and 1886 for Cuba, both Spanish colonies; and 1888 for Brazil. In most of these, abolishing slavery took place twenty-five to fifty years after the ending of the international slave trade. The forms of emancipation differed, some being immediate, some requiring considerable periods of time, some freeing all slaves, some freeing only those born after a certain date, and some requiring a period of continued plantation labor by the slaves. In all cases, except the United States, there was some compensation paid to the slaveowners in cash, bonds, or labor time, and in no case was any form of compensation ever paid to slaves (or, in Europe, where the serfs were freed at roughly the same time, to the serfs).13 An understanding of what freedom meant to those enslaved can be seen from observing the patterns of changes in societies after slaves were emancipated. Emancipation usually entailed changes in the institutional structure under which the economy and society operated and also shifts in political and economic power. Legal freedom did not, however, guarantee equality. Central issues such as the need to produce goods, to maintain a family life, and to relate to other members of society existed both under slavery and under freedom, but the changing rules of society with emancipation meant that they could be resolved in rather different ways. 13 See Stanley Engerman (1995), which also includes dates for the former Spanish colonies in South America and Central America.

SLAVERY, FREEDOM, AND SEN

87

The legacy of the past was present, but the availability of new opportunities permitted different outcomes even if some things, such as the allocation of land ownership, did not often change (Engerman 2000).14 A general pattern in most New World slave emancipations was the decline of the plantation system.15 The desire of ex-slaves was to move to smaller agricultural units, whether owned, rented, or labored on, resembling in size and structure the farms of those yeomen who had always been free. These, however, were often less productive than were the plantations, so that the end of slavery usually meant a decline in output, and in some cases it took societies several decades to reach the level of per capita output achieved under slavery. The archetypical case here was Haiti, where emancipation, despite some attempts of the new rulers to bring back a sugar plantation system, ultimately meant the development of an economy based on small farms, producing primarily foodstuff for local markets and with a substantial decline in labor productivity. Wherever small farms replaced plantations, the economic benefits of gang labor were lost. Haiti, once possibly the world’s richest area, today has a measured level of per capita output possibly below the level at the end of the eighteenth century, and is the one country in the western hemisphere to have an income at sub-Saharan levels.16 Ironically, given the successful end of the plantation system, in the twentieth century Haitian migrants produced sugar on plantations in the Dominican Republic and Cuba. There were exceptions to this general pattern, reflecting differences in natural or political constraints in different areas. On Barbados and Antigua, for example, the population density was so high that ex-slaves had nowhere to go and remained on the plantations at least until their external migration increased in the late nineteenth century. In areas such as Trinidad, Jamaica, and British Guiana, the aims of British colonial policy were initially the same, since there were large expanses of unsettled land. Although many of the ex-slaves were able to leave the plantations and establish small farms in the interior, not all were able to do so. In those areas where sugar cultivation remained highly productive, such as Trinidad and British Guiana, the plantation system was restored within 14 The one area with a dramatic change in land ownership patterns was Haiti, where the slaves freed themselves in a revolt and drove out their former owners. 15 Exceptions, in the British Caribbean, were Barbados and Antigua, where land shortages meant the workers had limited opportunities, and Trinidad and British Guiana, which attracted indentured laborers to work on their plantations. Later Cuba attracted immigrants to help maintain the plantation system. 16 Haiti’s initial difficulties were due, in part, to external interferences, such as the lack of desire on the part of most countries to trade with it, and a need to pay, after the 1820s, an indemnity to France in exchange for the opening of trade. Moreover, Haiti has long been politically and economically unstable.

88

STANLEY ENGERMAN

decades, worked not by ex-slaves, but with a plantation labor force consisting primarily of contract labor, for limited periods, drawn primarily from the Portuguese islands, India, and China. Even with contract labor, however, Jamaica was unable to successfully compete in the world sugar market. The decline of the plantation system meant a prolonged decline in measured output as the ex-slaves chose to move to smaller units to produce mainly foodstuffs rather than export staples. Presumably if exslaves had been willing to remain at work full-time on plantations, as perhaps they might have if they were given higher wages and if they had maintained both the same amount of leisure and labor intensity as prior to emancipation, the problem of falling output in these ex-slave societies could have been avoided. Then it would have to be asked what ending slavery had meant for the enslaved. The maintenance of plantation labor had been the goal of planters wishing to maintain profits. It was also the desired policy of many of those in the antislavery movement, who wished to demonstrate that ending slavery and shifting to free labor had no significant negative economic or social impacts and would possibly even result in increased production. In this sense, emancipation would be relatively costless or even beneficial to planters and the other free people in society. Emancipation did have significant legal impacts upon ex-slaves, who were now subject to the same laws as whites. What is less clear in most cases is the impact of emancipation on the material conditions of the ex-slaves, although there are indications that in the U.S. South some material conditions initially deteriorated. Clearly, the ex-slaves gained from being free of slavery and of the plantation system, and, where possible, moving to their own small, family-sized producing units and avoiding the gang labor routines of sugar and cotton production. In avoiding gang labor they made a choice paralleled by that of free whites and free blacks who, centuries earlier at the time of settlement, had chosen to avoid plantation labor as distasteful and to work instead on smaller units. It was the free population’s desire to avoid plantation labor that explained the initial demand for slaves in the Americas. As elsewhere, the ending of slavery in the southern United States meant the decline of the plantation as a producing unit and an expansion of small farms for the ex-slaves, who were still primarily producing cotton. There was, however, a decline in productive efficiency compared to the period prior to emancipation. Indicative of the impact of the efficiency decline in the U.S. South was the increased production of cotton on small farms by the white population, previously unable to compete with the more efficient slave plantation. The end of slavery thus increased the production choices of the white population in the South; it also led to

SLAVERY, FREEDOM, AND SEN

89

greater degrees of occupational divergence between black men and black women than had existed under slavery, as discussed below. The ending of slavery and the freeing of the labor force had a substantial impact in reducing overall output, particularly the output of plantation export crops, in most societies formerly based on slavery. Where it has been possible to prepare estimates, it has been shown that were declines in per capita income of significant magnitudes for several decades. Thus the free labor argument of higher outputs anticipated in the aftermath of slavery argued for by many did not occur quickly, and in several cases it required the importation either of contract labor from India and China or of free European labor from Spain or Italy. What is more difficult to analyze is the effect of emancipation on the consumption level of ex-slaves to determine this aspect of the benefits of emancipation beyond freedom. While the initial changes may not have led to dramatically higher income, freedom permitted the ex-slaves to reap benefits from increased productivity in future years.17 In the U.S. South the first decades of freedom did see increased mortality for exslaves, and there was an overall decline in southern production of foodstuffs. In the West Indies there were declines in food imports, suggesting a lessened availability of foodstuffs there. The estimates of food consumption needed to firmly establish the impact of emancipation upon living standards are still not conclusive, but declines in the initial years would not be implausible given estimates of overall production, and the important fact that redistribution of land ownership did not generally take place, so that some return to landowners continued. Whatever redistribution did occur did not mean that ex-slaves would gain their entire product after the end of slavery. In most places the gang system no longer existed after slavery, and it is probable that the hours and/or intensity of labor fell with freedom. However, even as freedom broadened the options for mobility and labor choice among ex-slaves, the pattern of the U.S. South, where life expectation declined for several decades, food consumption in rural areas was probably lowered, and dietary diseases became more frequent, suggests significant negative material effects. The relative impacts on males and females of the changing life expectation and magnitudes of consumption are not clear, but the nature of living arrangements and provision of consumption goods did change dramatically with the move from plantations and white-owned small units to small farms operated by blacks. 17 Food for slaves could be obtained either by master purchase and provision or by the slaves growing their own food on master-provided plots of land. For a discussion of food provisioning in the British West Indies, see B. W. Higman (1984, 204–218). Under the system of master provision, by purchase, of foodstuffs, females spent more time in field labor and therefore presumably learnt less about small-scale production of foodstuffs.

90

STANLEY ENGERMAN

Female farm labor was more seasonal after emancipation and took place without a gang system. Presumably the black male was now a stronger family leader, and some have argued that for black females all emancipation meant was a change in who controlled them.18 Accompanying emancipation was a southern black striving for education, particularly for children; a desire to purchase and own the farms on which they worked; and a demand for the right to obtain and use the ballot in order to vote for elected officials and to obtain benefits from state and municipal expenditures. The ability to purchase farms and urban dwellings out of accumulated savings after emancipation is one outcome of the drive to labor of ex-slaves, since it is doubtful that they were able to acquire very much low-cost credit in southern postbellum capital markets (Holt 2000, 52–99). Black women were frequently recorded in the census as unpaid family labor, although some worked as paid field hands, so their major contribution to land purchases was indirect. The racist response due to the decline in the cotton market in the 1890s in the U.S. South did reduce the benefits initially gained with emancipation, including education, land ownership, income, and wealth. The behavior and achievements of southern ex-slaves between 1865 and the 1890s suggest what many hoped would be the benefits of freedom. But this phase was ended by the dramatic changes in laws and in education, voting, lynching, transportation segregation, and occupational opportunities.19 For freed black females in the United States, the two key occupations were agricultural work and labor, often in white households, as domestic servants. Down to World War II about 90 percent of black women in the labor force were either laborers in agriculture or domestic servants, with the share of domestic servants rising over time, and accounting for over three-fourths of black female employment in 1930 (Goldin 1990, 74–75). These were low-income occupations, and it was not until after World War I’s movement north by blacks, and then World War II, that a broader range of occupational choices became available for black females. Domestic service often required living with whites rather than with blacks and imposed some limitations on marriage and fertility. Prior to emancipation United States slave fertility was unusually high, both compared to other slave populations as well as to all free populations outside the United States. After 1880 there was a decline in black fertility, roughly in tandem with the decline in white fertility (Engerman 1977). There was 18 For a discussion of the role of slave women in the antebellum South, see White (1985). On the changing role of women with emancipation in the British West Indies, see Bridget Brereton (1999). 19 Lynchings, as a form of social control, were almost exclusively perpetrated on males. Stewart Tolnay and E. M. Beck (1995, 269) estimate that only 3 percent of all lynchings of freed blacks were of females.

SLAVERY, FREEDOM, AND SEN

91

probably some decline in the percentage of black females married, an increase in the age of first birth, and an increase in the number of black childless women (Engerman 1983). Whether this reflected an increased control of birth and marriage patterns by black females or, as has been argued, an increase in diseases is not clear, but for United States blacks the end of slavery meant declining fertility, whereas in other parts of the Americas there may have initially been some increased fertility after emancipation (see, e.g., Roberts 1957, 216–72; Higman 1984, 347–73). In the U.S. South, as elsewhere where slave emancipation occurred, there were significant changes in family arrangements and gender relations resulting from the shift from plantations of several families to residence on small farms operated by individual families.20 There were declines in fertility after emancipation, in part because of the increase in the free family’s costs of raising children to adulthood with the ending of the plantation’s collective child-care arrangement and also the loss of the implicit subsidy to child-raising costs made by plantation owners. In most slave societies females had been valued at about 80 to 90 percent of same-aged males in the same work category, although until age 15 females were equal in price to males (Moreno Fraginals, Klein, and Engerman 1983). The wage differentials by sex among freed people were often larger than the price and hiring rate differentials had been under slavery. Much of the domestic work, such as food preparation and cleaning, had been undertaken during slavery by a limited number of specialized slaves, and a large portion of slave females specialized in field work. After emancipation household work was done in one-family households by women allocating their time among several different functions. Black women now had work patterns that, for whatever reason, resembled those of white females, particularly those of the working class. They spent more time outside the labor force, working in the home, and less in the field than when enslaved, and female wage earners most frequently worked as domestic servants. After emancipation in the United States South, ex-slave families frequently followed the pattern of two-parent households (whether coresident or in separate residence) which had existed under slavery, a pattern which was to change dramatically in the last half of the twentieth century, when the proportion of female-headed households, particularly in urban areas, increased. Whatever the explanation for this recent rise in the proportion of black female-headed households, it is of interest 20 The precise effects on fertility of the antebellum pattern of some slave couples living on different farms, or with separate residences, were, however, unclear. Under slavery, particularly in Virginia, there were cross-unit marriages with visiting, etc. Yet the areas in which these occurred seem to have been regions of high fertility.

92

STANLEY ENGERMAN

to note that these current levels did not occur until much later than emancipation. In recent years there have been some reinterpretations of the role of black women under slavery and afterwards. The labor force participation of slave women in plantation societies was greatly in excess of that of free white women, and this high participation may have reflected patterns of agricultural production in Africa (see, e.g., Robertson and Klein 1983). While the basic family pattern that emerged was structurally similar to that of white families—predominantly nuclear households—the nature of slavery and production patterns led to significant differences in actual arrangements in regard to food provision, child care, and power relations in slave families from those of whites. The relative equality in “earnings” of males and females under slavery suggested a more independent role for slave women than for free women, who were dependent upon the incomes of males. After emancipation there was some carryover of this labor force pattern, with significantly higher labor force participation for black than for white women and a higher proportion of unmarried black women, with or without children. To the extent that there was greater financial independence, black family relations, even with a predominantly nuclear household structure, need not have been identical to that of whites. The discussion among historians and other scholars of the nature of power relations between males and females in the black family and society has recently reemerged, regarding both what did happen and what would have been desirable. The black slave female was subject to the power of her white master, and possibly also of her husband in those slave families with a dominant male role. Ending slavery, which weakened white controls, need not have meant a shift in the power balance vis-à-vis male heads of households (Jones 1985; White 1985). The normative debates relate to the degree to which females should support males in order to improve black living conditions and to what extent females should take advantage of their relative financial independence to exercise power in their own interests, rather than to let themselves be dominated by males. This dilemma is, of course, similar to that confronted by many whites, but the background of lower incomes and the history of discrimination make this problem more acute among blacks. The study of the link between the black family under slavery and its contemporary circumstance is somewhat puzzling. Both abolitionists and proslavery defenders had believed that the slave family was weak, whether they thought it was because it consisted of slaves or of Africans (Fogel 1989, 162–186). The post–Moynihan Report arguments about female-headed households and the number of illegitimate births after emancipation drew upon these earlier debates, even though it took almost

SLAVERY, FREEDOM, AND SEN

93

one century after emancipation before the current rates of female singleheaded households began to appear. It is the dramatic change combined with such a long lag after slavery that makes positing the usual links problematic. That the recent sharp increases in single-parent households occurred in a time of economic improvement is also puzzling. Works such as Herbert Gutman’s (1976, discussed in Engerman 1978) point to the existence of a two-parent household under slavery, a point made earlier by the sociologist E. Franklin Frazier before he became influenced by observing the economic and family difficulties of the 1930s. There were other arguments about the impact of slavery on males and females such as those of W. E. Burghardt Du Bois (1909). Du Bois contended that the so-called “absent father” of the slave family need not have been physically absent, but he was forced to play a much weaker role in the family than did white fathers because of the controls imposed by masters. With the probable psychological costs to female slaves from forced sexual accessiblity by their masters, the two-parent slave household did not function in the same manner as did the free household. In the West Indies the single-parent household was more important in magnitude even at the time of emancipation, and its levels have long exceeded those in the United States, as has the share of illegitimate births (Roberts 1957, 263–306). In neither the United States nor the West Indies did the pattern of black family and gender relations both during and after slavery mirror the patterns in Africa prior to the transatlantic movement.21 Thus freedom, leading to variations in constraints, did have significant effects, but not all its effects were evaluated positively by contemporaries and by subsequent scholars.22

Conclusion This essay has used the systems of slavery and the transition from slavery to legal freedom of the previously enslaved to examine some of the issues raised about capabilities and freedom in Amartya Sen’s Development as Freedom. The discussion of slavery reminds us that over the past two centuries human capabilities, as described by Sen, have increased in large measure because of the worldwide ending of slavery as a legal institution. 21 The effect of planter policy on fertility is difficult to resolve. There were apparently more pro-natalist measures in the British West Indies than in the United States, possibly because they were thought more necessary there, given the lower fertility. Second, the decisions of women in one-parent vs. two-parent households may lead to differences in fertility, whether in slavery or in freedom. 22 Thus it is debated whether the prevalence of the mother-headed household reflects a desired outcome or rather is the outcome of various forms of social problems.

94

STANLEY ENGERMAN

While there is a broad range over which the different types of freedoms discussed are complements and mutually supportive, there are conditions, generally at low levels of income, where trade-offs between such freedoms might become necessary. The existence of voluntary slavery illustrates this. In some cases, freedom meant a lowering of the material well-being, health, and living conditions of ex-slaves. Current-day cases, where women remain in abusive marriages to obtain material consumption for themselves and their children, have been argued to demonstrate the operation of an analogous form of trade-off. A world that requires trade-offs is a rather difficult one for individuals, families, and society, and may require extensive state intervention to correct these adverse outcomes. Insuring increases in the capabilities of those at low levels of income, whether by influencing the market process or by using government funds or power to support activities such as providing resources to the poor, can permit positive social benefits. Legislation ending labor coercion, as well as laws permitting interference within parentchild, husband-wife, and master-servant (labor) relations, might also be regarded as necessary. The need for societies to choose between the rights of individuals and long-standing cultural practices raises a fundamental problem about which disagreements still persist. In Development as Freedom, Sen has raised these and other issues of importance for consideration and for resolution. In going beyond the use of income as a primary measure of welfare, Sen has broadened our concerns with social betterment and with the policies necessary to achieve that goal. Addendum: Voluntary Slavery I Slavery in the United States, and in the Americas more generally, was clearly involuntary, since at no step in the enslavement process from inland Africa to the societies of the Americas were individuals willing to become, or desirous of being, slaves. This is the basis, in another context, for the claim for reparations to the formerly enslaved or to their descendants. Yet it is useful when studying slavery to consider that at various times there had been forms of voluntary slavery of individuals or family members, and that voluntary enslavement played an important role in many societies. People were willing to accept the terms of slavery, either for a price paid to those family members still free, or else at zero cost, agreeing to serve as slaves to others for limited periods or for life, in exchange for the provision of subsistence. While this may be done in times of social chaos for protection and defense, the more frequent basis for voluntary slavery was to avoid hunger and starvation, whether due to long-term persistence of low income, or, more cyclically, to avoid the impact of harvest failures and famines. Freedom may be highly desired by

SLAVERY, FREEDOM, AND SEN

95

individuals, but as discussed in the foregoing, so is survival for themselves and their children. Voluntary slavery which existed in many times and places was used by different societies as a solution to perceived problems of overpopulation, and it was applied to adults as well as to children. In western Europe, however, where enslavement of other Europeans had been ended by the fourteenth century, excess births for the next centuries meant either abandonment, giving children to the church, or infanticide. Abandonment, at times, was based on socially accepted patterns that involved the expected retrieval of the children by members of the church or by individuals who would then provide the abandoned with permanent arrangements in exchange for labor. This was, in effect, a transfer of the rights to the child and its labor, at a zero selling price. Voluntary slavery declined when incomes rose, as starvation became a more limited threat, and when more successful relief and welfare institutions by the state, the church, and by individuals were devised. It is clear, however, that under certain conditions the loss of freedom could be permanent, and was considered acceptable not only to the enslavers, but also to the enslaved. The distinction between legal status and economic status, as well as between slave and free, was a central element in the debates on slavery in British colonies and in the United States in the early nineteenth century. It was claimed by proslavery advocates as well as by British workers (whatever their stance on abolishing slavery) that employers in Britain were able to force free workers to endure harsher working conditions, earn less income, and work longer hours than those of the West Indian slaves, backing these contentions with data from the two societies. And when it was argued that the British should encourage production of sugar by legally free laborers in India in the 1820s to replace the use of slave labor in the West Indies, the presumed benefits of free labor were questioned by pointing to the extremely low agricultural wages in India, and comparing free labor income in India with the consumption allowed slaves in the West Indies. It was not argued, of course, that all slaves had high standards of living or that many free workers might want to become slaves in order to benefit from improved living standards, but these cases do indicate that under certain rather important historical conditions slaves may have been better off materially than some free workers. This was so, even though the slaves obviously had the more limited legal freedom of choice. II While we generally regard slavery as a condition originating in compulsion and coercion, in many societies in Asia, Africa, and in pre-modern Europe, slavery was voluntarily entered into. Because of the low levels of

96

STANLEY ENGERMAN

income in these societies, sales of children and adults to the wealthy were seen as the only way for individuals to survive. While the legally voluntary nature of self-sale by adults is easily understood, the legal complexity of sales of children by their parents has a more uncertain status. The power of parents over children was a debated issue in early law codes, with little agreement on the conflict of the rights of the child (at whatever age) with the desire of the parent to insure the child’s survival, even if in a new residence and under rather adverse working conditions. Nevertheless, the policy of selling children into slavery in conditions of hardship is one that has long been implemented. As far back as the ancient Near East, it has been claimed by Mendelsohn (1949, 5, 14), “lack of employment, or debts, drove people to sell first their children and then themselves into slavery,” and these accounted for the “bulk of the Babylonian and Assyrian slaves.” He argues that this was due to being “driven from the soil by war, famine, or economic misfortunes,” in the “absence of any state or communal help.” The Hammurabi Code (Silver 1995, 118, quoting from paragraph 117) included a provision that “besides being able to borrow on personal security, an individual might sell himself or a family member into slavery,” although in this case, only for a limited number of years. Silver (1995, 117–122) notes the frequency of self-sale contracts in many parts of the Middle East. And, as occurred elsewhere later on, members of a family could be offered into slavery to pay off a debt or as punishment for a crime committed by an individual. In parts of ancient Greece “debt bondage” and the “enslavement or the sale of one’s children” was practiced, as it was elsewhere in the ancient world. According to Ste. Croix (1981, 163, 169–170), this occurred as “the result of extreme poverty” or of debt. In Rome, at times of difficulty, people would sell themselves or their children into slavery, since “a sheltered and tolerable slavery may be preferable to a precarious existence in freedom and poverty.” Under Roman law, according to Barrow ([1928] 1968, 2, 9–12), fathers were given the right to sell a child, and this did occur “especially in times of stress” in order to relieve the consequences of poverty (see also Bradley 1994, 35). While exposure of newborns was permitted, it is probable that self-sale was legally accepted only for the enslavement of war captives. In the nineteenth century the existence of slavery was described as “the Indian Poor Law” for its role in helping to avoid starvation. “In times of general starvation which periodically afflicted most parts of India, parents were often compelled to choose between allowing their children to die or entrusting them to someone else’s care” (Temperley 1972, 96). In certain areas of early-nineteenth-century British India it was, according to Banaji (1933, 44–52), considered that “the sale of children by their

SLAVERY, FREEDOM, AND SEN

97

parents or relatives, from poverty and inability to maintain them in times of famine or of other general calamity, was the most prolific source of slavery and the origin of almost the whole slave population.” In China the traditional pattern was that “destitute parents, especially during bad harvest years,” sold children, mostly females, since males were the source of parental support in their older ages (Watson 1980, 13). This was not an infrequent occurrence into the nineteenth and twentieth centuries, given the number of famines due to droughts and bad harvests that China experienced. Similarly, in southeast Asia voluntary slavery was frequent, reflecting times of famine, shipwrecks, and conditions after the volcanic eruption of 1814 when “people would ask to be accepted as bondsmen in return for food.” Indeed, according to Reid (1983, 159), “the ease with which these bonds were contracted was one of the things which most astonished foreign visitors to Southeast Asia.” In the Americas before Columbus, among the Aztecs (and the Mayans before them) there was self-sale, or sale of children or family members, “in time of hardship” to “secure survival.” This is described by Clendinnen (1991, 99–100) as a circumstance in which “slavery provided a social net” “for those suffering gratuitous misfortunes,” but also for those in this society who were regarded as “chronically shiftless individuals.” She, however, rejects the contention that these were the slaves who were used for human sacrifice, claiming that activity was reserved for outsiders received as tribute or war captives. Lovejoy’s (1983, 70, 149) discussions of African slavery points to cases where, “in times of famine,” children were sold and “self-enslavement to avoid starvation and the sale of children occurred to avoid the consequences of drought and famine.” In their introductory essay to their edited volume Slavery in Africa Miers and Kopytoff (1977, 12) note “children and even sometimes adults were bartered for grain in times of famine to save the rest of the group.” Cooper’s (1977, 126) description of slavery on the east coast of Africa in the second half of the nineteenth century points to the large increase in the slave trade in 1884 due to “a devastating famine” which made people “desperate enough to sell their neighbors, their children, and even themselves in order to survive.” Even later, in Africa, it was claimed that slaves themselves resisted “free” status when the latter resulted in lower standards of working and living, since some slaves did earn more than did the free population. Campbell and Alpers (2004) point to a case where slaves refused manumissions in order to maintain “their traditional privileges” and to avoid coercion doing “state imposed forced labor.” Self-sale and sale of children at times of famine thus occurred in parts of Africa in the eighteenth, nineteenth, and twentieth centuries, if not earlier.

98

STANLEY ENGERMAN

Voluntary slavery was made legal in Russia in the second half of the sixteenth century, to be applied when people lacked independent means. Voluntary slavery was generally entered into in response to crop failures or crop destruction that caused food shortages and famines, and served as a form of “welfare relief.” Hellie claims, however, that sales of children by Muscovite parents were “very rare,” the preferred (to parents, if not children) option in these circumstances being “to let them die.” He argues, “slavery was necessary” given the general inadequacy of “alternative relief or welfare institutions” (Hellie 1982, 39–41, 371, 377, 693–695). III Many of the early modern European advocates of liberty, including Samuel Pufendorf and Hugo Grotius, accepted voluntary adult and child slavery, but only if this was the best way to keep alive individuals who were in poverty, disabled, or otherwise incapable of earning a subsistence living. Jean Bodin ([1606] 1955, 14, 16, 17) noted the possibility that some people did become slaves “because they have sold or gambled away their liberty to another.” Although he pointed to the long-time and widespread existence of slavery in all parts of the world except Europe, he does argue that it was “contrary to nature.” He anticipated Condorcet in his pointing to the “base humiliations” and “cruelties” faced by slaves, and the difficulties these created in controlling slaves, also posing problems of adjustment after emancipation. Hobbes ([1657] 1968, 273) dealt with the issue of the enslavement of war captives, arguing that a person taken prisoner or threatened militarily can accept the condition of slave and become “the subject of him that took him; because he had no other way to preserve himself.” Such persons had the option of not accepting submission, and running away. Grotius ([1646] 1925, 103, 233, 255) argued that a father’s rights include that of selling a child “if it is necessary and there is available no other means of supporting him.” A person who enslaves himself “owes lifelong service in return for nourishment and other necessities of life” as “the lasting obligation to labor is repaid with a lasting certainty of support.” Grotius claims the latter benefit is often not granted those “who work for hire by the day.” In a more political context, Grotius goes further, to claim that “to every man it is permitted to enslave himself to anyone he pleases for private ownership, as is evident from the Hebraic and from the Roman Law.” Grotius, as did others of the time, wrestled with the issue of prisoners of war, but accepted the enslavement of war captives provided rather strict rules protected their treatment.

SLAVERY, FREEDOM, AND SEN

99

Pufendorf ([1688] 1934, 921, 936) argued that “In no case does it appear that nature allows a father to use his son as a pledge, or sell him, unless there is no other way to support him.” In that case slavery should be “easy,” and there should be some hope of becoming free. This would make enslavement provide a better outcome than if “he starve to death.” Pufendorf also claimed “the beginnings of slavery followed upon the willing consent of men of poorer condition, and a contract in the form of ‘goods for work,’ ” thus, “I will always provide for you, if you will always work for me.” Locke ([1690] 1963, 325–326), in contrast to Hobbes, argued that since man does not have power over his own life, he cannot “enslave himself to anyone” or “take away his life.” Yet, he went on to argue, when enslaved if “he finds the hardship of his slavery outweigh the value of his life, ’tis in his power, by resisting the will of his master, to draw on himself the death he desires.” Locke, in his drafting a constitution for the Carolinas, used the argument of a just war to justify Negro slavery. They had “forfeited their lives,” and being without property, were outside of civil society. Rousseau ([1762] 1947, 7, 8) argued against the possibility of a man enslaving himself, saying the “act is null and illegitimate,” since whoever would attempt this was “out of his mind.” The sale of children was also forbidden. He also redefined the issue of enslaving war captives, since there was, he argued, no “right to kill the conquered . . . deducible from the state of war.” Yet Rousseau (9) did comment that a state at war with an enemy “has a right to kill its defenders, while they are bearing arms,” but not once they surrender. Blackstone, in his Commentaries on the Laws of England ([1765] 1979, 411, 412, 127), argued that slavery is “repugnant to reason, and the principles of natural law,” and should not exist anywhere. Although the Justinian Code allowed the person’s sale of himself, this could not be a valid contract, since there is no equivalent quid pro quo “for life, and liberty.” He also argued against the customary argument justifying enslavement for war captives, since there was no longer a right to kill captives. Blackstone explained the absence of any acceptable form of voluntary slavery, since the English law “furnishes him with every thing necessary for their support.” “All the necessities of life” would be provided by “means of the several statutes enacted for the relief of the poor,” and based on “the more opulent part of the community.” There was, in Blackstone, no belief in the Aristotelian natural slave, a slave due to his inferiority— physically, mentally, and morally. Condorcet ([1781] 1999, 308, 313) regarded slavery as a crime “far worse than robbery.” He denied the possibility of self-sale into slavery, that being an infringement of natural law. There were, however, conditions under which a person was “incapable of exercising his rights” and

100

STANLEY ENGERMAN

if he would do so, he “would misuse those rights against other people,” and therefore his rights could be taken away. These other people include “young children,” and “idiots and madmen.” In discussing policies towards emancipation, he argued that so brutal was slavery and serfdom that the freed would be a threat to themselves and to others if emancipation were immediate, and they would be reduced to poverty. To Condorcet, ironically, the more brutal the slavery, the longer that slavery must exist prior to full freedom. Montesquieu made the argument that people could not sell themselves, basically because there was no enforceable quid pro quo possible. Nor could a person kill himself, since it would be a loss to the homeland. In discussing some cases in which people sold themselves into slavery, he explained a comment by a contemporary “ that the Muscovites sell themselves easily” as being “because their liberty is worth nothing” ([1748] 1989, 251). (Voltaire [(1768) 1994, 136–137] however, presented a different interpretation of the Russian case that dismissed Montesquieu’s argument.) IV The need (and desire) for survival throughout history led to sacrifices of freedom, whether by sales of children by parents, sales of wards by guardians, sales of adults by family members, self-sale, or by voluntary acceptance of enslavement without any payments. In these cases, presumably freedom would have meant being free to starve rather than free to choose. Similarly, late-nineteenth-century movements of indentured labor from India to the sugarplantations in the West Indies (and elsewhere) apparently were influenced by the extent of famine and poverty in India, out-migration increasing at times of harvest failures. Thus the wistful comments of the Trinidad Government Emigration Agent in Calcutta in 1904: “Recruiting conditions and prospects at present are abnormally bad in almost every district. The recent good harvest had done away with the chief inducement of laboring classes to emigrate and recruits are almost unobtainable,” (Great Britain, Colonial Office, 295/430 letter of March 10, 1904). And their minutes of December 22, 1904, commented that: “It is unfortunate that the Emigration Agencies must view with alarm a good harvest in India.” The desire to be re-enslaved or the refusal to accept freedom, while rare in the American South and other slave societies, was a decision made on consideration of the benefits and costs of freedom with those of remaining in the slave status. This could be based on their desire to remain with family members (since the freed often were to move to another state).

SLAVERY, FREEDOM, AND SEN

101

Others included the belief that better physical and political treatment was possible as a slave, particularly for the elderly or mothers with children, as in the 1857 Georgia situation where it was believed that “a kind master was better able to provide and care for her than she was herself.” The newspaper writer noted that the situation was “a nut for Yankee philosophers to crack” (Phillips 1909, 162; see also pp. 84–85, above). Genovese (1974, 399, 401) comments that in the “late antebellum period and early war years, pressure mounted to induce free Negroes to reenslave themselves and to restrict their rights further.” Few did, “but that some did suggests that often their legal and economic position was so precarious as to throw them on the mercy of a trusted white man.” By 1860, nine southern states had legislation “allowing free persons of color to enslave themselves” (Morris 1996, 31–36). Stampp (1956, 92–93, 216) argued that some Negroes might decline emancipation or seek a master since “given the general hostility toward free Negroes in both the North and South and the severe handicaps which they faced, the choice between ‘freedom’ and slavery did not always seem to be an altogether clear one.” He states that the numbers reenslaving themselves by turning down emancipation was small, even though most southern states during the 1850s “adopted laws authorizing the voluntary enslavement of those people and enabling them to select their own masters.” V The problem of voluntary slavery, and its long and frequent existence, does raise a number of important issues for the examination of slavery and emancipation. Were voluntary slaves treated better and made to suffer fewer psychological effects than were those forced into slavery? Patterson (1982, 27–28) states, “Voluntary servitude, however, was not slavery,” but the case is made based on the need for protection within Ashanti society. In some cases of voluntary slavery children inherited the slave status, but not always. Should, on theoretical grounds, voluntary slaves be eligible to receive reparations, or should they reimburse their owners? Since voluntary slavery could lead to labor exploitation, should voluntary slaves then obtain the same benefits as involuntary slave labor? The nature of voluntary slaves poses the important question of the tradeoffs between freedom and survival at low incomes, and the difficult choices imposed in societies where individuals received low incomes. The income levels of individuals will influence their behavior and helps define the meaning of freedom. Voluntary slavery has been one response to low incomes. Others include, government loans or expenditures, subsidized out-migration, and social redistribution schemes by individuals and

102

STANLEY ENGERMAN

associations. Political philosophers including Grotius and Pufendorf have argued further that extreme want would justify theft and as the early nineteenth-century lawyer Thomas Rutherford argued “necessity sets property aside” (Novak 1996, 72). In a broader sense, the overall social problem goes beyond the legal problems of slavery and freedom to examine those legal and other influences upon the relative incomes and statuses of different individuals and the perceived need for forms of redistribution to influence relative incomes, consumption, and welfare. How can the impact of the historical past be separated from current circumstances influencing income, and what differences are considered desirable for society to offset? Some of the current circumstances of families and individuals were influenced by decisions of their predecessors. Decisions as to whether to save (and invest) or to consume goods rather than enjoy leisure affect the incomes of future generations. Similarly, decisions concerning the number of children to have and raise, and the extent of the education to be provided them, will lead to differences among individuals in subsequent generations, that some believe society should eliminate. It has been argued by some that the population is owed, not only for slavery, but also for losses attributed to other forms of social organization. These losses can be economic, psychological, and cultural. Given legal systems, political power, and disparities in income, it is argued that since inequity in starting points is quite widespread it is up to society to equalize in the present. Thus our current debates about “socially necessary minimum income,” “living wages,” improved working conditions, and improving provisions of education, reflect concerns similar to those in the current discussions of slavery and the past and present debates on policies for social redistribution. VI According to most dictionaries, there are several different, not always overlapping, definitions of “redeem.” The ones of interest here include: to buy back, to make payments or, to free, to ransom, or liberate from bondage or captivity, and to recover by payment. In the present context the discussion concerns payments made to purchase and free slaves, the payments being made to the owners of slaves, at some agreed upon price. The payments could be made by the enslaved or his kin, the government, or by third parties interested in liberating the slave (generally some antislavery group). The basic argument for redemption payments is to free a slave, and to ending his lifetime in that coercive condition. The achievement of freedom, whether one at a time or of all slaves at one time, is considered an

SLAVERY, FREEDOM, AND SEN

103

improvement in human welfare, and is to be encouraged. There are, however, several arguments against redeeming payments, both moral and practical. If slavery is regarded as robbery, being the theft of an individual, any payment is a reward to criminal activities, a reward to those performing the crime and not to its victims. Given, however, that slaves in most societies were regarded legally as property, and the current owners are not necessarily those involved in initial enslavement, recognition of property rights has generally been the policy followed at the time of emancipation. A more practical argument against redemption is that if done on a one at a time basis, this increase in the demand and the reduction in the supply of slaves will raise the price of slaves and make more enslavements profitable to sellers. Thus the number of slaves need not decline, new slaves replacing those freed. Central to the redemption scheme must then be a mechanism to prevent any new enslavement. To place these points about redemption in context, it will be useful to look at the means by which past slavery has been ended. Individual freedoms have been the result of grants by owners, by self-purchase, or by purchases by third parties, whether antislavery organizations or, in the U.S. case, colonization societies. Manumissions were frequently accomplished by the right to self-purchase, not a grant of freedom by owners. Even in these societies in which manumission was more frequent than in the U.S. and British cases, it often entailed a payment by the slave out of his savings. Manumissions could be immediate or gradual, depending on the nature of the payment schemes. Runaways, if successful, were freed without redemption payments, a theft of self in response to the prior theft of enslavement. As seen in the case of Fredrick Douglass, runaways were liable to recapture if within the United States, and the only way to fully guarantee freedom was to redeem the slave from his owner. In the Americas and elsewhere, most slave societies ended slavery, not by individual manumissions, but by legislation or, in the case of Haiti, by revolution. In most cases it was by the legislative ending of slavery, which came with some form of compensation paid to slaveowners, and no payment to slaves (see Engerman 1995). Similarly, the ending of European serfdom meant benefits to landowners and not to the serfs. The form of payment reflected the belief in these societies in the sanctity of property rights, as well as the nature of political power when legislation was passed. Compensation or redemption payments took several forms. It could be in the form of cash, bonds, or labor time. Under forms of gradual emancipation there was some compelled labor time that provided labor to the owners. A frequent form of gradual emancipation was to consider all born to slave mothers after a certain day as free, but required to remain with the mother’s owner for 15 to 30 years, depending on the state or country. This was done to permit the owner to recover the costs of

104

STANLEY ENGERMAN

raising the slave by receiving returns during his productive years. In this sense the slaves were redeeming themselves by providing labor services for some prolonged time, thus minimizing costs for taxpayers.

Bibliography Banaji, D. R. (1933), Slavery in British India. Bombay: D. B. Taraporevala Sons. Barrow, R. H. ([1928] 1968), Slavery in the Roman Empire. New York: Barnes & Noble. Blackstone, William ([1765] 1979), Commentaries on the Laws of England, vol. I: Of the Rights of Persons. Chicago: University of Chicago Press. Bodin, Jean ([1606] 1955), Six Books of Commonwealth. Oxford: Basil Blackwell. Bolland, O. Nigel (2002), “The Hundredth Year of Our Emancipation: The Dialectics of Resistance in Slavery and Freedom,” in Verene A. Shepherd (ed.), Working Slavery, Pricing Freedom: Perceptions from the Caribbean, Africa, and the African Diaspora, (Oxford: James Curry), pp. 320–339. Bradley, Keith (1994), Slavery and Society at Rome. Cambridge: Cambridge University Press. Brady, Frank (1990), Citizen Welles: A Biography of Orson Welles. New York: Scribner’s. Brereton, Bridget (1999), “Family Strategies, Gender and the Shift to Wage Labour in the British Caribbean,” in Bridget Brereton and Kevin A. Yelvington (eds.), The Colonial Caribbean in Transition: Essays on Postemancipation Social and Cultural History (Gainesville, FL: University Press of Florida), pp. 77–107. Campbell, Gwyn and Edward A. Alpers (2004), “Introduction: Slavery, Forced Labour, and Resistance in Indian Ocean Africa and Asia,” Slavery and Abolition 25(2): ix–xxvii. Cartledge, Paul (1998), Democritus. London: Phoenix. Clendinnen, Inga (1991), Aztecs: An Interpretation. Cambridge: Cambridge University Press. Condorcet, Marquis de ([1781] 1999), “Reflections on Negro Slavery,” in David Williams (ed.), The Enlightenment (Cambridge: Cambridge University Press), pp. 307–316. Cooper, Frederick (1977), Plantation Slavery on the East Coast of Africa. New Haven, CT: Yale University Press. Du Bois, W. E. Burghardt (1909), The Negro American Family. Atlanta, GA: Atlanta University Publications Series. Engerman, Stanley L. (1977), “Black Fertility and Family Structure in the US, 1880–1940,” Journal of Family History 2(2): 117–138. Engerman, Stanley L. (1978), “Studying the Black Family.” Journal of Family History 3(1): 78–101. Engerman, Stanley L. (1983), “Contract Labor, Sugar, and Technology in the Nineteenth Century.” Journal of Economic History 43(3): 635–59.

SLAVERY, FREEDOM, AND SEN

105

Engerman, Stanley L. (1994), “Reflections on ‘The Standard of Living’ Debate: New Arguments and New Evidence,” in John A. James and Mark Thomas (eds.), Capitalism in Context: Essays on Economic Development and Cultural Change in Honor of R. M. Hartwell (Chicago: University of Chicago Press), pp. 50–79. Engerman, Stanley L. (1995), “Emancipations in Comparative Perspective: A Long and Wide View,” in Gert Oostindie (ed.), Fifty Years Later: Antislavery, Capitalism and Modernity in the Dutch Orbit (Leiden: KITLV Press), pp. 223–241. Engerman, Stanley L. (2000), “Comparative Approaches to the Ending of Slavery,” in Howard Temperly (ed.), After Slavery: Emancipation and its Discontents (London: Frank Cass), pp. 281–300. Fogel, Robert William (1989), Without Consent or Contract: The Rise and Fall of American Slavery. New York: W. W. Norton. Fogel, Robert William, and Stanley L. Engerman (1974), Time on the Cross: The Economics of American Negro Slavery. Boston, MA: Little, Brown. Franklin, John Hope, and Loren Schweninger (1999), Runaway Slaves: Rebels on the Plantation. New York: Oxford University Press. Genovese, Eugene D. (1974), Roll, Jordan, Roll: The World the Slaves Made. New York: Pantheon. Goldin, Claudia (1990), Understanding the Gender Gap: An Economic History of American Women. New York: Oxford University Press. Grotius, Hugo ([1646] 1925), The Law of War and Peace. Oxford: Clarendon Press. Gutman, Herbert G. (1976), The Black Family in Slavery and Freedom, 1750–1925. New York: Pantheon. Hale, Robert (1952), Freedom through Law: Public Control of Private Governing Power. New York: Columbia University Press. Hellie, Richard (1982), Slavery in Russia, 1450–1725. Chicago: University of Chicago Press. Heuman, Gad, ed. (1985), “Out of the House of Bondage: Runaways, Resistance and Marronage in Africa and the New World.” Slavery and Abolition 6(4). Higman, B. W. (1984), Slave Populations of the British Caribbean, 1807–1834. Baltimore, MD: Johns Hopkins University Press. Hobbes, Thomas ([1651] 1968), Leviathan. Baltimore: Penguin. Holt, Sharon Ann (2000), Making Freedom Pay: North Carolina Freedpeople Working for Themselves, 1865–1900. Athens: University of Georgia Press. Jones, Jacqueline (1985), Labor of Love, Labor of Sorrow: Black Women, Work, and the Family from Slavery to the Present. New York: Basic Books. Locke, John ([1690] 1963), Two Treatises of Government. New York: Mentor Books. Lovejoy, Paul E. (1983), Transformations in Slavery: A History of Slavery in Africa. Cambridge: Cambridge University Press. Maine, Henry Sumner (1885), Popular Government. London: John Murray. Mendelsohn, Isaac (1949), Slavery in the Ancient Near East. Oxford: Oxford University Press. Miers, Suzanne, and Igor Kopytoff, eds. (1977), Slavery in Africa: Historical and Anthropological Perspectives. Madison: University of Wisconsin Press.

106

STANLEY ENGERMAN

Montesquieu, Baron de ([1748] 1989), The Spirit of the Laws. Cambridge: Cambridge University Press. Moreno Fraginals, Manuel, Herbert S. Klein, and Stanley L. Engerman (1983), “The Level and Structure of Slave Prices on Cuban Plantations in the MidNineteenth Century: Some Comparative Perspectives,” American Historical Review 88(5): 1201–1218. Morris, Thomas D. (1996), Southern Slavery and the Law, 1619–1860. Chapel Hill: University of North Carolina Press. Nieboer, H. J. (1910), Slavery as an Industrial System: Ethnological Researches, 2nd rev. ed. The Hague: Nijhoff. Novak, William (1996), The People’s Welfare: Law and Regulation in Nineteenth-Century America. Chapel Hill: University of North Carolina Press. Nussbaum, Martha C. (1995), “Human Capabilities, Female Human Beings,” in Martha C. Nussbaum and Jonathan Glover (eds.), Women, Culture, and Development: A Study of Human Capabilities (Oxford: Clarendon Press), pp. 61–104. Patterson, Orlando (1982), Slavery and Social Death: A Comparative Study. Cambridge, MA: Harvard University Press. Phillips, Ulrich Bonnell (1909), Plantation and Frontier, Documents, 1640–1863. Cleveland, OH: A. H. Clark. Phillips, Ulrich Bonnell (1918), American Negro Slavery: A Survey of the Supply, Employment, and Control of Negro Labor as Determined by the Plantation Regime. New York: D. Appleton. Pufendorf, Samuel ([1688] 1934), On the Law of Nature and Nations. Oxford: Clarendon Press. Reid, A. (1983), “ ‘Closed’ and ‘Open’ Slave Systems in Pre-Colonial Southeast Asia,” in Anthony Reid (ed.), Slavery, Bondage, and Dependency in Southeast Asia (New York: St. Martin’s Press), pp. 156–181. Roberts, George W. (1957), The Population of Jamaica. Cambridge: Cambridge University Press. Robertson, Claire C., and Martin A. Klein, eds. (1983), Women and Slavery in Africa. Madison: University of Wisconsin Press. Rousseau, Jean-Jacques ([1762] 1947), The Social Contract and Discourses. New York: E. P. Dutton. Sen, Amartya (1980), “Equality of What?” in Sterling M. McMurrin (ed.), The Tanner Lectures on Human Values, Cambridge: Cambridge University Press, pp. 195–220. Sen, Amartya (1993), “Capability and Well-Being,” in Martha Nussbaum and Amartya Sen (eds.), The Quality of Life (Oxford: Clarendon Press), pp. 30–53. Sen, Amartya (1999), Development as Freedom. New York: Alfred A. Knopf. Silver, Morris (1995), Economic Structures of Antiquity. Westport, CT: Greenwood Press. Stampp, Kenneth M. (1956), The Peculiar Institution: Slavery in the Ante-Bellum South. New York: Knopf. Ste. Croix, G.E.M. de (1981), The Class Struggle in the Ancient Greek World. Ithaca, NY: Cornell University Press.

SLAVERY, FREEDOM, AND SEN

107

Steinfeld, Robert J. (2001), Coercion, Contract, and Free Labor in the Nineteenth Century. Cambridge: Cambridge University Press. Temperley, Howard (1972), British Antislavery, 1833–1870. London: Longman. Tolnay, Stewart E., and E. M. Beck (1995), A Festival of Violence: An Analysis of Southern Lynchings, 1882–1930. Urbana: University of Illinois Press. Voltaire ([1768] 1994), Political Writings. Cambridge: Cambridge University Press. Watson, James L. (1980), “Slavery as an Institution: Open and Closed Systems,” in James L. Watson (ed.), Asian and African Systems of Slavery (Oxford: Basil Blackwell), pp. 1–15. See also pp. 233–250. White, Deborah Gray (1985), Ar’n’t I a Woman? Female Slaves in the Plantation South. New York: W. W. Norton.

Chapter Five Freedom, Servitude, and Voluntary Contracts JONATHAN CONNING AND MICHAEL KEVANE

Introduction The literature on unfree labor has frequently found itself debating two purported paradoxes: a paradox of immiserizing emancipation and a paradox of immiserizing bans on the freedom to enter into “voluntary” forms of servitude.1 Together, of course, they form a double paradox: the first suggests that people may become materially worse off when freed to sell their own labor, the second says they are better off without restrictions on their freedom to sell their labor. Elements of both of these paradoxes are evoked in the preceding chapter of this book by Stanley Engerman in which he at first agrees with Amartya Sen that a ban on certain forms of bondage can at the same time expand both freedom and material welfare of laborers but later cautions that at other times, “freedom [has] meant a lowering of material well-being, health, and living conditions of ex-slaves,” and that “there are conditions, generally at low levels of income, where trade-offs between such freedoms might become necessary” (see his “Conclusion”). In this chapter we present an economic framework to revisit and reframe some important debates over the nature of free versus unfree labor and the economic consequences of emancipation. Our effort here is to suggest that such statements of paradoxes, while certainly valid at times, sometimes miss the larger point because standard partial-equilibrium economic analyses provide incomplete or misleading understandings of these situations. Consider the second purported paradox, often invoked by economists to point out the dangers of blanket bans or strict regulations 1 The observation that GDP per capita fell in post-emancipation societies including the U.S. South, Brazil, and sugar economies of the Caribbean is a widely noted and largely accepted empirical fact (Bush 2000; de Castro 2002) and one that was debated at length by abolitionists and political economists in the nineteenth century (Drescher 2002). Fogel and Engerman (1974) argued that the dietary and health well-being of slaves in the U.S. South may have been comparable or, along certain material dimensions, even superior to that of contemporary free laborers.

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

109

on certain types of servile but seemingly voluntary labor arrangements. In many post-abolition situations (that is, where coerced labor is forbidden) freed slaves or serfs and other types of poor laborers were often recruited to enter into new types of tied labor contracts. These included indenture and other forms of bonded labor as well as labor-service tenancy arrangements that tied peasant access to land to an obligation to supply labor service to a landlord. While each of these voluntary contracts has the potential to enhance efficiency and may be individually rational to accept as has so often been stressed in the literature on “interlinked contracts” (Braverman and Stiglitz 1982; Bardhan 1989), we argue that once all general equilibrium interactions are considered, these contracts may paradoxically reduce welfare for laborers as a class. That is because tied contracts can act as a barrier to competition, limiting peasants’ outside opportunities and therefore increasing the share of output that landlords can extract. In our model tied contracts or “servility” are “necessary” not because laborers are poor but rather as a strategy to help landlords sustain a collusive arrangement to pay workers wages below their marginal product and keep them poor. This can be seen by noting that an equally efficient (but less favorable to landlords) allocation of resources can be sustained via competitive factor markets without the need for any tying. We employ a simple general equilibrium framework in which labor can be either free or coerced and where land and labor will be exchanged on markets that can be competitive or manipulated, perhaps via nonmarket collusive arrangements.2 In addition to allowing us to analyze tied-labor contracts, the model will serve as a vehicle to formalize and reformulate some of the arguments and concerns that have animated the writings of Wakefield (1833), Marx (1887), Nieboer (1910), Domar (1970), and others regarding the relationships between conditions of labor scarcity and the “need” or demand for institutions of coerced labor. By working with variants of the same basic model under different assumptions about initial economy-wide factor endowments and asset ownership, we can compare equilibrium distributional outcomes under different institutional and contractual arrangements, including markets with free labor and free tenancy, slavery, and tenancy arrangements with labor-service obligations. This last type of arrangement has been ubiquitous throughout history (Morner 1970) and it was quite central to the organization of production under serfdom. Even today labor tying is seen in many regions of the developing world, and a careful analysis of this case offers important clues for understanding the origin and nature of other servile, yet seemingly voluntary, labor arrangements including several forms of bonded labor. 2

The presentation builds upon and extends Conning (2005).

110

CONNING AND KEVANE

The analysis yields insights that accord with common sense, but that are often overlooked or downplayed in academic debates, particularly amongst economists. Three insights are particularly important. First, landlords may at times forego an opportunity to enslave labor even where such coercion may be politically feasible to sustain. This will be more likely where laborers need incentives to apply certain talents and skills to productive tasks. Second, landlords often have the means and motive to collude to limit laborers’ outside opportunities even where there may be no explicit limits on labor mobility. We demonstrate how efforts to structure land and labor contracts as a bundle of tied or linked contracts may be an essential element of such a strategy. The frequent insistence in such contexts that labor be “servile” achieves similar ends. Third, state-led efforts to “emancipate” workers by banning certain types of voluntary or involuntary tied-labor contracts may raise laborers’ incomes but may also drive landlords to reach for other less effective but more distorting market manipulation mechanisms that can lower total output and continue to severely cap the growth of peasant incomes. Some of these insights contrast with conventional wisdom from both the modern and classical liberal traditions. Modern liberal thinkers hold that too much power in the hands of a few might lead inevitably to coercion, and forget that coercion has an opportunity cost. The powerful may sometimes be more interested in managing the choices of free laborers than compelling labor from a reluctant slave or serf. Likewise, ending coercive relationships may improve by little the welfare of those formerly coerced, if landowners have at their disposal alternate methods for extracting rents. Classical liberal thought holds that voluntary transactions must be welfare improving, and often downplays the importance of market manipulation and collusion in restricting the set of transactions available by assuming that collusive arrangements are fragile and will be undermined by competition. Yet banning the voluntary linked contract— forbidding a certain kind of freedom to contract—can raise laborers’ incomes by making collusive arrangements more difficult. Much like Adam Smith, then, the analysis presented here suggests that an embrace of freedom of contract should be tempered by an appreciation of the importance of guaranteeing self-ownership and insuring competition. In this regard our explanation differs from other recent attempts to explain why bans on child labor (Basu 1999; Baland and Robinson 2000) or voluntary forms of bonded labor (Genicot 2002) may be welfare improving. These explanations rely on multiple equilibria. For instance, it can be argued that low wages for adults may lead poor families to have few choices but to have family members enter into child labor or bonded labor, and that one of the reasons that adult wages remain so low is that there is such a large supply of child or bonded workers. Under certain

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

111

circumstances a ban on child or bonded labor may then lead the equilibrium adult wage to rise by enough that households no longer choose child or bonded labor. In these models the role of policy is to help the economy coordinate around a new Pareto-superior equilibrium. In our model the role of bans or regulations is to challenge the power of local elites by disrupting their strategies to restrict the choices of the poor and prevent peasant income from rising closer to its competitive level. Policy may improve the freedoms and the welfare of the poor, but there is less obvious scope for Pareto improvements compared to these other models, and output may well fall rather than rise. In short where these others are models of coordination failure, ours is fundamentally a model of conflict over property rights and freedoms. Before examining the details of the model framework, the next section places our analysis within a broader academic literature on unfree labor. This section reviews some of the arguments that have been made for why the material well-being of unfree laborers might fall following emancipation and why a ban on bonded labor and other forms of voluntary servitude could also do more harm than good. We turn then to the presentation of a simple general equilibrium model designed to pinpoint how different institutions of unfree labor might emerge as factor endowments, technology, and distribution of resources in an economy change. Debates over Unfree Labor Free and unfree persons are quite different. Even in very poor societies, free persons are typically able to choose their place of residence, seek work with more than one employer, accumulate property, and seek credit and insurance from the most favorable source. They can also make choices about whether to sleep late, chat the morning away, enjoy or refuse the company of others if they so desire, and consume unhealthy beverages during breaks from work. Unfree laborers rarely enjoy these freedoms except in stolen moments or at the whim and discretion of their masters or employers. Unfree persons are those persons constrained by forced labor arrangements such as slavery or serfdom, or by “voluntary” but servile labor relationships such as indentured or debt-bonded labor. Persons in these various categories have lost part of the everyday control over time and body that is so characteristic of the free person. Often with the explicit sanction of laws and societal norms, these individuals have found their opportunities for advancement sharply limited by the obligation to remain at the constant beck and call of their masters or employers. Even today, according to a 2005 ILO report, as many as 12.3 million people worldwide remain trapped in unfree labor relationships.

112

CONNING AND KEVANE

Unfreedom in the form of a forced labor arrangement is easy to define and to condemn, where constraints have resulted from the use of force or from obvious deceit or illegality. Since 1930 the ILO has defined the term “forced labor” by treaty as “work or service which is exacted from any person under the menace of any penalty and for which the said person has not offered himself voluntarily.” Few today would disagree that such practices ought to be banned and violations ought to be rigorously enforced. Yet many commentators argue that seemingly forced and servile forms of labor should not be so quickly and comprehensively condemned and banned. They fear that overzealous reformers might prevent poor workers from entering into long-term arrangements that improve their welfare. The argument for caution has two components. The first is that banning a voluntary and hence mutually agreeable arrangement may reduce well-being. Arrangements entered into voluntarily, by revealed preference, mean that the individual who chooses the relationship must have been choosing something that improved his or her situation. Circumstances make the person poorly off; the arrangement makes him or her better off. This argument is used frequently in discussions of child labor, where the benevolent parent chooses between two bad options: commit the child to a servile labor relationship or suffer the possible long-term harm of not being able to furnish the child with adequate nutrition (Basu 1999). The policy implications of this analysis are few, for, as a very recent economics survey of the topic by Edmonds and Pavcnik (2005) concluded, “Whether anything other than economic development is an effective, long-term solution to the widespread incidence of child labor is an open question” (p. 219). This first component is taken to be self-evident by many economists, who often dismiss the argument that persons have to be “forced to be free” by a benevolent and wise polity. Negligible is the possibility that people make systematic mistakes, or might act against their long-term or objective interests as one of their many irrational selves outmaneuvers the more rational selves. A voluntary agreement is always a meeting of reasonable and rational minds, in this view, and arguments to the contrary have a whiff of “moral fastidiousness” about them (Krugman 1997). If we respect the dignity of each human, should we not also respect his or her choices? Yet, a new literature on the “paradox of bans” suggests that contracts that appear to be Pareto improvements when analyzed in isolation, may more properly be seen as lowering the welfare of certain groups, when analyzed at the general-equilibrium level. This literature is a new and interesting application of the theory of the second best. The initial absence of complete markets implies that limits on existing markets and trades can, in some circumstances, improve well-being.

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

113

The second component of the argument for caution examines present and past economies, suggesting they may not involve coercion. In this view historical and present-day arrangements such as serfdom, sharecropping, labor-service tenancy, and other “tied” labor arrangements are not self-evidently the result of the application of force by the employer, as many have claimed. Rather, these may perhaps be contractual adaptations that help individuals sustain commitments in a milieu where asymmetric information, costly enforcement, and the absence of effective and impartial courts make commitments difficult or costly to sustain. Hundreds of thousands of poor European migrants, for example, used indentured servitude contacts to finance their passage to the Americas in the seventeenth and eighteenth centuries (Bush 2000; Galenson 1984). Often after five years of service the indentured servant earned his “freedom dues” and became a settler and possible property owner. These poor Europeans did not possess sufficient resources to finance the voyage and purchase land and did not have assets to pledge to assure a lender of compensation in the event of default. Given the distances and arduous passage, and the ease of movement to an ever-expanding frontier, lenders had reason to fear that loans would not be repaid. In this environment, voluntary relinquishment of rights for a temporary period of time seemed quite reasonable; a lender could sell the loan to an employer in the New World who might then collect repayment out of wages. The penal code helped enforce these contracts. A ban on indentured servitude would have made worse off the poor European workers who could not afford the voyage. By similar argument, a harsh penal code could be defended in the name of helping the poor to hold to commitments that would help improve their lot. Studies by Bauer (1979) and Knight (1986) of debt bonding in Latin American agriculture and larger comparative surveys by Northrup (1995) and Bush (2000) each emphasize the ways in which these contracts often responded to migrants’ and peasant farmers’ economic demands and often led to their advancement, even if they do agree that abuses did often occur. In a similar vein, North and Thomas (1973) saw serfdom emerging as a contractual arrangement that exchanged “labor services in return for the lord’s protection” and other public goods in a dangerous world (p. 20). They rationalized the labor-service component of these contracts as a substitution for money rents in an incompletely monetized economy. While many discussions on indentured servitude and serfdom have been broadly empirical in approach, a related debate on labor tenancy has been largely theoretical. The issue has been to show why tenants had to deliver labor services as payment for a plot of land in replacement for, or in addition to, monetary rent. The basic argument is that, in a world of

114

CONNING AND KEVANE

asymmetric information (or “transactions costs” in earlier parlance), bonding and tying may emerge as devices to help contracting parties to enforce commitments to expand trade, and therefore to expand the size of the economic pie (Sadoulet 1992). Each part of the argument for caution in public action with regards to servile labor arrangements—that voluntary arrangements are welfare improving and that many labor arrangements are more likely to be voluntary than previously thought—resonates with an extensive corpus of writing on the subject. But the argument for caution also has its zones of contention (Hart 1986). The theoretical argumentation relies on particular assumptions, and the major purpose of this paper is to show that under other reasonable assumptions some of the intuitions of the argument for caution in public action may be unwarranted. Likewise, empirical approaches rely on selective presentation of facts and statistics. Finding exceptional cases of voluntary servitude might obscure the more important fact of involuntary servitude. There is no shortage of reliable accounts of how persons have been threatened with bodily harm were they not to “consent” to an involuntary labor bond (Bales 2000). The threats sometimes emerge in the course of deliberate raids to capture labor, and sometimes as a judicially sanctioned punishment for failure to repay as debt. Sometimes children are rendered into the status of unfree labor as compensation or punishment for unpaid debts of the parents. In the same vein, in commenting on North and Thomas’s interpretation of serfdom as an efficiency-enhancing contractual exchange of tribute for protection and other local public goods, Robert Brenner (1996) compares the services rendered by lords to those of a protection racket. In effect, serfs were offered a contract they could not refuse.

A Model of Labor and Tenancy The following pages analyze variants of a simple general equilibrium model of an agrarian economy, modified and extended to focus attention on the interacting roles of factor endowments, technology, and the initial distribution of assets. These shape opportunities and incentives that elites have to use extra-economic coercion and market power to generate patterns of agrarian organization and distribution to their advantage. The framework is meant to be general enough to nest and allow for comparisons across several alternative institutional arrangements, including the standard textbook model of competitive and efficient land and labor markets, an economy with slavery, economies with both voluntary and involuntary labor-service tenancy contracts, and an economy where large

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

115

landowners withhold land from the lease market, giving rise to a “latifundia-minifundia” agrarian structure. This framework will permit a quick tour of several hypotheses and debates that have engaged political economists, economists, and historians for a long time. For example, we will state and analyze the NieboerDomar hypothesis on the causes of slavery or serfdom and mention some elements of its critiques. The model also provides a framework for understanding debates concerning the economic and political consequences of emancipation and the nature of the transitions or “paths” toward more free and competitive land and labor contracts (Byres 1996; de Janvry 1981). − We begin with an economy that has T = 100 units of land. There are M
2 identical landlord households that together own 80 percent (θ
0.8) of the total land area, or 40 units of land each. There are also N
80 peasant households that together own (or have customary property rights to) the remaining 20 percent of the land mass, or 0.25 units per household. Each peasant household owns H
1.25 units of household labor, and for simplicity we assume landlords do not supply manual − labor. The total labor force consists therefore of L = N . H = 100 units of − −− labor, and the over all land-to-labor ratio is t = T/L = 1 unit of land per laborer. Landlord and peasant households have knowledge of the same crop production technology that would allow them to produce a staple agricultural product sold at unit price set by world markets. In our first scenario, we assume that landlords begin with a higher initial endowment of nontraded farm management skills. Farm production technology is summarized by a standard constant-returns-to-scale (CRS) Cobb-Douglas production function q
T
LS1
, where T and L are land and labor farm inputs and S is the household’s holding of a nontraded farming skill or labor supervision ability. By normalizing the peasant’s endowment of factor S to Sp
1, we can represent the peasant’s technology more compactly by the restricted production function F(T,L)
T
L, which, as long as S plays any significant role in production (i.e., as long as
   0), will display decreasing returns to scale in land and labor input. The importance of nontraded skills in agricultural production is empirically well grounded and is widely used to explain the lasting prevalence of family farming and tenancy in many contexts (Hayami and Otsuka 1993). Landlords employ the same production technology but have a larger endowment of the nontraded input, Sr  1. The landlord’s technology can then be represented as G(T,L)
AF(T,L) where A = Sr(1−−). Intuitively, the landlord’s assumed higher level of managerial skill can be thought of as raising the productivity of the land and labor inputs in the

116

CONNING AND KEVANE

restricted farm production function. For the simulations below, we shall assume 
0.45, 
0.45, and that landlords start with Sr
10, or ten times more of the nontraded factor compared to a peasant household.3 If land and labor markets are allowed to be fully competitive, then households will allocate their own factor inputs and hire in or hire out land and labor until the marginal value product of each factor is equalized to a common market land rental or labor wage. Suppose the efficient equilibrium level of peasant land and labor use is denoted as (Tpe, Lpe). The equilibrium level of input use can be easily deduced by noting that the equalization of marginal value products across farms implies that farms will operate using the same factor proportions (i.e., same land-tolabor, skill-to-land, and skill-to-labor ratios). Since landlords have Sr
10 times more management skill in an efficient equilibrium, they must also operate using ten times more land and labor compared to peasant farms. In equilibrium, labor demand from the N peasant farms (N . Lpe
80Lpe) plus the labor demand from the M landlord farms (M . Sr . Lpe
− 20Lpe) must add up to the economy’s total labor supply (L
100). From this (and applying similar reasoning to the land market) it is easily calculated that Lep =

L =1 M ⋅ Sr + N

Tpe =

T =1 M ⋅ Sr + N

Hence in the efficient competitive economy each of the 80 peasant households would operate a farm with one unit of land and one unit of labor, while each of the two landlord households would run a larger farm with 10 units of land and labor. Given the assumed distribution of ownership over land and labor, we could view this equilibrium as consisting of each landlord leasing out 30 units of land (0.75 units of land to each of 40 subtenants) while operating his own demesne or plantation on the remaining 10 units of land by hiring 0.25 units of labor from each of 40 households, for a total of 10 units of labor hired. Note that the competitive equilibrium distribution of operational farm sizes is unique and is determined independent of the initial distribution of land or labor ownership. Efficient operational farm sizes are affected only − We’ve chosen Sr such that MSrN
L. This is without loss of generality, and mainly a convenient normalization to help simplify later expressions. 3

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

117

by the overall land-to-labor ratio and the distribution of nontraded factors. Factor prices, equal to marginal products, are easily calculated to be w = FL (t , 1) =   t 



r = FT (t , 1) =   t 

= (1 −  − )  t 

 −1



where w, r, and  are, respectively the market wage and rental rates and the shadow price of a unit of nontraded S. This last quantity is also the value of peasant farm profits. The wage rate w increases monotonically − with t as labor becomes scarcer (or, what is the same thing, as land becomes more abundant) while the competitive land market rental rate falls. Total income to household g is given by the sum of farm profits plus factor sales:  g (t , 1) = Sg  F(t , 1) − rt − w1 + wL g + rT g − At an initial land-to-labor ratio of t
1, landlords in our simulated economy will be essentially rentiers, deriving 95 percent of their income, or 19 units, from their 40 units of land valued at the market rental rate r
0.45, and only 5 percent, or one unit of income, from farm profits. Since all farms face a common market wage, a peasant laborer should be, on the margin, indifferent between working another hour for a landlord or on her own farm or tenancy. This implies that there should be no compelling reason for any party to tie labor and tenancy contracts. Some peasants might of course rent from and work for the same landlord, but this would emerge by chance. Landlords might want to clock their laborers’ hours on the job (one could perhaps interpret S as a required input into this monitoring process), but a landlord should not care at all about what any given laborer does on her own family farm or tenancy on her own time. This is because each hour of work that that laborer withdraws from the landlord can be easily and immediately replaced by hiring another laborer at the same spot market wage.

Labor Scarcity and Coerced Labor: The Nieboer-Domar Hypothesis Suppose that, due to discovery or conquest, the economy’s land endow− ment is now expanded from 100 to t > 100 units, while the labor force − remains constant at (L
100). Assume furthermore that it is either the

118

CONNING AND KEVANE

peasant households, or some third party such as the government, or a previously non-landowning military service class, that establishes property rights on the newly expanded land frontier. We assume that the members of this new group do not have the required skills to operate farms themselves so they will simply turn around and rent out (or sell) any newly acquired lands to existing landlord and peasant farmers at the new competitive market rental rate. A new efficient equilibrium should emerge where every existing household production unit with farming ability S expands its use of land proportionately to absorb the economy’s increased acreage. Since the labor supply and distribution of nontraded farming skills has not changed, each farm employs as many labor units as before (one unit per operational farm unit) but each unit of labor is now more productive. From the equations above and intuition, it should be clear that the higher land-to− −− labor ratio t
T/L leads to higher equilibrium labor market wages and farm profits and lower land rental rates. Unless the original M landlords are able to appropriate significant amount of this frontier land for themselves at below market prices, their total incomes will surely decline. The solid line in figure 5.1 indicates how the income to a landlord who owns 40 units of land falls as a function of the economy’s rising land-to− labor ratio t as we add land to this parameterized economy. Faced with the prospect of declining rents and rising wages as population density declines, landlords may try to expropriate the returns of peasant labor by acting to severely limit peasant labor mobility or through direct coercion. This is, in essence, the argument made in Evsey Domar’s famous 1970 essay, “The Causes of Slavery or Serfdom: A Hypothesis.” Building on earlier ideas by E. G. Wakefield (1833) and H. J. Nieboer (1910) and motivated by the descriptions of the imposition of Russian serfdom in the sixteenth century after the great expansion into the steppes by historian V. O. Kliuchevsky (1968), Domar conjectured that coerced labor arrangements were most likely to emerge in economies where labor had become scarce relative to land. Hence, Domar wrote, as “the central areas of the [Russian] state became depopulated because of peasant migration into the newly conquered areas in the east and southeast,” serfdom emerged, “under the pressure of the serving [landlord] class . . . [as] the government gradually restricted the freedom of peasants . . . to move” (pp. 18–19). Domar’s hypothesis has been subjected to some valid criticisms, for example by Patterson (1977) and Engerman (1973). Informed commentators such as Kolchin (1987) argue nonetheless that while the hypothesis certainly cannot explain all cases, the concern over the scarcity of labor probably “was the most basic element behind the rise of slavery and serfdom in several important cases including Russia and the slave labor

119

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

Landlord/slaveowner total income

20

18

Slave

16

14

Fre e

12

10

1

1.5

2

2.5

3

Economy Land-to-Labor Ratio Figure 5.1 Landlord sector income as a function of the land-labor ratio: free markets vs. slavery.

plantations in the Americas and the US South” (p. 19). We do agree with the criticism that Domar, much like Wakefield (1833) and Nieboer (1910) before him, appears at times to suggest that landlords’ incentive to clamp down on laborer’s mobility is largely due to the existence of an open frontier to which peasants might otherwise be drawn to set up independent farms. In fact, an open frontier is not an essential element of the argument. As suggested by the model above, what will matter more generally is the distribution of property rights over land as well as anything else that might affect the value of laborers’ outside opportunities.4 Domar was not very clear about the nature of the difference between slavery and serfdom or about how technology or the initial distribution of property rights over land might shape these institutional choices and transformations. To clarify these issues, consider the following simple extension which models the imposition of slavery in a very crude way. Suppose that landlords can, as a result of a campaign of violence and intimidation, kidnap half the existing peasant population and compel them to supply their labor at a subsistence wage w per unit. We assume 4 The argument that an open frontier is not essential was also made by Engerman (1973) and was an essential element of Marx’s elaborate critique of Wakefield in the last chapter of Capital, volume I.

120

CONNING AND KEVANE

that the new slaveowner can compel a slave to provide manual labor but cannot compel the slave to supply nontraded farming skills. One way to interpret this assumption is that it captures elements of an idea, traceable to Adam Smith, that slaves cannot be motivated to supply certain types of labor effort as efficiently as free laborers, because slaves are, by definition, limited in what property they can own and accumulate which might serve as incentives. Although this economy clearly violates slaves’ rights, we shall assume that property rights over slaves are secured and tradable and that both land and labor are transacted on competitive markets. The competitive renting or selling of slave labor was commonplace in many slave economies such as that of the southern United States. To simplify, slave labor and free labor are assumed to be perfect substitutes in production so that an hour of slave labor will trade on the market at the same market wage w. Given these assumptions, the principal (but not the only) difference between this slave economy and the earlier described competitive equilibrium with free land and labor contracting lies in the distribution of property rights over labor. Total landlord income is, as before, given by the sum of farm profits plus factor market sales: − − [G(T rs, Lrs) r .T rs w .Lrs]rT r(w w)Lr The latter includes the “exploitation rents” the slaveowner receives from expropriating w worth of labor services from each unit of slave labor but paying each slave only w for those services. In our numerical example each landlord now owns 10 slave households, or 12.5 (
10  1.25) units of slave labor. Landlords face a trade-off in choosing between a slave economy and a free labor and tenancy economy. Domar said relatively little about the political arrangements required to sustain coercion, or its costs to the elite, and instead simply pointed out that a landowning aristocracy could not easily survive in a land-abundant environment due to the high cost of free labor. Domar and Nieboer did not, however, appear to have considered another trade-off, that our model does capture. Even if slaves have been imported from outside of the economy, landowners must ask whether more can be earned by employing these workers as slaves, or whether they could earn more by transforming these slaves into tenants so as to try to extract via land rentals at least part of the return the economy would enjoy from bringing forth into production the nontraded farming skills S that these new peasant-tenants otherwise might not supply. The nature of this last trade-off will be determined in large part by the importance of these nontraded skills in the economy’s crop technologies.

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

121

Figure 5.1 compares landlord income under an economy with free tenancy and free labor market arrangements as earlier analyzed (solid line) to that under a slave economy where slave laborers can be compelled to work but do not use nontraded skills S (dashed line) for different land-tolabor ratios.5 As computed with the parameters given above, landlords earn higher incomes in the free economy when the land-to-labor ratio is somewhere approximately below 1.25, but prefer the slave economy in more labor-scarce environments. This comparison does not net out the costs that slaveowners might have to pay to coerce a slave labor force, nor does it consider the possibility that slave laborers could be either more or less productive than free laborers.6 As drawn, landlord incomes at first fall but later rise as the land-labor ratio increases. This reflects the tug of war between falling income from land rents and rising income from owning slaves. The dispossession and enslavement of half of the peasant population means that land and labor that these households had employed in a free economy are now put on the market for the remaining production units to absorb. In the new “efficient” equilibrium, the size of landlord farms and the remaining free farms must expand proportionately to absorb these resources, both wages and land rental rates must fall (factor productivity falls as less S is being used), and farm profits must rise (as S is now more scarce). Total output would fall compared to an efficient free and competitive equilibrium.7 Independent farm producers who are not enslaved may either gain or lose from this change in factor prices, depending on the importance of wages and farm profits in total household income. The political implications of this analysis are rather interesting. It suggests for example a reason why white small-farm producers in the 5 The parameters are as described so far in the text and w
0.45, the free labor market − wage when t
1. Raising w simply raises the costs of owning each slave, lowering landlord − income in the slave economy and therefore raising the cutoff level of t above which slavery becomes financially more lucrative. 6 A large and important literature has debated this latter question. Adam Smith and many abolitionists often argued that slave labor was by its very nature less efficient than free labor because a slave who could not own and accumulate property would have less incentive to supply quality labor. Several nineteenth-century political economists challenged this view (Drescher 2002). Fitting into this debate, Fogel and Engerman (1974) have argued that slave labor in the antebellum South produced more output per worker than free labor because of the extra labor and “economies of scale” that they argue could be achieved under a coercive gang labor system. The dominant view amongst economic historians seems to be that American slavery was on the whole a profitable institution to slaveowners at the time of its abolition (Kolchin 2003). Our model can be easily adapted to either assumption. 7 If slaveowners could extract sufficiently more output per worker from a slave than from a free person through coercion, then total measured output could fall following the abolition of slavery, the explanation suggested by Fogel and Engerman (1974). Without denying that this effect could have been important, we shall suggest a complementary explanation.

122

CONNING AND KEVANE

U.S. South would have opposed abolition even though they did not own slaves, and why they would later support establishing impediments to independent farming by freed slaves. Allowing freed slaves to put their own farm management skills to work would have raised market wages and land rents in ways that would have lowered profits for these existing farm operators. The Roads to and from Serfdom: Tied Labor Contracts The Russian landlords and the tsars did not establish slavery. What instead occurred was the slow but steady growth of legislation over the course of the sixteenth and seventeenth centuries that made it increasingly more difficult for laborers to move from one landlord to another. This transformed existing modes of “voluntary” tied labor contracting including varied forms of “voluntary” bondage, “self-pledging,” and labor-service tenancy into the involuntary tied arrangements that came to constitute Russia’s new serfdom (Kliuchevsky 1968; Kolchin 1987). Conversely, when centuries later the Russian tsar abolished serfdom by decree in 1861, or when slavery was abolished in the sugar economies of the Americas and the U.S. South, these economies did not snap into new competitive tenancy and labor markets of textbook analysis. Instead, in almost all cases these societies transitioned to economies where indenture, debt bonding, and varied forms of labor-service tenancy emerged. Our model provides a simple explanation for the widespread use of voluntary, tied-labor contracting. Tying labor can serve as a device to help landlords collude and improve their ability to extract land monopoly and labor monopsony rents. Collusion is difficult to sustain in practice because laborers will naturally seek higher returns by contracting outside of the collusive arrangement, or by more intensively utilizing whatever land of their own or common property resources they have available. Anticipating this, elites will seek to preempt these outside options and try to either limit labor mobility outright (in which case the system becomes involuntary serfdom) or strictly regulate peasant access to other resources and trading opportunities in the “free” economy. Hence, almost universally, we find elites in post-abolition societies rushing to sponsor anti-vagrancy laws, and in general discouraging peasant production through the imposition of taxes, limits, or prohibitions on the ability of freedmen to own land, to obtain credit, to move to or settle in certain “white” areas or frontiers (Ransom and Sutch 2001; Binswanger, Deininger, and Feder 1995; Lundahl 1992). The question arises as to what determines whether an economy that bans coercion and abrogates restrictions on mobility will transition to the efficient competitive equilibrium described above, or to some other type

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

123

of equilibrium. Suppose slavery or serfdom were suddenly abolished. To add a touch of realism, suppose also that the transition were somewhat disruptive so that for a brief period households retreated to a semiautarkic equilibrium. Recall that the “slave economy” of the last section was privately profitable to the slaveowners but it was also inefficient in so far as slaves’ nontraded skills S were not being utilized. Whether this efficiency cost is large or small depends on the importance of S in production and the extent to which S is held by slaves. Rather than allow the emergence of a competitive market where landlords have to compete for labor by offering the highest wage or for tenants by offering the lowest land rent, they would much prefer a more collusive arrangement. Consider the following collusive contracting arrangement: landlords agree to divide up the peasant population equally amongst themselves and to not contract with each other’s peasants. Each landlord then offers each peasant household a take-it-or-leave-it contract with the following clauses: − The landlord will lease Tpe
t units of land (i.e., the efficient level described in the competitive equilibrium above) to a peasant in exchange for a lump-sum rent or tribute payment R (value set as described below). The lease is provided on condition that the tenant additionally agrees to supply 0.25 units of labor service to the landlord. The tenant shall not sublease or allow any other workers to work the land without the landlord’s explicit consent. To insure that the contract remains an entirely voluntary transaction for the peasant, the landlord cannot set the tribute payment R above a level that would push the tenant-laborer’s welfare below their reservation utility, or what they could earn elsewhere if they refused the contract. The peasant’s reservation utility depends on the availability of alternate contracting opportunities and his ownership of land (defined broadly to include customary land rights or access to common property areas). If all other peasants in the economy are accepting similar contracts, then the only alternative to contracting with the landlord becomes to remain or retreat into autarkic (Chayanovian) production. In either case, the most a peasant household can earn is what it could get from using its available land and labor − endowments. In our numeric example this would mean devoting Tp
0.125 − − 8 and Lp
1.25, to home production to yield income F(Tp, Lp). To assure voluntary participation the tribute rent R must be chosen to make sure that the peasant earns slightly more from accepting than from rejecting the contract: F(t , 1) − R ≥ F(T p , L p )

124

CONNING AND KEVANE

The colluding landlord therefore can set the rent/tribute payment as high as just below R = [ F(t , 1) − F(T p , L p )]

or



 t  − 0.434 = 0.566 per tenant when t = 1. As long as landlords can sustain a collusive equilibrium, this set of voluntary tenancy contracts with labor-service obligations reproduces the efficient competitive production allocation exactly, except that here landlords take a far larger share of total output. Since each landlord contracts with 40 peasants, he earns a total income of G(Sr t , Sr ) + 40R − or 32.65 in our numerical example when t
1, which is well above the 19 units that he would earn under a competitive equilibrium in the same economy. No doubt some economic historian many decades later looking back at the records left of these voluntary transactions would be led to proclaim that these contracts had all been “Pareto enhancing” and therefore to the advantage of the peasants involved. As evidence they would point to the certain fact that compared to what a subsistence peasant earned those “lucky” to obtain these contracts were made better off. While the logic of the arguments seems impeccable, the economic historian’s conclusions are in fact quite misleading in our example economy. The historian’s problem here is not a failure of logic but one of imagination—the failure to imagine the counterfactual of what the distribution of payoffs would look like in a truly competitive economy if landlords had to compete and every factor were paid its marginal product. When this is the counterfactual, labor-service contracts can be instead interpreted as a strategy by landlords to limit competition and sustain collusion, and therefore something that immiserizes rather than improves peasant lives. It should also be obvious that, starting from the proposed collusive equilibrium, a peasant household with a labor-service tenancy contract would have incentives to try to breach the contract and divert labor from the landlord’s demesne to his rental plot. That is because the peasant’s, − marginal return from extra labor on his own plot is FT(t , 1), compared to zero on the landlord’s farm where labor is obligated. If the application 8 Transacting with other peasants who have not accepted landlord contracts offers no better alternative over this reservation payoff, since we have assumed these other peasants have the exact same endowment. Throughout this discussion we continue to assume that − − T
L
100, θ
0.8, and all other parameters are as described above.

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

125

of labor time or effort could not be perfectly observed on the landlord’s farm, the situation would essentially become a classic principal-agent or moral hazard problem. The landlord would complain that his peasants lacked a work ethic and were diverting resources to their individual plots rather than fulfilling their contractually agreed-upon legal obligations to provide labor service. Landlords as employers would also argue, again on seemingly firm liberal grounds, about the need to maintain a system of criminal sanctions against contract breach (Steinfeld 2001). How, after all, could free markets be made to function properly unless workers could be kept to their promises, and how could this be for anything but for the workers own good, when they entered into these contracts of their own free will? Without legal enforcement, the argument would go, landlords might have little choice but to respond to this situation of moral hazard by underinvesting in agriculture and rationing peasants to smaller-thanefficient tenancies, with a consequent fall in total output. This moral hazard problem is a very familiar one that has been at the center of important explanations of the choice of contract forms. The crucial point to note here however is that the “problem” emerges primarily as a consequence of the fact that landlords are trying to get away with paying labor a wage far below its marginal product. Since the wage for each extra hour of work on the landlords’ demesne is essentially zero, peasants are strongly tempted to cut back on an hour of labor service and divert that − labor to their own tenancy, where on the margin they earn FT(t , 1). A peasant therefore has strong incentives to feign sickness or find other ways to cut back on labor-service hours.9 There is no similarly strong incentive to avoid wage labor in a competitive economy, because the marginal gain peasants earn by using that labor on their own farm would − then be exactly offset by the wage w
FT(t , 1) they lose by working an hour less for the landlord. Can this collusive equilibrium be sustained? The terms of the contract act as a barrier to competition for several reasons. Clearly stipulated and strictly enforced labor-service obligations aim at stopping labor from being diverted to peasant plots as well as to other landlords. In actual practice a way to control labor’s outside opportunities was to mark class differences and insist that the servant adopt a servile and deferential attitude and to remain attentive to the master’s every “beck and call.” Although historians have sometimes sought to root such behaviors in cultural and military traditions, these behaviors obviously also served the 9 The landlord cannot ameliorate the problem by offering a larger tenancy (and hence residual claimancy) to the laborer and then extract rents through a higher lump-sum tribute payment, because this would fail to fully utilize the landlord’s own farming skill S.

126

CONNING AND KEVANE

useful economic purpose of limiting a peasant’s chances of discovering or utilizing outside “moonlighting” opportunities. Similarly, the stipulation that the peasant not hire outsiders without the landlord’s explicit consent is aimed at making sure that a competing market for wage labor does not develop within the peasant sector. Without this clause a peasant who refused to accept a labor-service tenancy on the landlord’s terms might be able to find work as a laborer on another tenant’s plot (where the marginal product of his labor exceeds what landlords are offering as wages). The important point of this discussion has been that the “moral hazard problem,” and the attendant conflict that the landlord must work so hard to control, is deeply rooted in the landlords’ attempts to sustain a collusive arrangement that pays labor-service a wage well below its marginal product (essentially zero in our simple example) and charges tenants tribute rent well above the shadow factor price of land. If the collusion were broken, and landlords were forced to compete to hire in labor and lease out land at competitively determined factor prices, then the “need” to tie or to insist upon a servile labor force would disappear or be greatly reduced.10 Even if the terms of labor-service contracts could be perfectly enforced by “the law,” the collusive equilibrium could be undermined by competition between landlords. Since each contract is lucrative, landlords might be tempted to poach each other’s peasants by offering labor-service contracts with lower tribute R, or by simply hiring away workers at a wage. Serfdom solved this problem by imposing lordly jurisdiction that strictly limited mobility and controlled and regulated peasants’ freedom to engage in activities that might divert labor or effort away from tasks that could be subjected to lordly taxation (Brenner 1996). Where laborers’ movement could not be perfectly limited through coercion, landlords often found it convenient to employ the device of a credit advance to bring tenants under the purview of the law and penal codes to enforce the terms of labor service. The voluntary contract above could be modified to provide the peasant with an up-front credit advance, ostensibly to compensate the landlord for the cost of the peasant’s dwelling or perhaps the “cost of passage” that the landlord has provided. For example, prior to the imposition of involuntary servitude in Russia, peasants often “self-pledged” by accepting a loan backed by a promissory 10 We are well aware, of course, of a large literature that makes clear that ex ante commitments to limit an agent’s outside opportunities may be Pareto-improving to both landlord and peasant in many moral hazard contexts (Holmstrom and Milgrom 1990; Braverman and Stiglitz 1982). Virtually all of this literature starts however from the partial-equilibrium, bilateral contracting approach that takes agents’ reservation payoffs as given, and therefore ignores the primary focus of our analysis here: how contract terms may affect those reservation payoffs in general equilibrium.

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

127

note in which they agreed to pay off their obligation by serving the landlord as a servant or a labor-service tenant (Kliuchevsky 1968). It was not unusual for a contract to have a worker “serve for usury,” or in other words, work merely to pay off the accumulation of interest, leaving the level of the principal or the bond unchanged. Any pledger who broke service or was declared insolvent by a court could be turned into a permanent bondsman who could then only be freed at his master’s will, a condition of service that became hereditary. These arrangements still remained “voluntary” and contractual so long as the law generally allowed the peasant to ransom his or her freedom and, although only during officially sanctioned days of the year, to change employers. This system of voluntary bonding evolved into bondage in perpetuity or involuntary serfdom over the course of the late sixteenth and seventeenth centuries as new laws were introduced that made it increasingly difficult for peasants to change employers or to ransom their freedom. V. O. Kliuchevsky saw these restrictions as a landlord reaction to the conditions of “acute labor shortage” that developed as “masses of the peasantry” started to flee the central Russian provinces in search of land and opportunities in the newly expanded frontier” (Kliuchevsky 1968, 182).11 As this discussion suggests, the dividing line between voluntary and involuntary tied labor contracts has oftentimes been blurred. Economists often assume that a system of voluntary contracts must, almost by definition, lead to improvements to the laborers who accept them, yet as we have argued here this need not be the case once the possibility of landlord collusion and general equilibrium interactions are considered. History is replete with other examples of laws and state action aimed at facilitating collusion amongst landlords to keep “voluntary” labor arrangements from becoming too competitive (Binswanger, Deininger, and Feder 1995). For example, after the Thirteenth Amendment to the Constitution was adopted in 1865 banning involuntary servitude in the United States, southern legislatures rushed to immediately pass “Black Codes” to limit freed slaves’ outside opportunities by making it more difficult for them to switch employers, own land, access forests and other common property resources, or even to remain idle. This legislation was often quite naked in its obvious attempts to sustain collusive arrangements. 11 Kliuchevsky underscores how landlords used the legal device of debt bonding to tie down their workers, for instance by converting practically all the remaining servants who had previously worked for wages and without promissory notes into bondsmen. The terms of bondsmen’s contracts also became more and more onerous as time passed, adding clauses in which bondsmen had to for example agree to “live as a peasant under my master for the rest of my life and not run away anywhere,” and which obliged the peasant to now pay damages on top of his debt for leaving (Kliuchevsky 1968, 184).

128

CONNING AND KEVANE

Consider the following excerpt of a law passed by the Louisiana legislature in December 1865, cited by Ransom and Sutch (2001) in their classic study of the postbellum southern economy. The legislation limited the ability to switch employers by requiring “all laborers to make contracts for the following year by January tenth” (p. 67) and by stipulating “that laborers leaving employers would forfeit all wages due them and could be imprisoned as a lien against any alleged losses.” Just as importantly, the law also threatened that anybody who should “entice away, feed, harbor, or secrete any person who leaves his or her employer . . . shall be liable for damages to the employer, and . . . shall be subject to pay a fine of not . . . less than ten dollars, or imprisonment in the parish jail for not . . . less than ten days, or both” (p. 67). Even if such types of legislation have by now been taken off the books in many nations, servile labor and other forms of voluntary bondage still persists where local laws, customs, and norms of behavior help sustain collusive arrangements that severely limit workers’ outside opportunities, often on the basis of race, caste, nationality, or ethnic or gender differences (ILO 2005).12

Property Rights, Asset Inequality, and the Paradoxes of Bans Efficiency can be achieved, in the model economy, either through competitive factor markets or via a collusive but “efficient” system of voluntary or involuntary tied labor-service contracts that extracts surplus from peasant households via lump-sum tribute and labor-service obligations. If overall factor endowments consist of 100 units of land and labor, the efficient allocation of resources under our assumed technologies should yield a total of 100 units of output for the economy, regardless of the initial distribution of land. But efficiency need not be an outcome of the economy. Figure 5.2 is useful for summarizing and explaining how income and distribution change under different institutional arrangements and initial distributions of property over land. The efficient potential level of income, as discussed above, is indicated by the horizontal “Total—efficient” line passing through point E in the figure. The peasant household sector’s share of total output is obviously very different under these two alternative scenarios. In a competitive market, total peasant income— 12 Modern economic theories such as the theory of indefinitely repeated games with reputational equilibria (Kandori 1992), models of asymmetric information and stigma (Akerlof 1976), or evolutionary game theory and agent-based analysis of social dynamics (Durlauf and Young 2004) have provided solid microfoundations to explain how collusive arrangements and discriminatory norms might arise and be sustained over long periods of time.

129

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

Total - efficient

100

E Total - distorted

D 80

Income

C

Peasant - competitive

60 V A

Peasant - ban, distorted

40

Peasant - tied

20 0.65

0.7

0.75

0.8

0.85

0.9

0.95

Land Figure 5.2 Total income and peasant sector income under different regimes.

given by the sum of farm profits plus the market value of factors owned—is as indicated by the solid line labeled “Peasant—competitive” passing through point C. This has a slope of minus the competitive rental rate on land r because the value of land owned by peasants declines as the fraction of total land owned by landlords θ increases. Point C on the diagram indicates that when landlords own 80 percent of the land stock, the peasant sector nonetheless still takes home 62 percent of total income in the economy by virtue of their ownership of labor and other factors.13 Under a collusive arrangement of tied contracts, peasant incomes could in principle be pressed down to their autarky levels. That would leave the peasant sector earning just slightly more than they would from withdrawing from transactions with landlords and instead making − − use of their own land (T p
(1 )T /N) and labor (H) resources. This − would earn them their Chayanovian farm income F(T p, H). Point A indicates that at θ
0.8, where the typical peasant household has access − to T p
0.25 units of land, this income would be F(0.25, 1.25)
47.4, which is considerably below their income with competitive markets. 13 Recall that the efficient competitive equilibrium had w
r
0.45. Total imputed income from the peasant sectors’ ownership of 20 units of land is therefore 9
0.45  20, their income from owning 100 units of labor is 45
.45  100, and total farm profits (or returns to the nontraded S) from the 80 peasant farms equals 8
0.1  80. This sums to 62.

130

CONNING AND KEVANE

The figure also shows that the gap between competitive incomes and incomes in an economy with “tied contracts” increases rapidly as initial land ownership becomes more concentrated in landlords’ hands. This is because as θ rises peasants have ever smaller autarky plots and therefore worse and worse fallback positions. Suppose the distribution of property rights were such that the economy was at collusive equilibrium A in the diagram. The arrangement may have been sustained by limits on labor mobility. Consider what might happen next if a newly emboldened state intervened to limit all forms of landlord coercion and lift all barriers to peasant mobility within the agrarian economy. The state declares that existing property rights over land will be respected and enforced, but that involuntary servitude and coercive barriers to peasant mobility are prohibited. The one and only interference with the freedom to contract is that tied voluntary laborservice obligation contracts will be effectively banned via the legal stipulations that all land and labor transactions must be remunerated at “market-determined” wages and rental rates. Legislation of this sort has been common in Asia and Latin America as part of efforts to regulate or suppress bonded labor and labor-service contracting arrangements.14 Does a ban on tying lead to a quick transition to a new efficient competitive equilibrium at C with much higher peasant incomes, or somewhere else? Historical evidence suggests that the most common immediate outcome of slave and serf emancipation was the significant collapse of output in many societies (Bush 2000; de Castro 2002) and perhaps even a decline in some of the material conditions of the now free workers (Engerman in this volume). This question, on the nature of the transition paths and the kinds of agrarian capitalism that may emerge following expansion into a new territory or the abolition of coerced labor and other forms of voluntary “servile” labor, has been one that has vexed and exercised Marxist and non-Marxist historians for several generations (Byres 1996). When does the economy follow an “American road” leading close to a Jeffersonian ideal, in which production is dominated by family farmers and very active and competitive land and labor markets, and when does it follow a “Prussian” or “Latin American” road, where production instead becomes dominated by wage labor plantation agriculture or latifundia, in which peasant incomes remain compressed and peasant production confined to inefficiently small tenancies and minifundia? A simple extension of the model above suggests why the economy may very well not snap to a new competitive equilibrium, but rather settle on 14 For example, de Janvry (1981) discusses the impacts of twentieth-century legislation in Chile and other Latin American countries that required landowners to pay agricultural labor in cash rather than kind as well as enforcing minimum agricultural wage laws.

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

131

a distorted one with lower total output and only a scant improvement in peasant welfare. The argument, which is laid out in some more detail in the appendix and in Conning (2005), is that landlords’ reaction to a ban that deprives them of the use of tied contracts as an efficient mechanism for surplus extraction may be to turn to other privately profitable but socially inefficient mechanisms to earn rents. As discussed above, elites frequently acted to close down the outside income–generating opportunities of their newly untied laborers using both legal and illegal means. Our argument here is that that landlords would likely also seek to limit peasants’ outside opportunities via economic market mechanisms, without the need for (but in practice very likely in addition to) this kind of extra-economic barrier building.15 Suppose that following emancipation the economy temporarily moved to the competitive equilibrium indicated by C in figure 5.2. Starting from this efficient outcome, can landlords change resource allocations to raise their incomes? If landlords control a large enough amount of land, the answer is they can. Consider the simplest case where our M
2 landlords collude but cannot use tied contracts, so they must transact at “market” wages and rental rates.16 As discussed previously, competitive landlords would be rentiers, as the bulk of their income is from leasing out large amounts of land to tenants, and only secondarily from farm profits. If in such a context landlords have enough land to be able to exploit market power, they should limit the amount of land they lease to the market to push up land rents. This is the classic partial-equilibrium argumentation for a monopoly markup on land rents. But in this general equilibrium context there is an additional important effect that landlords must also consider. Landlords maximize profits plus the value of owned land. As they restrict land leased out, they increase earnings from land but also end up pushing down available land per worker in the peasant sector. With less land to work with, the marginal product of labor declines on peasant farms, leading peasant households to optimally increase their supply of labor to the market at any given wage. Since landlords are the largest (and in our constructed case, the only) net employer of wage labor, this benefits them by increasing profits. In short, the exercise of monopoly power over land also creates monopsony rents from labor. 15 In this sense the model offers microfoundations to help explain the pattern of elite behavior described by Ransom and Sutch and others, by providing a framework within which to identify and quantify the costs and rewards elites stood to gain via political and economic actions. It also offers a microeconomic explanation for the kind of inverse relationship between asset inequality and political and economic outcomes that Sokoloff and Engerman (2000) and others have identified in explaining the divergent growth paths of regions. 16 Similar results, albeit in somewhat more muted form, emerge if we instead model landlord interaction as a noncooperative Cournot game (Conning 2005).

132

CONNING AND KEVANE

Viewed slightly differently, the landlord strategy is to push down the land-to-labor ratio on peasant farms to lower their wage bill and at the same time raise land rents. The one thing limiting how far this strategy can be pushed is that there is a cost to running inefficiently large and land-intensive farms because of their inefficient use of the nontraded factor S. Under an efficient equilibrium nontraded skills are allocated evenly and efficiently throughout the agrarian economy, as the tenancy market works to match land to farming skills. Suppressing the tenancy market can only lead to a more inefficient allocation and therefore to a fall in overall output. Landlords will expand the size of their farms up to the point where the marginal gain from earning higher monopoly rents on land and/or monopsony rents from labor equals the marginal increase in this cost. The dashed line running through point D in figure 5.2 shows total income in the distorted economy at different levels of initial land inequality under the assumption that landlords collude. The line passing through V in figure 5.2 indicates the level of peasant sector income in the postabolition distorted economy. Peasant incomes remain well below their competitive market potential and scarcely above what peasants earned in their serf-like existence. Note that at higher levels of inequality θ, where landlords find it easier to exercise market power, the market distortion becomes even more pronounced. The gain to the peasant sector from a ban on voluntary or involuntary tied labor-service tenancy is smaller in an economy where peasants have few property claims to land, as landlords can then easily find “market” mechanisms to compel their labor at low cost. Taking stock of the argument so far, the model can be used to demonstrate that, although it may be individually rational for a peasant to accept a tied contract when all other peasants have entered into similar contracts (their payoff can be raised from A to some small amount above A), paradoxically a general ban on all tied contracts can raise peasant wages and incomes. Peasant welfare gain can be as large as moving from A to C if the ban leads to a new fully competitive equilibrium, but the gain is likely to fall considerably short of this if landlords react with new distorted allocations to extract rents. Peasant welfare then rises only from A to V. Landlord income and total output falls if the ban leads to a new distorted equilibrium. Although we have argued that peasant income rises, it is not too hard to imagine situations where peasant welfare could actually end up falling if the landlord reaction extended to include extraeconomic measures such as those described above and by Ransom and Sutch, or if as described below it creates new property rights conflicts over land. The model offers a plausible complementary explanation for why output should fall following abolition that is different from Fogel and

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

133

Engerman’s (1974) for the U.S. South that coerced labor produced more output per hour than free labor. Our model shows that the size of the output collapse and the resulting level of peasant welfare following emancipation, are affected by several factors including the initial distribution of property rights over land, the ability of elites to collude and of peasants to organize, and the role of the nontraded factor S. Consider further the role of land inequality. While the efficient way to produce 100 units in our numerical example was to allocate 80 percent of total production to peasant farms and the remaining 20 percent to landlords, simple simulations for our benchmark parameters show landlord farms expanding to account for more than 50 percent of output at θ
0.8 and expanding ever more rapidly at higher levels. As one increases θ to higher and higher levels, tenancy suppression becomes more active until the point where (at around θ
0.87) landlords find it optimal to close down the tenancy market completely. At this point, landlords decide that they earn more from monopsony rents on labor (since they have pushed the market wage far below the marginal product of a hired laborer), and they simply ignore land rent earnings (Conning 2005). At yet higher levels of initial land concentration , simulations show that landlords will actually start to encroach on peasant lands via reverse tenancy—large landlords will be observed hiring in land from small landholders. This behavior may make sense in a model where we have assumed that peasant property rights to land are scrupulously protected by the state, since landlords then only have markets to use as a strategy to manipulate factor prices and obtain the land they desire. In more realistic contexts, where property rights cannot be assumed to be costlessly enforced by a third party, one is likely to instead see the flaring up of conflict, for example land-hungry peasants squatting on landlord farms and common lands, or perhaps landlords hiring thugs not only to drive off squatters but also to restrict peasant access to frontier lands.17 One might interpret the drive to enclose common lands as responding in part to a similar logic. The discussion suggests that an “American road” trajectory, with active tenancy markets and prosperous family farmers, is likely to take root only following emancipation or the opening of a new frontier where land is distributed in a more egalitarian fashion (θ is low). A more distorted “Prussian road” such as that followed in Russia, Prussia, and many parts of Latin America, where landlords suppress tenancy and land sale markets, is more likely when initial land inequality is high. 17 Conning and Robinson (2007) explore a related model that shows how endogenously determined property rights conflicts would further suppress the operation of market for tenancies.

134

CONNING AND KEVANE

Recasting Debates The freedom to sell labor is a touchstone of classical liberal philosophy. Social conservatives and libertarians agree on this, even as they disagree over the value of freedom to decide what to do with one’s leisure time. By contrast, some modern liberals question whether certain contracts to sell labor should be regarded as inviolable privileges of personal freedom. Perhaps some kinds of freedom result in unfreedom. Some kinds of trade liberalization, it has been argued, may render a people worse off (Deardorff 2001). And democratic triumph by an undemocratic party may destroy a democracy. Clearly, constructing models or hypotheticals of exceptions to a basic intuition is a perilous pursuit. Special interests are always seeking justifications for antiliberal restrictions on freedom. One thinks, for example, of the many restrictions on women’s employment that favored male union workers (Goldin 1988, 1991). In the political domain, a stance of refusing to grant exceptions to the principle of freedom of contract may be a healthy strategy. But in very poor countries, people are living generation after generation in miserable drudgery and suffering early death. Asset inequality is high and political and economic power is often locally concentrated. These human beings, and the economies they inhabit, are of special interest. Might an otherwise healthy political strategy render disservice to the very poor and newly free? Sometimes former slaves, serfs, and bonded laborers have been emancipated into a daunting world of freedom of contract. Exceptions to the intuition of the desirability of freedom to contract labor matter more for them. Models and hypotheticals demand careful attention rather than cavalier dismissal; it is vital to think carefully about whether a poor laborer should be allowed or forbidden to accept a contract to bond his labor for a decade to an employer, and why such contracts are emerging in the first place. Models that analyze the choices that people make when they are free to buy and sell labor presuppose a broader account of the determinants of self-ownership and control over earnings of others. Several factors may determine which strategies might emerge; among them are the overall ratio of land to labor, the distribution of property rights and the concentration of land holdings, the ease of collusion, the importance of nontradable skill or managerial talent, and the role played by outsiders, the state, and other political factors. We have argued that a large variety of institutional settings can arise out of the interactions of landlords and peasants as the ratio of land to labor changes. Because of this, it may not be the case that liberating formerly coerced laborers always unleashes their productive potential so that economies grow over time. Moreover, it may be the case that there are

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

135

at times good reasons to regulate the freedom of contract. Prohibiting a landlord and a peasant from contracting would appear to stand in the way of obvious gains from trade. Yet in a purely competitive market there is no need for tied contracts, and use of tying by landlords may point to an attempt to control laborers’ outside opportunities. A prohibition on tying may result in less efficient collusive arrangements but can deliver a greater share of the lower output to the hands of the laborers. This is another example of the general theory of the second-best that is well used in economics: where labor markets may not be competitive to begin with and where other markets (e.g., for insurance) may not be perfect, government regulation or banning of certain voluntary contracts may improve general well-being or redistribute well-being in socially desirable ways. Stopping one household from using bonding or child labor worsens their welfare, but a uniform ban on bonded labor contracts (Genicot 2002) or child labor can produce general equilibrium effects that may lead to a better equilibrium (Basu 1999; Baland and Robinson 2000). Our explanation is related to existing arguments, yet in an important sense is also fundamentally different. Each of these other accounts focuses on externalities on labor and/or credit markets and the role of public policy in helping to solve the coordination failures that may prevent society from reaching a Pareto-superior equilibrium where every member of society can in principle be made better off. Our account focuses on the strategic action of local power holders to actively limit the outside opportunities of laborers.18 Society must overcome not a coordination problem but rather the local power structures that restrict the choices of the poor and prevent peasant welfare from rising closer to its competitive level. The discussion of an exception to the intuition that free contracts are invariably socially desirable should not make one sanguine about public policy. Governments are more likely to be handmaids of welfare-worsening institutions regulating labor relations, as in Russia, where the state prevented serfs from changing employers. The analysis of freedom and contracting in the context of labor must proceed dialectically, asking how an initial distribution of freedom and property generates a pattern of contracts, and how this pattern might then change subsequently. A famine might induce many laborers to willingly become serfs; their serf status may drive down the wage in the normal, nonfamine economy, inducing still more laborers to enter serfdom. An expanding frontier might induce laborers to leave their ancestral homes and employers in search of higher returns. Their former 18 Basu (1986) explores similar ideas in his analysis of examples of “triadic” power. Our approach allows for potentially more general comparative predictions across economies by virtue of being embedded within a more canonical neoclassical general equilibrium framework.

136

CONNING AND KEVANE

employers might respond to the rising cost of labor by using force or collusion to limit the opportunities available to laborers, and extra-economic coercion and anticompetitive labor tying might emerge as a reprehensible strategy in these circumstances. The limited freedom of laborers that comes from enslavement is however not necessarily the best choice for employers, who may occasionally find that less aggressive tactics garner greater returns. So the bright line between involuntary and voluntary labor arrangements is better redrawn as continuous shading. Appendix Here we provide some additional details and derivations. The reader is referred to Conning (2005) for a more detailed and general presentation of the problem. Efficient Competitive Equilibrium In an efficient equilibrium the marginal value product of each factor must be equalized across farms. If (Tpe, Lpe) is the equilibrium level of land and labor on a peasant farm, then efficient landlord farm inputs must be (Tre, Lre)
(SrTpe, SrLpe), since landlords use the same technology and have Sr times more of the nontraded factor. Factor market balance requires M ⋅ Sr ⋅ Lep + N ⋅ Lep = L and M ⋅ Sr ⋅ Tpe + N ⋅ Tpe = T from which we obtained the expressions for factor use stated in the text and which give (Tpe, Lpe)
(t, 1) for the assumed factor endowments. − − Peasant output is then qp
F(t ,1,1)
[t ] and equilibrium factor prices are w = FL (t , 1) =   t 



and r = FT (t , 1) =   t 

−

− At t
1 we have qp
1, w

0.45, and r

0.45. The shadow price of a unit of the nontraded factor S, which is also the value of peasant − farm profits, is
(1  )[t ], which is 0.1 for our parameters. The Slave Economy If 40 peasant households are enslaved and the other 40 remain free, equilibrium production must now takes place on two landlord and 40 peasant farms. Given the distribution of S, we must have Trs
10Tps, and

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

137

Lrs
10Lps , and land market equilibrium now requires 2 ⋅ 10Tps + 40Tps = T Tps =

T 10 = t 60 6

A similar calculation using the labor market equilibrium equation reveals that Lps
10/6
1.67. The institution of slavery blocks 40 households from becoming direct producers, which then makes room for each of the remaining farm units to expand approximately 67 percent compared to the equilibrium without slavery. Wage rates and rentals must fall as slaves’ nontraded farming skills are no longer being utilized as in the competitive economy. The Market-power Distorted Economy When Landlords Cannot Tie When tying is not allowed landlords must hire-in labor at a “market” wage w and land rental rate r. A “landlord cartel” would however collude to distort resource allocations to manipulate prices to their advantage. They would act to choose factor use on the typical landlord farm (Tr, Lr) to maximize G(Tr , Lr ) − r(Tr , Lr ) ⋅ Tr − w(Tr , Lr ) ⋅ Lr  + r(Tr , Lr ) T / M   Market labor wage w and the land rental rate r will be determined by the marginal product of labor w
FL(Tp, Lp) and the marginal product of land r
FT(Tp, Lp) on the competitive fringe of peasant farms. In the expression above we have written w and r as a function of Tr and Lr, by − using the factor market balance equations, Tp
T /N MTr /N and − Lp
L /N MLr /N. Market factor prices are determined by the marginal productivity of factors on peasant farms, but these will be manipulated prices because landlords will take into account how their own production choices affect the price and availability of land and labor in the peasant sector. The pair of equations that make up the first-order conditions for this problem are highly nonlinear, but the optimal solutions to maximize landlord cartel income are easily found numerically for given parameter values, as discussed in the text. Bibliography Akerlof, George A. (1976), “The Economics of Caste and of the Rat Race and Other Woeful Tales,” Quarterly Journal of Economics 90(4): 599–617.

138

CONNING AND KEVANE

Baland, Jean Marie, and James A. Robinson (2000), “Is Child Labor Inefficient?” Journal of Political Economy 108(4): 663–679. Bales, Kevin (2000), Disposable People: New Slavery in the Global Economy. Berkeley: University of California Press. Bardhan, Pranab, ed. (1989), The Economic Theory of Agrarian Institutions. Oxford, New York, Toronto, and Melbourne: Oxford University Press. Basu, Kaushik (1986), “One Kind of Power,” Oxford Economic Papers, N. S. 38(2): 259–282. Basu, Kaushik (1999), “Child Labor: Cause, Consequence, and Cure, with Remarks on International Labor Standards,” Journal of Economic Literature 37(3): 1083–1119. Bauer, Arnold J. (1979), “Rural Workers in Spanish America: Problems of Peonage and Oppression,” Hispanic American Historical Review 59: 34–63. Binswanger, Hans P., Klaus Deininger, and Gershon Feder (1995), “Power, Distortions, Revolt and Reform in Agricultural Land Relations,” in Jere Behrman and T. N. Srinivasan (eds.), Handbook of Development Economics (Amsterdam: North Holland), pp. 2659–2772. Braverman, Avishay, and Joseph E. Stiglitz (1982), “Sharecropping and the Interlinking of Agrarian Markets,” American Economic Review 72(4): 695–715. Brenner, Robert (1996), “The Rises and Declines of Serfdom in Medieval and Early Modern Europe,” in M. L. Bush (ed.), Serfdom and Slavery: Studies in Legal Bondage (London: Longman), pp. 247–276. Bush, M. L. (2000), Servitude in Modern Times. Cambridge: Polity Press. Byres, Terence J. (1996), Capitalism from Above and Capitalism from Below: An Essay in Comparative Political Economy. New York: St. Martin’s Press. Conning, Jonathan (2005), “On The Causes of Slavery or Serfdom and the Roads to Agrarian Capitalism: Domar’s Hypothesis Revisited,” Hunter College Department of Economics working paper no. 401, http://ideas.repec.org/p/htr/ hcecon/401.html. Conning, Jonathan, and James A. Robinson (2007), “Property Rights and the Political Organization of Agriculture,” Journal of Development Economics February. de Castro, Stephen Anthony (2002), “Wrong Incentives for Growth in the Transition from Modern Slavery to Labor Markets: Babylon Before, Babylon After,” texto no. 249, Departamento de Economia, Universidade de Brasilia, http://www.unb.br/face/eco/cpe/TD/249Oct02SDeCastro.pdf, accessed 9 July 2005. de Janvry, Alain (1981), The Agrarian Question and Reformism in Latin America. Baltimore: Johns Hopkins University Press. Deardorff, Alan V. (2001), “Rich and Poor Countries in Neoclassical Trade and Growth,” Economic Journal, 111(470), 277–294. Domar, Evsey D. (1970), “The Causes of Slavery or Serfdom: A Hypothesis,” Journal of Economic History 30(1): 18-32. Drescher, S. (2002), The Mighty Experiment: Free Labor Versus Slavery in British Emancipation. New York: Oxford University Press. Durlauf, Steven N., and H. Peyton Young, eds. (2004), Social Dynamics. Cambridge, MA: MIT Press.

FREEDOM, SERVITUDE, AND VOLUNTARY CONTRACTS

139

Edmonds, Eric, and Pavcnik, Nina (2005), “Child Labor in the Global Economy,” Journal of Economic Perspectives 18(1): 199–220. Engerman, Stanley (1973), “Some Considerations Relating to Property Rights in Man,” Journal of Economic History 33: 43–65. Fogel, Robert William, and Stanley Engerman (1974), Time on the Cross: The Economics of American Negro Slavery. Boston: Little Brown. Galenson, David W. (1984), White Servitude in Colonial America: An Economic Analysis. New York: Cambridge University Press. Genicot, Garance (2002), “Bonded Labor and Serfdom: A Paradox of Voluntary Choice,” Journal of Development Economics 67(1): 101–127. Goldin, Claudia (1988), “Maximum Hours Legislation and Female Employment: A Reassessment,” Journal of Political Economy 96: 189–205. Goldin, Claudia (1991), “Marriage Bars: Discrimination against Married Women Workers, 1920 to 1950,” in H. Rosovsky, D. Landes, and P. Higgonet (eds.), Favorites of Fortune: Technology, Growth, and Economic Development since the Industrial Revolution (Cambridge, MA: Harvard University Press), pp. 511–536. Hart, G. (1986), “‘Interlocking Transactions: Obstacles, Precursors or Instruments of Agrarian Capitalism?” Journal of Development Economics 23 (1): 173–203. Hayami, Yujiro, and Keijiro Otsuka (1993), The Economics of Contract Choice: An Agrarian Perspective. Oxford: Oxford University Press. Holmstrom, Bengt, and Paul Milgrom (1990), “Regulating Trade among Agents,” Journal of Institutional and Theoretical Economics 146(1): 85–105. ILO (2005), “A Global Alliance against Forced Labour: Global Report under the Follow-up to the ILO Declaration on Fundamental Principles and Rights at Work,” in 93rd Session 2005, Report I (B): International Labour Office. Kandori, Michihiro (1992), “Social Norms and Community Enforcement,” Review of Economic Studies 59: 61–80. Kliuchevsky, V. O. (1968), A Course in Russian History: The Seventeenth Century. Chicago: Quadrangle Books. Knight, Alan (1986), “Mexican Peonage: What Was It and Why Was It?” Journal of Latin American Studies 18: 41–74. Kolchin, Peter (1987), Unfree Labor: American Slavery and Russian Serfdom. Cambridge: Harvard University Press. Kolchin, Peter (2003), American Slavery, 1619–1877. New York: Hill and Wang. Krugman, Paul (1997), “In Praise of Cheap Labor: Bad Jobs at Bad Wages Are Better than No Jobs at All,” Slate, 20 March 1997. Lundahl, Mats (1992), Apartheid in Theory and Practice: An Economic Analysis. Boulder, CO: Westview Press. Marx, Karl (1887), Capital, vol. I: The Process of Production of Capital. Moscow: Progress Publishers. Morner, Magnus (1970), “A Comparative Study of Tenant Labor in Parts of Europe, Africa and Latin America 1700–1900: A Premilinary Report of a Research Project in Social History,” Latin American Research Review 5(2): 3–15. Nieboer, Herman Jeremias (1910), Slavery as an Industrial System: Ethnological Researches. The Hague: M. Nijhoff.

140

CONNING AND KEVANE

North, Douglass, and Robert Thomas (1973), The Rise of the Western World: A New Economic History. Cambridge: Cambridge University Press. Northrup, David (1995), Indentured Labor in the Age of Imperialism, 18341922: Studies in Comparative World History. New York: Cambridge University Press. Patterson, Orlando (1977), “The Structural Origins of Slavery,” Annals of the New York Academy of Sciences 292: 12–34. Ransom, Roger L., and Richard Sutch (2001), One Kind of Freedom: The Economic Consequences of Emancipation. New York: Cambridge University Press. Sadoulet, Elisabeth (1992), “Labor-Service Tenancy Contracts in a Latin American Context,” American Economic Review 82: 1031–1042. Sen, Amartya Kumar (1999), Development as Freedom. New York: Knopf. Sokoloff, Kenneth, and Stanley Engerman (2000), “Institutions, Factor Endowments, and Paths of Development in the New World,” Journal of Economic Perspectives 14(3): 217–232. Solow, B. L. (1993), Slavery and the Rise of the Atlantic System. New York: Cambridge University Press. Steinfeld, Robert J. (2001), Coercion, Contract, and Free Labor in the Nineteenth Century. Cambridge: Cambridge University Press. Wakefield, E. G. (1833), England and America: A Comparison of the Social and Political State of Both Nations. London: R. Bentley.

Chapter Six Slavery and Slave Redemption in Sudan JOK MADUT JOK

As I read the thoughtful pieces by my co-contributors in this volume, I am struck by the complexity of attempting to apply the terms of debate about the “effects” of slave redemption to the particularities of the Sudanese experience. That this is so, applies as to the nonacademic protagonists in the debate as much as to anyone else. Throughout the late 1990s, when the reemergence of slavery and slave redemption in Sudan were hotly debated issues, especially in the United States as well as in Sudan, one of the main characters in this drama was the government of Sudan, which argued bitterly against the claims that slavery is practiced in Sudan. Upon the realization that these abduction and enslavement issues were too embarrassing internationally, the government of Sudan felt that it needed a body to respond to the allegations in a more concerted way. It set up an organization known as the Committee for the Eradication of Abduction of Women and Children (CEAWAC). Despite the documented involvement of the government itself in raiding, abduction, and enslavement, CEAWAC came to the defense of the government and tried to explain away the whole phenomenon as the work of detractors who wanted to smear the good name of the Islamic government in Khartoum. The “Americans” and the “Zionists” were named in the mission to defame Sudan. Another character in the debate was Christian Solidarity International (CSI), a Zurich-based evangelical organization that underwrote most of the slave redemption programs and whose representative actually traveled to south Sudan on a rather frequent basis to undertake the actual buying back of the slaves through Arab intermediaries to secure their freedom. Conducted clandestinely and without clear methods of verification by independent researchers or journalists, CSI’s work became the subject of concern for those who regard slave redemption as a wellintentioned practice but one which can invigorate slavery, as it promotes slave-taking by providing the market for abductees. Other influential parties contributing to this debate were two child welfare agencies, United Nations Children’s Fund (UNICEF) and Save the Children (SC-UK). These two were among those that regarded slave

144

JOK MADUT JOK

redemption as fueling slave-taking, and were heavily critical of it. It will be recalled that, contrary to some allegations, UNICEF was not denying the existence of slavery in Sudan, but was only critical of the myriad responses it generated. In fact, it was their own UN Commission for Human Rights that had documented this most dehumanizing of human rights abuses through the special rapporteurs, Câsper Bîro, Leonardo Franco, and Gerhardt Baum. All of the three rapporteurs provided unequivocal evidence for the government’s use of slave raids as a counterinsurgency measure in its war against the South (see for example High Commission for Human Rights 1999b). Based on these reports, UNICEF and other aid agencies and human rights groups also made the point that instead of placing a monetary value on human freedom—something that is inherently free—it would be better to work with the government of Sudan or pressure it to desist from the use of slavery as a way to fight its war against the South by proxy. Instead of paying for the freedom of these slaves, the UN and major human rights groups such as Amnesty International and Human Rights Watch argued, the effort should be one where the government of Sudan is forced to live up to its responsibility of protecting its own citizens, as a government is expected to do. As for the involvement of the American public in the debate, this was the work of the American Anti-Slavery Group (AASG), a Boston-based activist group that raised awareness not only about the presence of abduction and slavery in Sudan, but also about the ever-present modern forms of slavery throughout the world. They rallied much of the American sympathy for Sudan’s slaves and were able to generate money to support CSI redemption campaigns. The campaigns, led by AASG to spur the American public as well as the international human rights community, were so effective that they managed to persuade a wide array of people to take collective action, forming a Sudan constituency of varied sectors of the American society that would otherwise have very little in common. The members of this constituency ranged from schoolchildren who collected their lunch allowances for this cause, to elderly Jewish retirees in Florida, to black Christian congregations in California, all of them donating money for redemption and for awareness campaigns. This was possible through public protests and by the use of the internet— AASG runs one of the more well-organized human rights websites, antislavery.org. They wrote petitions to both the executive and the legislative branches of the American government, urging the officials to come to the aid of Sudanese slaves. AASG also testified in the U.S. Congress, picketed the Sudan Mission to the United Nations in New York, and persuaded such influential civil rights activists as the reverend Al Sharpton and Washington D.C. radio talk show host Joe Madison to speak against the atrocities committed by the government of Sudan. Joe Madison himself

SLAVE REDEMPTION IN THE SUDAN

145

traveled to Sudan to redeem slaves to make a journalistic point that slavery exists in Sudan and that in this day and age, one can still purchase slaves for 50 U.S. dollars a piece. AASG got a hold of an escaped Sudanese slave and recruited him as a spokesman for the organization. Francis Bok, a tall, slender, and engaging young Dinka man, toured the United States and gave emotional speeches about his own captivity. He gathered sympathy for the Sudan slaves from rock music concert crowds that AASG enabled him to address and he moved church congregations across the United States. But the most surprising success scored by AASG was in creating the strangest legal bedfellows by getting Johnnie Cochran, the famed OJ Simpson defense attorney, and Kenneth Starr, the special prosecutor in the Clinton-Lewinsky sex scandal, to join hands in defense of two Sudan activists who had been arrested for chaining themselves to the front gate of the Sudan Mission to the United Nations in New York in protest of the Sudan government’s use of slavery as a war tactic. The debate about redemption quickly polarized the opponents of Sudan’s slavery into two camps. On one side there were agencies, including the UN and the major human rights groups, that advocated that the focus in the fight against slavery in Sudan should not be on slave redemption but rather on pressuring the very government that perpetrated the practice to desist from it and protect civilians. They suggested the use of diplomatic pressure on the government to end the practice as well as working with it to find ways to release the slaves and punish the culprits. And on the other side there were groups, including AASG, CSI, and a number of other groups and Sudanese diaspora communities, that denounced such a stance as amounting to doing nothing or even direct complicity. Both groups have generally agreed that redemption is not the solution, but the latter group says that engagement with the government of Sudan, a government well known for its intransigence on the slavery issue, without an alternative plan to immediately rescue the abductees is irresponsible. “To sit and do nothing while the slaves rot in captivity until the government has decided to be more humane is negligent if not outright complicit,” said one “abolitionist” at a Harvard and Tufts University student demonstration in Boston in 2000. The groups that have supported slave redemption, including a member of the British House of Lords, Baroness Caroline Cox,1 have all made the point that the current Sudan government, a repressive Islamic extremist regime, has been in 1 Baroness Cox ran a human rights group named Christian Solidarity Worldwide, and has herself made over twenty missions to redeem slaves in south Sudan. When reports of impropriety regarding redemption began to surface, she was much taunted and considered naïve by a number of journalists, a labeling that only increased her resolve to expose the crime of slavery.

146

JOK MADUT JOK

power since 1989 and no one knows when and how it will leave power; and while the endeavor to influence its policy vis-à-vis the slaves is a noble effort, the approach to the question of slavery must necessarily be multifaceted. Slave redemption is definitely one facet, which, while it does not tackle the root causes of the phenomenon, makes a difference in the lives of individual slaves whose freedom might thus be secured. The result of this debate has undoubtedly been the elevated awareness of the American public about the plight of South Sudanese slaves. It is particularly the infamous redemption trips and the attendant media coverage that put the problem of slavery in Sudan squarely on policymakers’ agenda. For this, many South Sudanese organizations and communities have expressed their appreciation of these groups, and their indebtedness to the Americans who are involved. But in reaction to this, the government of Sudan’s most popularized rebuttal was an effort to make the usual unconvincing connection between these antislavery campaigns and Zionism. Indeed, antislavery advocacy, slave redemption, and the entire South Sudanese opposition has been framed in Khartoum as the work of anti-Arab, anti-Muslim Zionists. The fact that AASG has been headed by a man who happened to be Jewish also came in handy for the Sudan government’s apologists. The increasing visibility of Christian groups from around the world in this debate, and indeed the framing of Sudan’s conflict as one between the Muslim north and the Christian south, has armed the government of Sudan in its effort to attract military, economic, and moral support from the Islamic world. Thus, the slavery question has increased the polarization of the two sides to the Sudan’s conflict in more ways than most issues have done in the past. Since 2000, however, many events have happened that have reshaped this debate. One is the ascendance to the presidency of the United States of George W. Bush, who, contrary to expectations that he would not concern himself with foreign issues, has taken up the attempts to resolve the conflict in Sudan as a priority foreign policy issue. This has moved the peace process between the government of Sudan and the opposition Sudan People’s Liberation Army (SPLA) to the negotiating table and the signing of a comprehensive agreement to end the two-decade-long war. This development has led to halting of slave raids. So although there has been no talk of how to get the slaves already in captivity out of the North and return them to the South, there have been no new slave raids since the winter of 2001. Another event was the media reports claiming that the famed slave redemptions were a hoax all along, faked, and that these unsuspecting redeemers were duped. A Washington Post report written by Karl Vic, a similar story carried by the Irish Times, a TV special program by the American journalist Dan Rather, and a host of other reports, all spoke of redemption having been a scamming enterprise. Local authorities

SLAVE REDEMPTION IN THE SUDAN

147

were accused in these reports of taking redeemers for a ride by presenting fake slaves for redemption, rounding up some children and women from nearby villages, and presenting them as slaves who had just been brought to the South by Arab middlemen from the north. These reports have sparked suspicion not just regarding redemption, but about the nature of slavery itself. The fact that there were so many reports about redemption being a hoax, the fact that these slave buy-back missions were not open to independent verifiers, and the tendency among some organizations to inflate the volume of slavery with the assumption that it would earn the slaves more sympathy, have all weakened the position of these antislavery groups. On the one hand the opacity of the process challenged their credibility, and on the other it provided the government of Sudan with room to explain away the entire phenomenon of slavery by arguing that it had come to be an issue in the first place because Western Christian and Zionist zealots and their friends had seized on unproven allegations in order to bring down an Islamic government they would have wanted to bring down anyway. Government agents have seized upon any slight discrepancy in the facts as proof that the whole saga has been a lie all along. The most convincing argument against slave redemption, however, is the economic one alluded to by UNICEF and Save the Children: that paying for slaves only encourages slave-takers to abduct more in order to sell them back to the redeemers. This point was the subject of a number of journalistic pieces, including a rather well-argued one published in Atlantic Monthly. From an empirical perspective there is no evidence that slave redemptions have led to increased raiding, or have increased volume of slaves taken since the programs began. Two factual points need to be illuminated here. One is that slave-raiding is not merely driven by the economics of slavery. Because Sudan’s slavery is driven almost solely by its usefulness as a war tactic—to destabilize the support base of the Southern opposition armies by destroying property and displacing the civilian population—it has been able to go on for as long as it has regardless of whether or not redemption happens. The second point is that redemption had been an ongoing practice since the start of the current wave of slavery in the 1980s, initiated by families to get their loved ones back through their contacts among the Arabs of Darfur and Kordofan, long before the involvement of international groups. The resurgence of slavery in the early 1980s relates to three main factors. The first is the ethnic, ecological, and economic complexity of the region known as the “transitional zone,” where the Dinka and the Baggara share pastures and a major river (Kir) that falls right between the two groups. Both groups’ economic activities are tied to a cattle culture, as

148

JOK MADUT JOK

they follow a seasonal pattern of migration to and from the pastures. One important event that changed this balance was the drought that hit the Baggara territory in the 1980s, causing a major famine. The result was that the Baggara sought more pastures in Dinka territory further and further south across the Kir River, far beyond the zones traditionally permitted by the Dinka on a seasonal basis. The second development was the rise of mechanized farming in Darfur and Kordofan in which the Khartoum government supported specific groups to take over large tracts of land to the disadvantage of others, whether farmers or herders, causing socioeconomic dislocation in the process. Related to this were the government’s large-scale agricultural irrigation schemes, which undermined the food security of the rural population by reducing their access to arable land, therefore propelling them toward competition with each other. A combination of these factors created a rather explosive situation of what had already been a tense relationship between the Baggara and the Dinka on the one hand, and the state versus marginal farmers and pastoralists on the other hand. The result was that the Baggara have tried to deal with their losses and impoverishment by turning their wrath toward the Dinka (Johnson 2003, 157). Having been unhappy with the central government and having even militarized their criticism of the government, Khartoum has seized on the Baggara discontent and tried to appease them by supporting their activities against the Dinka, by offering them arms and turning a blind eye to their atrocities in the south. The third issue was the resumption of the second round of the NorthSouth war in 1983—the first round had ended in 1972. As the Khartoum government tried to counteract the insurgency and finding itself fairing poorly on the battlefields, it resorted to the use of tribal militias as a cheap counterinsurgency weapon—a way to fight the war by proxy. The abovementioned Arab cattle-herding tribe, the Baggara, proved most effective in this regard, and were armed and deployed to raid the Dinka, who were in turn stigmatized as the main supporters of the opposition armies in the South. In other words, the government scored on two accounts: it muzzled the Baggara political and economic complaints by appearing to side with them, and through the use of Baggara militias, it denied southern opposition armies their support base by destabilizing the Dinka areas through slave raiding. These factors made the resurgence of slavery primarily a function of the government’s use of slave raiding as a weapon of terror aimed at civilians who are suspected of giving support to the insurgents. It is also related to the question of dwindling resources in the transitional zone, the increasing Islamic radicalism, and the tense Afro-Arab race relations. And the most effective way the government has tried to execute its

SLAVE REDEMPTION IN THE SUDAN

149

agenda of counterinsurgency, as shown by its actions toward the people of the South, is by targeting the family—women and children (Johnson 2002, 157). Such technique is made possible by means of military activities to insure physical disruption of family life through raiding, stripping or destruction of assets, killing of men, and forced relocation of women and children to government-controlled areas. This has been clearly evident, since the start of this round of the war, in Khartoum’s attacks on villagers everywhere in the south. These attacks are vicious, aimed at civilians, and are executed either directly by the regular army, through aerial bombardment, or by use of militias recruited from the northern tribal groups hostile to southerners (Christian Aid 2001). Despite refusal to apply the term “slavery” and referring instead to “abduction,” a 1998 UN report admitted for the first time that a large number of women and children were abducted in that year, and were transferred to the areas of the Arab Baggara (UN Commission on Human Rights 1999a). The abduction was carried out by the Baggara-based militia known as the Murahileen, which captured these women and children as war booty. This UN report, although a sign of an overdue international recognition of the problem, was a mere scratch on the surface of a much more elaborate government plan of genocide, ethnic mixing, or cultural reconfiguration that targeted the South. The government denied involvement or complicity in slavery but it took no action to halt these practices. But some earlier studies have shown that there were really three different phenomena in the practice of slavery and abduction. First, there was armed and organized raiding in which the role of the government was not clear, and was likely complex. There is information that sometimes the government provided arms, other times the groups of Murahileen went off on their own. Northern tribal groups have been known to organize raids with “representatives” from other Arab groups; returning with children, women, and cattle taken in these raids, all of them have had a common celebration (Jok 2001 and Collins 1992). Then there was the state-owned military supply train escorted by the Popular Defense Forces (PDF) which traveled slowly from the north down through contested territory of Aweil and Wau in Bahr el-Ghazal. It is evident that there was formal recruitment by the government of militias to guard the train from possible SPLA attacks. These Murahileen (or Mujahideen, holy warriors, as they were sometimes called) then went out from the train and attacked villages suspected of supporting the SPLA on the way from Babanusa to Wau and back. According to eyewitness accounts that we have received over the years, the Murahileen used to ride on horseback along both sides of the railroad tracks, fanning out to a radius of up to 50 km, and systematically raided villages, torched houses, stole cattle, and killed men who put up any resistance. Their booty consisted not

150

JOK MADUT JOK

just of goods, but also of women and children to be subjected to various forms of enslavement and other slavery-like practices. Often, abducted women and children were taken up to the north and would remain in the possession of the captors or other persons until such time as they escaped, got released, or were redeemed through payment. Many more, the exact number being unknown, remain in captivity or otherwise missing to this day. Finally, it was often reported that there were joint punitive raids carried out by the government and the Murahileen, who, under the Popular Defense Act, enjoy a status as state-sponsored militias, and were called the Popular Defense Forces (PDF). Throughout the years of the war up until the cessation of slave raiding in 2001, the raids by the militias constituted a major source of violation of human rights, as they were not subject to any legal measures for their destructive activities (Jok 2004). Our interviews and many NGO reports contain lengthy and detailed testimonies of women and children abducted and kept in these circumstances, who regained freedom only by escaping or through ransom. The captors were often referred to as PDF, Murhileen militia, or sometimes even as government soldiers. It is, therefore, doubtless that regular troops also took part in or facilitated the raids. According to certain accounts, the perpetrators were said to be wearing uniforms, whereas Murahileen and other militias usually wore plain cloths. Although an auxiliary force, the PDF were directly under the control of the Sudanese authorities, and sometimes conducted raids jointly with the regular Sudanese troops. Moreover, the militias were armed by the Sudanese authorities and received ammunition from the Sudanese army. Raid preparations followed established patterns, leaving little doubt as to the Murahileen’s intentions of capturing women and children for slavery practices. The overall abduction and enslavement situation that emerges is complex, but with a distinctly violent character. In areas where the government of Sudan’s military forces and their militia allies confronted the main forces of the Sudan People’s Liberation Army, there was intense fighting taking place in the context of rough strategic parity. In the South, this was chiefly in the more westerly parts of Upper Nile province and Bahr el-Ghazal province. Extremely heavy and destructive fighting in the Aweil area of the Bahr el-Ghazal region was characteristic of the period between the start of the war in 1983 and the ending of raiding in 2001, as reported by a number of sources over the years.2 Civilians, of course, continued to pay the heaviest price. For example, aid agencies estimated 2 In the course of my research for the book War and Slavery in Sudan, I conducted over 200 interviews with figures of authority, heads of families, chiefs, and SPLA military personnel who had fought in defense against the Murahileen, all of whom were sources of information on raiding patterns.

SLAVE REDEMPTION IN THE SUDAN

151

that these wars were responsible for nearly half of the 4.5 million internally displaced persons (IDPs) in Sudan. Among the displacements due to slave raiding, estimates have put the Dinka of Aweil IDP population at 200,000 civilians, with the death of thousands more in the period between 1998 and 2001 alone (CARE International 2001). The raids have caused disruption of Dinka economic activity as well as relief services such as child immunization campaigns and feeding programs for undernourished children. On the other hand, where the government of Sudan did not encounter major SPLA military resistance, their rapidly growing ascendancy became clear. Here the military situation was ominous in the extreme for the civilian populations. In an extension of the scorchedearth tactics that created displacement of populations throughout the transitional zone since 1983, Sudan gained the reputation of being the country with the world’s largest internally displaced population. Although a tense debate over Sudan’s slavery has gone on for over two decades, such a debate has been over the volume, the causes, who benefits from it, or how to stop it; there has been little doubt that slavery exists. One important development that has helped put that doubt to rest is the return of a number of Dinka children from captivity. These children had been in bondage for some time, and were able to return either through escape or redemption. The redemption was the result of longstanding efforts by Dinka families to locate and return their children from enslavement. By the time international redemption groups became involved in the mid-1990s, redemptions had been going on since the start of the current wave of slavery in the 1980s, initiated by families to get their loved ones back through their contacts among the Arabs of Darfur and Kordofan. Redemption had operated on the basis that a Dinka family would sell some of their belongings, usually cattle, and seek an Arab intermediary; trace the whereabouts of their abductees, and initiate a negotiation with the abductor or the current holder and settle on the price; the abductees would be released to the intermediary, who in turn guided the abductees back to their families. Although it had gone on for a long time, this was a limited-scale operation because people could not afford the cost, and so when antislavery groups found out about it, they began to support many more families to secure the freedom and return of their members through these clandestine channels. This support surely went a long way for the families who gained the freedom of their members without incurring too much expense in terms of property. When more and more people began to be released through these channels, more money began to pour in, and with more money coming into the small cash-poor economy of south Sudan, the local authorities began to get involved, which initially meant nominal taxes levied on the incoming foreign currency through control of the exchange rate. More slaves bought

152

JOK MADUT JOK

back resulted in more media coverage, and in turn more money collected from sympathetic individuals, particularly in the United States. With significantly larger amounts of money involved, the local authorities then instituted a policy that all redemptions had to be channeled through them. This led to suspicions that by imposing these regulations, the SPLA authorities wanted to get a cut of the redemption money. But in fact what the authorities wanted was the opportunity to garner hard currency by being the ones to change dollars for Sudanese dinars—the currency used for redemptions. The dollars were then put to use in procuring military equipment from the neighboring countries with a view to preventing further raiding even as the families got their abducted members back. This is what brought the issue of slavery to the world’s attention. The government of Sudan was infuriated by the news reports that slavery was thriving in Sudan, and it went to great lengths to deny these reports. It became a question of national pride for government officials to defend the nation from such damning reports, and both slavery and redemption were described in the Sudanese state discourse as a part of effort by Zionists and American Christian zealots to destroy the good name of an Islamic state. So, whether or not redemption programs might in theory have the economic effect of fueling slavery, in practice, redemption served to underwrite the military efforts that prevented further raiding by giving the local military authorities the chance to convert their local currencies into foreign money that they could use to purchase arms in the neighboring countries. And it did this without the process becoming a direct assistance toward military defense. Further, since mass redemptions were not possible in such a volatile environment, the redeemers provided a great deal of help to the local economy, but not the kind of help they thought they were giving. Redemption could not have fueled slave taking, since the money was not really being passed on to the alleged slave-takers in foreign currency; but it facilitated local trade between the Dinka and traders in the neighboring countries in East Africa. For example, once foreign redemption agencies arrive with dollars, the local traders and the opposition armies would go and buy the dollars from them; the redeemers would then go one direction to buy back the slaves with Sudanese dinars, and the Dinka would go in the other direction to obtain goods in Uganda or Kenya that they could not have bought with the Sudanese currency. There have been other sources of information that put this discussion almost to rest. Between 2001 and 2004, the Rift Valley Institute has carried out a project to register all the abductees from the seven counties of northern Bahr el-Ghazal most affected by raiding. This was possible through the use of the Dinka kinship and administrative structures, a system that allows every individual to be accounted for. The project employed 48 local researchers who traveled from village to village to record

SLAVE REDEMPTION IN THE SUDAN

153

details of raids, names of authority figures, names of people killed, estimates of property destroyed, and the names of abductees, including their gender, their identifying clan, and the name of the village or cattle camp from which they were abducted. This registry, although not complete, has generated a database that looks like a missing persons report, and has 12 thousand abductees in it—not estimates, but real people with real names whose relatives are known. This database will assist in tracing and reunification, should the political situation becomes amenable to such an exercise. But beyond this, the database dispels without a shred of a doubt the question about the existence of slavery and other gruesome rights violations. It has to be noted that enslavement is only the sharp end of a long process of violent activities, starting with an attack on a village, burning of property, killing of those trying to defend the village, and chaining and dragging of captives across the rough terrain between the south and the north. So, long before the fate of the abductees is decided by the captors, and long before the much-debated redemption becomes an issue, there are a host of rights violations that are worthy of their own attention.

Addendum During my trip to south Sudan in the summer of 2004 it was clear that the abduction issue was subordinated to the hopes for peace, and there seem to be a desire, among both the officials of the SPLA and some of the chiefs, to focus on the future rather than push for restoration of justice to the victims of abduction and slavery. It is surprising that this nonchalant attitude about the right of the victims to assistance to return home or for compensation is happening in an environment where the families of the victims, some former slaves, and the international human rights circles are discussing how measures of such restoration of justice could be built into the peace process. Apparently, because these voices are calling upon the international community, especially on the mediating governments in the peace process, to have abduction and slavery addressed at the peace talks, there is an increasing fear in Khartoum that evidence of the government’s involvement in abduction and slavery might become clear when a peace deal is reached and an open investigation into these crimes may finally be conducted. This fear is especially intense because of the failures of CEAWAC to achieve its goals. CEAWAC had been set up to paint a picture that the government is working to end abduction, compensate the victims, and facilitate the release and return of the abductees. The funds for CEAWAC operations had been provided by UNICEF, and expertise in child protection issues was provided by Save the Children. But at the

154

JOK MADUT JOK

realization that CEAWAC was a costly and ineffective agency, and perhaps even a façade set up in order for the government to be seen as doing something about abduction, the funding was stopped. The result was that CEAWAC resorted to the Sudanese president’s office for funding. And the central government, we were told, decided that they wanted to deal with this abduction issue once and for all. The government then approved an amount said to be about 200,000 U.S. dollars for CEAWAC in 2003. It has to be borne in mind that any information that involves statistics or numbers of any kind must be taken with a grain of salt, as the deficit in accurate information is one of the consequences of the political conflict in Sudan. It is especially important to use caution when dealing with this kind of accusatory discourse, which is a common feature in Sudanese contexts and leads to blurring of the lines between real information and rumor. In any case, the money was to be shared between CEAWAC and other member agencies such as the Dinka Chiefs Committee, a Khartoum-based agency purporting to trace and reunite abductees with their families. This group of agencies were then given instructions to find these abductees and facilitate their return to Bahr el-Ghazal. The agencies then purchased trucks and a number of four-wheel-drive vehicles to use in transporting whatever abductees they can find. Not having done their homework, the agencies rounded up any Dinka persons in Khartoum and in a number of towns in south Darfur and west Kordofan who were allegedly abductees and willing to be transported to Bahr al-Ghazal. Many of these people were actually displaced persons, some of whom had moved to the north voluntarily to escape the fighting or famine. Such preemptive migration was a common occurrence in the late 1980s and again in 1998, when a major famine occurred in Bahr elGhazal. We interviewed some of these returned abductees. Some of them confirmed that they had indeed been abductees, but that they had gained their freedom somehow and had established a degree of normal life in the northern cities. Some of them said that they had jobs or businesses and owned property in the north; and now “we were repatriated by force, we have been abducted again,” they said. The agencies had loaded them onto the trucks, most of them women and children, and they were taken to such places as Malualkon, Akon, Warawar, Alek, and many other locations in northern Bahr el-Ghazal. The exact number transported in this manner was not ascertained, as we were not able to travel to all the locations where these alleged returning abductees were dropped, but the conditions they found themselves in were appalling. In other words, in response to the imminent peace deal between the north and south to end Africa’s longest war, the victims of slave raiding were abducted and forcibly returned to the South as the government embarked on a program to white wash its involvement in slavery just before peace came. But the

SLAVE REDEMPTION IN THE SUDAN

155

conditions of the returned abductees were so incredibly pitiful that the main welfare agencies operating in the area such as UNICEF and Save Children refused to be part of this operation. We were told that these agencies had disagreed with the manner in which the operation was conducted. There was no adequate coordination with the agencies or with the local authorities. There were no systems in place to speed up a process of unifying the returning abductees with their families upon arrival. The agencies felt that the whole operation was not done right. No provisions for the protection of child rights were made. There was no proper investigation of the status of individuals to ascertain whether they were abductees or IDPs of another kind. There were no proper interviews conducted with the abductees to insure their willingness to return to the south. No measures of care were set up in the receiving locations prior to their return so that proper reception was set up beforehand. The result was that when the abductees were returned and “dumped” (as the abductees describe their return) in the south, this was done without basic services being provided. No provisions were made to feed them during the time it would take to find family members, not to mention the lack of clean drinking water, cooking utensils, water containers, shelter, and other basic needs. This was especially deplorable considering that identification and reunification was not an immediate process, the original home villages of the abductees were not identified beforehand, which meant that some abductees were returned to areas so distant from their own villages that no one among the host communities could recognize them. This situation became much harder in the case of people who may have been abducted at a young age and could not be clear about their families, clans, figures of authority, or names of villages. Many such people were still stranded in the camps at the time of our visit in July 2004, and the possibility of ever finding their relatives is next to nil, especially since no one is actively looking into their cases. We met and interviewed a few such boys in Alek in Gogrial county. Some of them speak of a return to the north. Indeed, a few have reportedly already returned to the north. Given that they had been rounded up without insuring the consent of the abductees, the international welfare agencies such as UNICEF refused to get involved in the care of them. However, CEAWAC and the SPLM authorities in Bahr el-Ghazal had agreed that CEAWAC would set up a camp and provide services for one month, during which tracing of relatives and reunification would take place. But such services were not provided, and much suffering ensued. Throughout the summer of 2004 some women and children died of starvation, and two women gave birth in open air. Some field staff from other agencies whom we interviewed

156

JOK MADUT JOK

said that they had reported this situation and recommended intervention on behalf of the abductees to save them from perishing, but were ignored or were told that they should not become part of this haphazard forced return. The local authorities also had a very surprising reaction to this situation. They initially expressed enthusiasm about the “return of our people,” we were told. But upon the arrival of these people, both military and civil authorities of the SPLA/M (Sudan People’s Liberation Army/ Sudan People’s Liberation Movement) did utterly nothing to receive them. Indeed, they were observed to have expressed no interest whatsoever in helping the returnees to find their families, not to mention their outright negligence in the eventual death of some of the abductees. Surprising developments indeed came out of this operation. Because there was no proper coordination between CEAWAC/DCC (Dinka Chiefs Committee) and the agencies on the ground to receive the returnees, the government eventually allowed CEAWAC to send a team of medical students from Khartoum to travel to Malualkon in Aweil East County to supposedly look after the returned abductees. Team members were welcomed by military authorities in the area and were allowed to travel all the way up to the military headquarters of the SPLA area command, a development that eventually caused a great deal of conflict within the SPLA, since some officers viewed this as a breach of SPLA security. The team ended up not being able to attend to the abductees whom they had come to care for. The other surprising development was the revival of Christian Solidarity International interest in the area of abduction. CSI, formerly unable to see eye to eye with CEAWAC, and its bitter enemy over the question of abduction, slavery, and redemption, was now turning a blind eye to this mass repatriation. Again, this was a clandestine operation where the returnees were once more faked abductees rounded up on the streets of Khartoum and northern towns and dumped in the South without their consent. In the past UNICEF, Save the Children, and CEAWAC were trying to work together, and CSI was bitterly opposed to the approaches of both international child welfare agencies and to CEAWAC. Now alliances and disagreements are being revisited between CSI and CEAWAC. While the agencies continue their war among themselves, the victims of abduction continue to suffer. The real abductees remain held somewhere in northern Sudan, and very little is being done to find them. It is the displaced persons, who undoubtedly need a different approach to assist in their return, who are now being repatriated in a manner unsuitable to their status. They are being rounded up and dragged to the South with neither their

SLAVE REDEMPTION IN THE SUDAN

157

agreement nor provisions to look after them or integrate them into their communities. Because of this experience, many people, among both the returnees and the local population, say that the official line was hard to believe because of the apparent lack of interest among the local staff about “returning our people.” People were particularly unhappy with the way the local authorities did not seem to be bothered by this “dumping,” to use their word. The returning abductees themselves spoke of having been abducted again from the north and forced to come back to the south. Bibliography CARE International (2001), Internal Displacement: Causes and Remedies. Atlanta: CARE International. Christian Aid (2001), The Scorched Earth: Oil and War in the Sudan. London: Christian Aid. Collins, Robert (1992), “Nilotic Slavery: Past and Present,” in Elizabeth Savage (ed.), Human Commodity (London: Frank Cass), pp. 210–231. Deng, Francis M. (1995), War of Visions. Washington, DC: Brookings Institution. Harker, John (2000), Human Security in Sudan: The Report of a Canadian Assessment Mission (part 1). Prepared for the Minister of Foreign Affairs, Ottawa. Hendrie, B. et al., eds. (1996), Operation Lifeline Sudan: A Review (independent consultant report). Human Rights Watch (HRW) (1999), World Report: Events of December 1997– November 1998. Washington, DC: HRW. Johnson, Douglas (2003), The Root Causes of Sudan’s Civil Wars. Oxford: James Currey. Jok, Jok Madut (2001), War and Slavery in Sudan. Philadelphia: University of Pennsylvania Press. Jok, Jok Madut (2004), “The Targeting of Civilians as Military Tactics” in Ann Lesch and Osman Fadl (eds.), Coping with Torture: Images from the Sudan (Trenton, NJ and Asmara: Red Sea Press). UN Commission on Human Rights (1998), Report of the Special Rapporteur on the Situation of Human Rights in the Sudan. Geneva: Document E/CN.4/1998/66. UN Commission on Human Rights (CHR) (1999a), Question of the Violation of Human Rights and Fundamental Freedoms in Any Part of the World: Situation of Human Rights in the Sudan, E/CN.4/1999/38/Add.1. UN Commission on Human Rights (CHR) (1999b), Situation of Human Rights in Sudan: Visit of the Special Rapporteur Mr. Leornardo Franco, 13–24 February 1999, Addendum to E/CN.4/1999/38. UN General Assembly (UN GA) (2000), Human Rights Questions: Human Rights Situations and Reports of Special Rapporteurs and Representatives: Situation of Human Rights in the Sudan, A/55/374.

Chapter Seven Dilemmas in the Practice of Rachat in French West Africa E. ANN MCDOUGALL

“We have bought a slave. We have shown that the traffic still exists.” [Timbuktu, 1958]

“The Koran says that the best thing a man can do is to give a slave his liberty. This man [from whom the slave was purchased] has only sold a slave. It is you . . . who have given Ibrahim his freedom.” [Saharan freeman, Timbuktu]

“Though I know that I am free in Timbuktu, I also know that I still belong to my master. For how can I be really free? I know that when the French leave the country my master can take me again. . . . That is why I still feel myself a slave.” [Freed slave, Timbuktu]

“[He can go to the Commandant. . . .] And the Commandant will tell him that by law he is not a slave, and by law his master is not his master. But the slave probably knows that already. In his head he knows that he is a free man. But in his heart he does not believe it. He knows he is . . . a slave. If he buys his freedom from his master, that is different. But otherwise he believes that he belongs to his master—and so do his children.” [Saharan freeman, Timbuktu]

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

159

Introduction On the eve of independence in most of French West Africa, it was reported that slavery still existed and that slaves were still being bought and sold.1 The British and Foreign Anti-Slavery Society undertook to assess for itself this horrific legacy of colonialism. Timbuktu, once an obsession for European travelers who willingly gave their lives in the quest to “discover” it, became newly infamous as a place where anyone—even a European—could buy a slave. Robin Maugham’s Slaves of Timbuktu (1961) showed that it was also a place where Europeans could still buy freedom for Africans.2 Both concepts, buying a slave and buying freedom, were anathema to postwar Europe. Maugham expressed unease at his venture—purchasing the freedom of one slave in order to prove the continued traffic in so many others. But this was less because of any moral reflection on the inherent contradiction in the process than guilt that he was misleading the local commandant, who treated him well (1961, 204–205). Maugham’s aim was to shock his benefactors and those who would support the Society into “abolitionist action” with his stories of contemporary slavery, cruelty, and slave-marketing. But his achievement from an historian’s perspective was to reveal, through his various interviews with slaves, slave masters, and “freed” slaves, the complexity of the issues surrounding the whole question of “slave redemption.” The selection of quotations with which this chapter is introduced suggests a number of those issues, including some implicit contradictions. On the one hand, it seems that being “freed” did not necessarily mean being “free,” and that to whatever degree it did, it was dependent on freedom being obtained in a particular way. Both descriptions of (former) slave and (former) slave master seem to be assuming a common Muslim identity, the former indirectly with the reference to being a “slave in his heart,” the latter directly with the statement that he has “only sold a slave.” The former could realize “freedom” only if he purchased it himself; the latter could realize being a good Muslim only if he personally “freed” the slave. On the other hand, the three quotations attributed to local free/freedmen all raise the specter of the French as cultural and political interlopers, whose role as “abolitionists” is at best ambivalent. If when the French left, it was believed by slaves and masters alike that a master could (and would) retake his slaves, then clearly “law” had 1 Rapport des Nations Unis sur l’esclavage, 1950 (Archives Nationales du Sénégal, AOF. 2K15 174 Mauritanie). 2 The process of purchasing a slave is recounted in detail in Maugham (1961), pp. 201–204.

160

E. ANN MCDOUGALL

not supplanted “culture” in spite of the fact that slaves had been freed during the colonial experience. Finally, the initial statement by Maugham himself—“We have bought a slave. We have shown that the traffic still exists”—reminds us that individual “freedom” may not always have been the only item on the colonial agenda. The nations that emerged from France’s desert and desert-edge territories, Mauritania, Morocco, and Mali (where one finds Timbuktu), were what the French called sociétés esclavagistes—slave societies. In such societies, where slavery in some form has been integral to defining customary social relations, you cannot simply “move past” it—you must engage with slavery. But engagement also requires that you acknowledge the role of “culture” in shaping slavery’s context. One of our errors in establishing the parameters of the discussion here is to seek them principally in the American experience, and to seek them only in the nineteenth century. If we want to look at “slavery” and “redemption” in Africa, then it is in Africa that we will find our relevant contexts. In fact, the bulk of contemporary discussion around slavery and redemption focuses jointly on the West African countries of Mauritania and Sudan.3 From a western (largely American) perspective, both are regarded as countries dominated by white Arab Muslims in which black non-Arab non-Muslims are enslaved. Actually, Mauritania does not conform to this profile in two important ways: the black non-Arabs are also Muslim, and among those black non-Arab Muslims are many former owners of black slaves. These “details” are usually overlooked in the interest of painting the larger picture of religious persecution and racial exploitation.4 The “joint” profile also overlooks the important fact that each country is the creation of a different colonial power (Mauritania being French, Sudan British) and they therefore reflect two different historical realities. Those colonial histories, as was suggested at the outset, were critical in shaping colonial and immediate “post-colonial” slavery; consequently, contemporary slavery, however we come to understand it, is effectively a legacy of those earlier experiences (and perceptions 3 That said, Niger has begun to attract similar attention. While a May 2004 antislavery law making the practice of slavery punishable by thirty years in prison has gone largely unnoticed internationally, the controversy over the March 2005 “freeing ceremony” in a region bordering Mali is currently attracting considerable media interest. Niger may well soon compete with Mauritania and Sudan for notoriety. (For example, see: “Niger Govt to Free 7000 Slaves,” 4 March 2005, http://www.afrol.com/articles/15830; “Confusion over Niger Govt Stand on Slavery,” 7 March 2005, http://www.afrol.com/articles/ 15847 “Timidria” [the Nigerian antislavery society], http://www.antislavery.org/homepage/ antislavery/award/timidriabackground2004.htm.) 4 This discussion is expanded, with references, in McDougall, Brhane, and Ruf (2003), 67–88. Particularly egregious examples of these caricatures are posted to the “Coalition Against Slavery in Mauritania and Sudan,” http://members.aol.com/casmasalc/.

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

161

thereof) among Africans involved. Others in this collection are addressing the situation in Sudan, where “slave sales” in recent years have already provoked considerable international controversy. I would like to turn to the other “villain” in the region, Mauritania, where efforts of redemption societies like Christian Solidarity International have not yet manifested themselves in spite of ongoing reports of slavery and slave sales.5

Colonialism and Rachat: The Buying of Freedom The Islamic Republic of Mauritania was created by policy that reflected French concerns for French West Africa (AOF) more generally; it was also carved out of a Sahara that generated policies unique to the desert and encompassed part of southern Morocco and eastern/central Mali (home to Timbuktu’s Tuaregs). This ambiguous position was well reflected in the evolution of French policy regarding slaves and rachat— “buying back.” The first thing to make clear is that French concern to end slavery—or more specifically, to end it for African masters—was nothing new when Maugham encountered it in the 1950s. On the contrary, it echoed policies dating to the initial French conquest of West Africa in the 1880s and 1890s. The establishment of villages de libérté or “liberty villages” was envisaged to provide refuge for slaves seeking escape from masters; the villages were also intended to absorb slaves the French administration “taxed” (as they taxed all goods in kind) from caravans passing by their scattered frontier customs posts. Once registered as residents of these villages, these slaves were to be given “liberty.” However, the villages were provided with almost no resources (including land), and by dint of their location near railheads and administration posts, they served largely as labor reserves. Slaves who registered worked for a three-month period, during which time they could be reclaimed by masters. They could also be hired out, with arrangements made for part of their salary to be put towards their “price of liberation.” Initially, that price was some Fr 300, a price well above the cost of “slave replacement” and certainly well beyond what any wages could accommodate. (In the late 1890s, the price was reduced to Fr 200.) At the same time, masters could also claim up to Fr 1 per day of their slave’s wages (presumably, before reclaiming them); few slaves were actually paid more than that. 5 The national antislavery society working actively to bring slavery and slaves to the attention of international audiences is SOS Esclaves Mauritanie. An internet search reveals the extent to which it has been successful in disseminating information; its 2001–03 reports are available at http://sos.iabolish.com/report-2001-French.rtf; http://sos.iabolish.com/ SOS-Report-2003-French.rtf.

162

E. ANN MCDOUGALL

Often new arrivals worked on short-term jobs without pay, and “pay” was frequently nothing more than rations. Those who agreed to enlist in the French “African Army” (the tirailleurs) had only one month to work before being “freed” to the employ of the army (Bouche 1968, 79–112 passim, 146–147; Klein 1998, 84–88). In short, the notion of “liberty villages” seems literally to have been rooted in the belief that “work shall set you free”; in reality, the villages placed slaves in positions where they had various options that might—or just as often, might not—result in their actual liberty. Ultimately, one way or another, while satisfying French labor needs, they were working off the price of their own redemption. It is perhaps not insignificant that before the French banned the word from their reports, the villages were referred to by local administrators as “villages de captifs” (Bouche 1968, 156), the concept of “liberty” being as conspicuously absent from their descriptions of labor as from the slaves’ experience of reality. Even as this policy of rachat was being worked out in the 1890s by colonial administrators both in Paris and on the ground, it was being expanded by the activities of an unlikely partner—the Catholic missions. As early as 1885, Pope Leon XIII had recommended in a letter to Cardinal (later the “abolitionist archbishop”) Lavigerie, that missions not wait for government action to end slavery in Africa but rather “racheter” as many slaves as possible, as quickly as possible. Missions had already been doing that in East Africa (site of the famed “Arab trafficking in slaves” reported by David Livingstone), but perpetually complained that resources were too few to be effective; their activities had been limited to “picking up a few children” from passing slave caravans. Lavigerie initially argued that rachat was simply not practical to tackle the millions in slavery; he preferred the idea of creating military units to attack the merchants and slavers in the caravans themselves. What is most interesting in light of current discussions, however, is his argument that if the Church undertook a public campaign to buy back slaves, this would encourage slavers “to multiply their captures in order to satisfy the growing demand.” The military forces failed, however. And following Lavigerie’s death in the mid-1890s, the former director of one of the Saharan contingents took over the mission at Timbuktu. Despairing at the mission’s failure so far, he asked the Anti-Slavery Society of France for Fr 24,000 to establish the first mission-sponsored liberty villages. He received Fr 12,000 in 1987 (Bouche 1968, 177–180). By the turn of the century, religious and secular efforts to “buy freedom” for West African slaves through “liberty villages” were largely indistinguishable. Initially, most mission villages were peopled by systematically purchasing slaves; limited monetary resources also limited village growth. One report suggested that missions could only afford to pay

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

163

Fr 100 on average per rachat and that in some areas, traders refused to do business with them. Moreover, whether buying from local communities or slave merchants, this price could only purchase the old, the sick, and those ostracized for crimes like sorcery or intended as part of funerary sacrifices; often they were children—usually in poor health, already abandoned (Bouche 1968, 192, 255). “It was rare that missions were presented with well-constituted children: their masters knew that they could get much more profit elsewhere, anywhere from Fr 180, 200 to 300” (Bouche 1968, 194). If in principle the acquired freedom of any slave should have been equal to the freedom of any other in the battle to end slavery (and populate villages), what is clear from these complaints is that missionaries’ zeal for slave redemption was tied to their recognition that the Christianizing of Africa was rooted in rachat. The missionaries’ very raison d’être was reflected in these pitiful, purchased slaves; they were not yet ready to admit that these might not prove suitable pillars of faith (Bouche 1968, 194). Soon these villages, like those of the administration, began to attract refugees in search of food and sustenance during times of drought or famine. But like slaves who were excessively expensive, these refugees too were often turned away (Klein 1998, 117). Reports of these instances and ongoing complaints from the missions about inadequate resources gained an ear in France. The Anti-Slavery Society committed itself to fundraising for slave purchases. There were several schemes proposed in letters and advertisements in the society’s Journal. It was suggested that “liberty villages” be named after donors who contributed funds for purchasing slave residents; in 1896 one prominent cardinal wondered publicly if prospective donors might not be attracted by the possibility of naming newly purchased children. The White Father’s Bulletin similarly suggested that donors contributing at least the price of one slave (that is, Fr 100) should be permitted to give the baptismal name; to be permitted to name a whole village would be “even more glorious.” In the 1902 edition of the Anti-Slavery Society’s journal the priest of one of Paris’s most populous parishes proposed raising Fr 6000 among parishioners to create a new village that would carry the name of their patron saint, Saint Joseph. And an elderly woman proposed sending Fr 6000 to establish a village in the name of a dear, deceased friend; she expressed the hope to be able to do the same for herself, before she died (Bouche 1968, 184–185). It is clear that rachat, the idea of “slave redemption,” had as fully captured Catholic imagination as it had colonial policy; but it is also clear from the failure to follow through on most of these proposals that the goal of creating free Christian souls was no more capable of generating resources than the goal of creating “free” marketplace workers. “Liberty” was conspicuously absent from the “villages” to which its name

164

E. ANN MCDOUGALL

lent legitimacy. And this was not only true from the perspective of those in charge. In 1925, a League of Nations report condemned the villages de libérté, noting that “the natives consider this institution as an attempt by European authorities to steal their ‘serfs.’ . . . In the French Sudan they call them les villages des captifs du gouvernement” (Bouche 1968, 254). The fact that such villages were always established where labor was needed, that they would be broken up to disperse labor as needed, and that up to three months’ labor could be exploited from a slave who could still, then, be reclaimed by his or her, master was understood locally for what it was. Moreover, that masters could claim large parts of wages, that wages were in fact often no more then food rations (remembering that “traditional” slave masters had to provide clothing and bride price as well), and that the “price of self-purchased freedom” was exorbitant, reinforced the perception. A well-known scholar of French colonialism and slavery puts it this way: in the early days of colonial rule, the French army and French administrators were as much a part of the “savagery” as any others; moreover, “their African agents often exploited French power to accumulate [their own] wealth and slaves” (Klein 1998, 124–125)—this with the full knowledge and participation of the civilisateurs. Judgment passed on the policy during its final days was hardly less harsh. Not only did the inhabitants of the villages not gain any material advantage, they gained no “moral” prestige either (Bouche 1968, 156). “The ‘liberty’ given by these villages was, in fact, a slavery often worse than one claimed bound slaves. ‘Slaves of the Whites’ is how the indigenous people of the Sudan call inhabitants of the ‘liberty villages’; the Commandants of the cercles hardly consider them otherwise. They are, in fact [slaves] by the obligatory residency, the forced labor, the delivery of a [food] ration by way of salary or welfare. . . . People of the ‘liberty villages’ are not able to be accepted as equal to other ‘free natives’ because for them this ‘act of liberation’ was arbitrary and spoiled [not properly done]; nor are they able to force their acceptance [as ‘free men’] by means of their lifestyle or by their [nonexistent] class unity” (Bouche 1968, 156, 159). Mission villages and their inhabitants were no different. Missionaries dealt with slave merchants. One missionary reported a merchant eyeing the various cloths, beads, knives, and other goods in the mission’s small store and openly regretting that he had not more children to sell (Bouche 1968, 255)—clearly dispelling the myth that Africans believed the mission was a source of liberty. Moreover, children’s labor had become indispensable. As one report reflected, “these ‘children of the mission’— bought [rachatés] by the missionaries, working for them, brought to the village with the support of the [colonial] administration—when they try to run away, what else are they but ‘slaves of the Mission’? And so it is

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

165

understood by the local populations who honestly return the fugitives, exchanging them for goods with the Mission. . . . The villages de libérté (or the Christian villages formed by purchased slaves) permitted the missionaries to install themselves in a ‘slave milieu,’ to be accepted there, to endure” (Bouche 1968, 255). The missions sometimes sought to purchase the “freedom” of relatives of their converts or workers and they were known to put pressure on masters to set a modest ransom for these special cases. They also took in many “pawns”—though it was often unclear to local people whether they were giving refuge or actually purchasing them. Pawning was an important survival strategy in both the Sahara and the Sahel. “Pawns,” usually children, were used to help pay off debts, the intent being to redeem them when the debt and its interest was covered. They were also used in times of great difficulty, like the droughts and famines that hit these regions periodically. Children of working age would be exchanged for a sum of money; they would be assured food and lodging, and the rest of the family could use the money to buy themselves the same. The aim was for the family to repay the “loan” in better times and recoup the pawn. In reality, the interest levied on the loan often necessitated leaving the child indefinitely with the debtor—essentially as a very cheap slave. When missions took in or “bought” pawns, then, it was seen as exactly that: a means of obtaining very cheap labor. They were acting just as a wealthy merchant or powerful family might act in the village or clan. Indeed, as the missions “inserted themselves” in these sociétés esclavagistes, they were accepted largely because they took on such a role. While they may have convinced themselves that they were rescuing slaves, masters simply saw them as relieving their own burden in difficult times: increasingly, they would turn their slaves loose to seek refuge at the missions, or offer their slave children as pawns—then return to reclaim them later (Klein 1998, 117). Consequently, far from “ending slavery” by buying freedom, French policies and practices (broadly defined to include the secular and religious) reshaped slavery by buying more slaves. The impact was not so much felt through the “market,” as no form of rachat ever seems to have been at high enough prices to encourage expanding supply or even significantly raising profits. It was felt more at a social level, where French administrators and French missionaries created a role for themselves—or perhaps more accurately, had that role created for them, that was, for the most part, compatible with local cultural customs. Buying slaves was . . . buying slaves. Using slaves for labor was . . . using slave labor. Doing both was . . . becoming a master. Negotiating with masters was . . . acknowledging masters’ authority and responsibility. African or European, authority was authority. It really did not matter.

166

E. ANN MCDOUGALL

The Case of Mauritania: “Escaping Slaves Are Like runaway children . . .” By 1911, villages de libérté had officially all but disappeared from the AOF. It is ironic that almost two decades later, they began to reappear along the desert edge. At least four such villages were identified in 1927, another was created in an existing riverside agricultural center, and several “unofficial” river villages continued to receive desert refugees even after the official village apparently ceased to exist (Bouche 1968, 162–163). The source of these slaves was Mauritania—an exception to the general “rule” regarding slavery. While in the other colonies and territories the French administration declared all inhabitants free (or “potentially free” under some form of rachat), in Mauritania, masters were openly supported in their claims to recover escaped slaves. In 1905, one administrator had emphasized that the “nature and law of slavery [are] different” here, that total abolition would cause much political and social distress, that “mass liberation” would constitute a “revolution” (McDougall 1988, 362–363). More than two decades later, in the Adrar (the central desert region where the French had established an administrative and military center), a circular from the commandant warned that “nonjustified” slave escapes were to be regarded as an affront to the authority of the family head. They were equivalent to situations of runaway children in France; consequently, such refugees should be returned to their “families” (Bouche 1968, 162). As late as the mid-1940s, the administration continued to levy a fixed redemption payment for those still enslaved, for those who had been slow to realize their “new right to freedom” (Bouche 1968, 163). It is no wonder that in 1949-50, a United Nations commission found significant numbers of “servile peoples” in Mauritania, many of whom were unquestionably slaves. It also found French officials repeating the same excuses they had provided for nearly half a century—concern over “theft and vagabondage” by people with no resources, the “ruin of the economy”; this seeming anachronism in the mid-twentieth century needs to be understood against the backdrop of rachat in the region (McDougall 1988, 362–363). In the Mauritanian case, possibly because of their belated appearance, villages de libérté never played the important policy role they did elsewhere. The concept of rachat in this desert colony was rooted at the local level and used to deal with individual cases that, if records are to be believed, plagued some administrators for years. The cases in which administrators reluctantly became embroiled most often involved both women and their children in “marriage” situations, or slaves “inherited” through Islamic (customary) law (McDougall 1988, 370–379). The legacy of nineteenth-century practices in the AOF of divvying up captured slave

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

167

women and children between army officers and soldiers (many of whom were African) and local African notables and interpreters lived on in the twentieth-century Mauritanian practice of providing locally based tirailleurs with “slave” wives. Soon “marriages” that were little more than slave prostitution prompted the authorities to insist that the slave woman be purchased before the marriage. Then fears that these wives would in fact be re-enslaved when they returned to Senegal or Mali with their husbands led to a prohibition on taking Mauritanian wives out of the colony. Children of these marriages, on the other hand, were a different issue: tirailleurs could claim and take children as long as it could be proven they had been born within wedlock—otherwise, they were regarded as slaves belonging to the (former) master of the woman/wife (McDougall 1988, 370–371). These rather bizarre regulations reflected an interesting attempt by the French colonial authorities to respect several not necessarily compatible sets of values: patriarchy and marriage as defined by French law and culture; Islam and slavery as defined by Mauritanian law and custom; colonialism and abolition as defined by French policy and practice. The mix saw cases spanning ten or more years as family members fought over inherited slaves, husbands and masters fought over “wives,” and everyone fought over children. Women seldom won (unless they were also “masters”—and a significant number were), children were never asked. The most vulnerable became even more so thanks to rachat and abolition. Masters often accepted “payment” for female slaves: they called it bride price for a woman still enslaved, the French called it “purchase” of a woman now “free.” Meanwhile tirailleurs enjoyed (exploited) what were essentially concubines until they left Mauritania; if they were lucky they also got children out of the deal. Masters who exploited both Islamic law and French policy generally retained their female slaves in the end, having been compensated handsomely for their temporary absence. Local administrators found themselves mired in the middle, often for a very long time. That this situation still pertained in 1950 drew international attention. Indeed, this was the backdrop against which the British and Foreign Anti-Slavery Society launched several investigations of its own, one being the excursion by Robin Maugham to Timbuktu that we encountered at the outset of this chapter. Another had taken place just two years earlier, this one to Morocco and Mauritania. Or, more precisely, to Morocco— permission was never granted for the investigators to enter Mauritania. Consequently, they sought sources in the southern Moroccan town of Goulimine that could shed light on slavery to the south (Fox-Pitt 1957, 19–20, 25). And they found them. Commander T. Fox-Pitt published findings in the Anti-Slavery Reporter in 1957 that included two “testimonies” of former slaves (26, 27). These are clearly problematic from a

168

E. ANN MCDOUGALL

political point of view: they are indictments of French human rights policies at a time when Britain (among others) wanted to see an end to French control of the region; and they come from Goulimine, home to the recently formed Free Liberation Army whose goal was to overthrow the French protectorate in Morocco, as well as to help “absorb” Mauritania into the newly independent nation. The interviews were arranged by the mayor of Goulimine, whose family was active both in the Liberation Army and in maintaining influence in Mauritania (24, 25). It suited everyone’s purpose to recount stories of slavery and cruelty at the hands of the French in Mauritania, in contrast to the “liberated” environment of Morocco. The French were arguing that “residual pockets” of slavery continued to survive in Morocco, and if the protectorate were lost, fullfledged slavery would immediately reappear [19, 20].) That said, the testimonies are revealing to us as we search for an understanding of the intersection between being “freed” and being “free” in the Mauritanian (and by extension, southern Moroccan) context following fifty years or so of colonial rule (Fox-Pitt 1957, 26, 27; Italics below are my emphasis). Abide El-Barka ben Mbarak [bin M’barak]: ‘I was the slave of Mir [Emir] Hamed ould Ahmed Aida who had many other slaves. . . . He sold some of them. He sold the children at any age but he also sold the grown-up slaves. My master lived in an encampment but he wanted to live in the town of Atar and asked the French if he could keep his slaves in the town but the French took us slaves away by force from him and put us in prison. I expect they bought us as we worked for them. I think they paid about 20,000 Fr for us. I escaped from the French prison and was flogged with a cravache whip . . . then I went back to my master. The French only allow slaves in the desert not in the towns. It was when I was back with my master that the Caid Dahmane who is now Caid of Goulimine bought me to set me at liberty. . . . He paid the money to my master. I have two sons still in slavery with my old master. . . . The French always stop slaves escaping to Goulimine but some escape in spite of this. There are others living here in Goulimine who have escaped but they are away now. I live very well here in Goulimine with plenty to eat and good clothes. I am a mason and work at that trade, but am free now in Morocco’ (Fox-Pitt 1957, 26, 27). M’Barjarek ben [bin] Bilal: My master in Mauritania was called Hamidi [Hamody] and I was his slave. He had more than 10 other slaves. There were both men and women. He had the right to sell any of us but he mostly sold the children. He would sell them at any age. Most often they were about ten years old when they were sold. The French knew there were slaves belonging to my master. When I first

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

169

escaped I went to the French at their office in the town of Atar. They held me until my master arrived and then they handed me back to him. I was beaten with a camel whip called a cravache. My master made a hole in the lobe of my ear. . . . He put a thong through this and led me with my hands tied behind my back walking beside him as he rode a camel. This is how I was taken back to his camp. I wore nothing but rags when I was a slave. I looked after the camels. Later I made a successful escape from slavery. I now live free and well in Goulimine. I have food and clothes and work to do and lead a prosperous life. My father escaped from slavery before me and my mother is dead, so I was able to leave without leaving anyone behind. I came alone.

Rachat from a Slave’s Perspective Ould Ahmed Aida’s slave draws our attention to the now familiar but still ambiguous policy of the colonial rulers “buying the freedom of slaves.” On the one hand, he recounted that he (and the other slaves) were “forcibly” removed from their master when the latter moved into town, on the other, that this same master was paid the price of Fr 20,000. First they were “freed,” and then they were “imprisoned” and forced to work. Then, they were beaten and our informant ran away, back to his master. Contradictory though these descriptions seem, they are not inconsistent with the historical experience of rachat as we have come to understand it—or more importantly, as the former slave himself understood it. He was “purchased” by the French and told he was free. The administration’s policy of having slaves “work off” their redemption (thereby keeping them from that feared “theft and vagabondage” which lack of resources would induce, while simultaneously getting work done) was “prison” and “forced labor” to bin M’barak. So, if he was not going to be “free” in any meaningful sense, he would rather return to his master— whom he clearly continued to think of as “his master”—and be a slave. The fact that he did not comment on a return to slavery per se suggests that at best he had seen himself as “freed” by the French, not as “free.” He spoke of returning to his master as returning from his “new master,” the French, to his “old one,” Emir ould Ahmed Aida. It is consistent with a belief on his part—as well as his master’s—that he was still a slave; consequently, it was natural to then speak of Caid Dahmane buying him (for the second time), “to set him at liberty.” Caid Dahmane ould Beyrouk ould Abidine, mayor of Goulimine, generously purchased bin M’barak from his master for Fr 50,000 (according to his own testimony, also recorded by Fox-Pitt [1957, 26]). This of course is in stark contrast to the rachat (of Fr 20,000) reported by bin

170

E. ANN MCDOUGALL

M’barak, above, as given by the French for more than one slave. He was from the Beyrouk family, a family famous for trans-Saharan trade who actively engaged in slave trading between the eighteenth and the early twentieth century. Nineteenth-century European travelers’ accounts probably exaggerate their holdings; nevertheless, it is clear the Beyrouk family owned more slaves than most others in the region, and not infrequently engaged in trafficking in them (McDougall and Mohamed 1999). Even as Caid Dahmane, mayor of Goulimine, spoke to Fox-Pitt, his daughter was the widow of a very wealthy Saharan merchant, Mohamed Barka, who had both imported slaves from Timbuktu earlier in the century and purchased them locally. Elsewhere, I have chronicled the history of one of them (McDougall 1998), a woman named Fatma (also called “Faytma” by her master), who was to have been “freed” rather than inherited on the death of her master in 1929. However, Dahmane’s daughter (Barka’s second wife) decided otherwise and kept Fatma as a slave for several years thereafter. It is unclear exactly when she was finally freed; the family did not want to provide her with papers—“what would anyone want of you?,” understood as “why would you need to prove your liberty?” (288–289, quotation p. 310). Nevertheless, even after her “freedom” was announced and she undertook work as a bread baker and seller traveling to different markets, she remained close to the family, dependent upon it in many ways (protection of her property, assistance with her son’s marriage), and answerable to it as long as there were family members alive to call upon her (287–289). Her story also involves a local slave who became a well digger for the administration, whom she married in the early 1930s, and another named Messoud, with whom she worked in Barka’s household. Although Fatma remembered Messoud as coming from south of the Sahara, another family member confirmed that he had been a “freed-slave” of Barka’s deceased business partner whom Barka had “inherited” along with other property (288, 307). The world of Fatma and of Goulimine in the 1940s and 1950s was one of slaves and freed slaves. It was not always clear which was which. That much Fatma’s story makes clear. According to bin M’Barak’s testimony, he “live[d] very well . . . in Goulimine with plenty to eat and good clothes . . . and work [as a mason] . . . [he was] free in Morocco.” Similarly (one might remark, too similarly), the other informant, bin Bilal, also recounted that he “live[d] free and well in Goulimine . . . [with] food and clothes and work to do.” Apart from the suspicion raised by this parallelism in conceptualizing and expressing what being “free” in Goulimine meant, with particular reference to food, clothes, and work, we might well question how it was that they were “free” while others in Dahmane’s own family were, at best, “freed.” One might also note that in a footnote to the report, Fox-Pitt corrected the name given in the testimony by

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

171

“Abide El Barka . . .” to “El Barka.” He explained that abide was the plural of abd, meaning slave “and not correctly part of a name” (1957, 27). Other evidence from Goulimine would suggest that the meaning, if not the name itself, did indeed apply; and it is not impossible that he was literally referred to in this way. “Slave of el-Barka,” which was what the name meant, could well pertain—he was ‘bin M’Barak, son of M’Barak, among the slaves of Barka. This raises a rather interesting possibility: Fatma’s master and Dahmane’s son-in-law was Mohamed Barka. Just a year before her death, Fatma received her first ever “national identity card.” She had given her legal name as “Faytma Barka” (McDougall 1998, 305). Had bin M’Barak also been “a slave of Barka” following on his arrival in Goulimine? Was his “free” life as a mason akin to Fatma’s life as a bread seller (and also, at one stage, a sardine-canning factory worker [288, 289])? If this was the case, there is no question that “free” meant “freed”—haratin in the Mauritanian sense (see below), even if he himself did not use the term. M’Barjarek bin Bilal (“son of Bilal,” the black slave of the prophet Mohamed, affirming his inherited slave status) recounted an equally revealing, if more suspect, biography. His account of being punished and returned to his master as a fugitive slave is consistent with early French policy; it is perhaps less clear that he would have been so readily returned to his master in the 1950s. But it is possible. However, as an escaped slave arriving in Goulimine, supposedly alone on a camel (which he would have had to have stolen from his master’s herd), he would have had no choice but to seek out a patron from one of the leading families. As it happens, we know his master “Hamidi” had commercial dealings with the merchants of Goulimine at this time, including the Beyrouk (McDougall 1988, 381).6 There is no way that one of his known slaves would have been permitted to claim both his own freedom and the value of the camel without someone, quite probably the Beyrouk themselves, paying a price. A hartani experience comparable to that of Fatma, Messoud, and (probably) bin M’Barak was more likely to have provided the “food, clothing, and work” he enjoyed. Food, clothing, work—the essence of both slavery and freed-slavery. And also, apparently “freedom.” 6 The article says “Morocco”; my oral interview work subsequently in Goulimine (several occasions between 1993 and 1996), with “Hamidi’s” son Mohamed Said ould Hamody (2000, Washington, DC) and most recently with several haratine related to Hamody’s family confirms the Goulimine-Beyrouk connection. Hamody frequently sent camels north with his haratine in order to bring back Moroccan merchandise. Two prominent Beyruk cousins also established themselves in Atar. Relations between Hamody and the Beyruk family included marital and “milk-kin”; there were also relations between their respective slaves. Both sets of relations are being explored further as part of my current work on the Hamody family.

172

E. ANN MCDOUGALL

The ambiguities of his status aside, the most interesting aspect of his account for us lies in what he reveals of his former Mauritanian master “Hamidi,”—a rather famous entrepreneur named Hamody who profited from the presence of the colonial army to establish a business supplying the army with meat and hides. He successfully translated this economic prosperity into property—land, palm trees, buildings and slaveholdings (sizable ones from all accounts) (McDougall 1988, 380–382). Bin Bilal’s reference to Hamody having ‘more than 10 other slaves’ was an understatement. That they were both men and women is absolutely true—indeed, Hamody had a child by one of the latter, Medeym, the same year the Anti-Slavery Reporter article appeared (382).7 What our informant omits to tell us, however, is that Hamody was himself the son of a former slave-concubine who was “freed” by her Goulimine master upon giving birth.8 She subsequently married another “freed” slave, a hartani. As the son of two “freed slaves” Hamody was himself hartani, a “freed slave”— not, it is important to note, “free” (380). He chose to invest much of his wealth in slaves. The French supported him in this choice, as long as he continued to contribute to the development of the region and serve as a local representative (chef ) of his clan (381). Giving birth to Hamody’s son would have brought Medeym her freedom, had the child been recognized by his father. As it happens, the relationship was not made public until after Hamody’s death and the birth of a second son; technically Medeym remained his slave. However, in fact Hamody provided her and her two children with a house. Later he married her to one of his haratine, Bilal. The children remained with Medeym and Bilal until Hamody’s death, at which time the latter’s paternity was revealed. Hamody’s family raised one of the sons, Mohamed, while the other, Isselmou, choose to stay with his mother.9 Bin Bilal’s presentation of Hamody as “master,” without qualification of Hamody’s haratin status, is significant. To be master did not necessitate being “free,” only being “freed.” Hamody’s experience, as I have elsewhere described and interpreted it, suggests that the right—the 7 About two years later he had a second son with the same slave concubine, according to interviews with Hamody’s children, the ex-wife of one of his sons, Selka, and Medeym’s son “Isselmou,” December 2004–January 2005, Nouakchott and Atar. I am currently working on a history of the family, based largely on these recent interviews. None of the material has yet been published, nor is the research complete. However, what I have done to date not only expands upon the information in McDougall (1998), “Topsy-Turvy World,” but corrects certain points, perhaps most important of which is that Hamody never married Medeym. 8 She stayed with him until his death, however, as was common. Only then did she “activate” her freedom by leaving. 9 Interview (2000) with Mohamed Said ould Hamody, Washington, DC; interviews with Hamody’s daughter Swaifya, his ex-daughter-in-law, Selka, and his son “Isselmou,” December 2004–January 2005.

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

173

ability—to buy slaves affirmed the difference between being freed and being slave (1988, 382–384). Bin Bilal’s testimony confirms this to have been the case, at least from his perspective. If ownership of slaves was not the prerogative of the free, freedom was not the antithesis of slavery. Not, at least, in Mauritania. In short, the cast of characters in these Goulimine vignettes that supposedly vilify Mauritania and the French while acknowledging the progress and humanitarianism of Goulimine— where, according to Caid Dahmane and bin Bilal, so many Mauritanian slaves sought refuge from the cravache—generates nothing but continued questions, not about what slavery was, but about what freedom was not. These testimonies, the odd anecdote (often unattributed), some recent interviews with former slaves—they are not enough to give us the extensive “slave perspective” we truly need here. However, they are enough from which to argue that the issues are complex ones, varying from person to person, from region to region (and, I would argue, from men to women). One anecdote (repeated in a 1912 publication to explain why the village policy failed) cited an anonymous “liberated slave” from one of the villages as responding to the question “how are conditions?” with the complaint that “they had no slaves to help them in the fields” (Bouche 1968, 159). The acceptance of “slavery” as a social and economic reality continued even among those who had been enslaved. This anecdote from the era of the villages de libérté may be apocryphal or perhaps exceptional—but Hamody was neither. Not only did slaves who had been freed accept the concept of slavery, they embraced using slave labor to better their own position, and owning slave “people” to reinforce their own social distance from slavery. Because the issue was not “freedom” versus “slavery” but rather “freed” versus “non-freed,” the acquisition of rights to slave labor or slave ownership affirmed one’s “non-slave” status. There were other issues as well, as Fatma’s story in particular illuminates. Fatma was Muslim. Her telling of her tale emphasized several details about how important her slave status was to her Islamic one. Among references to the pious nature of her master—in a sense, a kind of “mirror” to her own—was the insistence that she was not his daughter (although she’d arrived with him in Goulimine as a child) but that he had bought her. She was his slave. This statement was repeated at least twice in a context that to me seemed not to warrant this specificity. On the contrary, if she could have been mistaken for his daughter (she was Malian, but so was her master’s wife at the time), would she not have wanted that to be the case (McDougall 1998, 300–301)? It was only after her death, when I found out that she had been Mohamed Barka’s concubine, that I understood: in Islam, concubinage is completely acceptable. Honourable, even. However, only slaves can be concubines. It is forbidden to take free

174

E. ANN MCDOUGALL

women as concubines. Her whole “legitimacy” depended upon her being very publicly recognized as a slave.10 And her continued relationship to the family similarly was rooted in the recognized responsibility of that family to her as a former slave. While the status of haratin seems to vary within the broader Muslim world, the notion that a good Muslim master will see to a former slave’s well-being, and that a good Muslim slave will continue to respect a former master’s family, is generally shared through common belief in the Qu’ran. Each remains, in fact, a “resource” for the other, as well as a touchstone for affirming “good,” pious behaviour (McDougall 1998, 434–439). Fatma may be exceptional, but she is not the exception. Belief in Islam seems to have shaped at least some slaves’ perception of “colonial freedom” from very early on. A French scholar who studied the villages de libérté extensively, reported the following exchange in 1884 between a slave in Saint Louis, Senegal, looking for work and a French administrator. The slave indicated he was looking for work so that he could earn the price of his freedom. The administrator told him he had only to request a certificate of freedom from the French authorities and it would be done. The slave refused, apparently saying that he could not just leave his master without reimbursement for what he had cost him. If he behaved “properly” (which was to say, buying his freedom), God would forgive him all his sins, past and present, and allow him to benefit fully from all his prayers and good deeds. He added that he knew his master would not ask for more than he was capable of paying, so “I will buy my freedom or I will die a slave” (Bouche 1968, 158–159). This is not inconsistent with many stories that have emerged in oral interviews with elderly former slaves throughout contemporary Senegal, Mali, and Guinea, stories of continuing to pay zakat (Islamic tax) to former masters, building community mosques, continuing to accept (albeit often hidden) benefits available only to slaves who retained “socially subservient” relations with former masters.11 One story in particular has relevance for us. This is the tale of an old male slave who, informed that he no longer needed to purchase his freedom in modern Senegal, insisted on doing so anyway—and then leaving immediately on pilgrimage to Mecca. This desire to make the hajj after being freed was apparently very widespread (Klein 1998, 246–247). While it might be suggested that this is nothing more than an imitation of “being noble,” of “speaking the master’s language” (Klein 1998, 247), I would argue that it is a much stronger statement indeed. I would go so far as to say her honor and very identity were thus defined. McDougall (2000): reference is to several works by Martin Klein in which he recounts different stories of these “Islamic symbols.” My interpretation of their significance in the lives of former slaves differs from his, however. 10 11

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

175

The fact here is that it was not the purchase price that brought true freedom for the slave, it was the pilgrimage: one could redeem oneself with money; legal freedom did not necessitate the hajj. But one could not make the hajj without being legally free. It was in this act that Islam and Freedom intersected, and one came to define the other. To return full circle to Robin Maugham’s experience in Timbuktu in the late 1950s and to the quotations with which we began this exploration, it was clear that Islam, “faith,” shaped masters’ and slaves’ views of each other and of their world. The Tuareg master who sold his slave apparently had second thoughts about the relative gain he had made: he was no longer certain that the money he had received was worth the blessing he had lost by selling rather than freeing the man.12 From the slave’s perspective, the explanation given Maugham as to why a slave would not just ask for his freedom, namely that he would know he was a slave in his heart unless and until he had properly purchased his own freedom, was another way of expressing the same Islamic beliefs as those held by the slave from Saint Louis. Similarly, another slave Maugham met in Timbuktu had been “released” many years before by his master to fend for himself. Now growing watermelons for a living, he still paid his master (and he called him that) an annual “head tax”; in good years he gave him an extra “present.” “I still belong to my master” the man said, “and so do my four children. I know that the law today prevents my master from forcing my sons to go back to him. But if my master sent for them they would feel bound to go, for they know they are his slaves” (Maugham 1961, 169). Maugham also met a former slave who recounted a situation in which it was his master who reaffirmed “traditional” Muslim beliefs and custom, albeit in a less than benevolent fashion. The slave told Maugham that back in the 1930s, he had left his master and joined his runaway slave parents in Timbuktu. There, with his father’s help, he had received the support of the commandant to remain, ostensibly, “free.” His master had soon arrived to take him back, and the commandant had intervened, telling his master that under French law he was free. His master had sought time alone with him and essentially threatened him with continual harassment and possible re-enslavement if and when the French ever left—unless, that is, he bought his own liberty according to a price he, the master, would fix, in which case he would be “free” for good. “So I saved every franc I could. And at last 12 This point was made repeatedly by people I interviewed from Hamody’s immediate and “extended” (hartani) family: to sell your slaves was “honte”—shame. One would sell almost anything—animals, houses, land—before selling a slave. The fact that so many were sold to Hamody in the 1940s in particular, and that people sought him out to buy their slaves, underscores the economic penury experienced by the region at this time.

176

E. ANN MCDOUGALL

I had saved twenty thousand francs which together with two donkeys made up the price he wanted. Then we went to the mosque of Dkinguereber, and I paid the ransom in front of the Imam and in front of witnesses, and I received my liberty” (1961, 166, my emphasis). This religious ceremony gave legitimacy to that liberty for all involved—slave, master, community—which French colonial law never could. One is led to wonder, then, just how “real” the freedom was that Maugham ostensibly bought for “his” slave, or more precisely, for whom that freedom was real. I do not interpret this as the oft-cited “submission to Islam” per se, the easy explanation that Islam makes everyone a slave to God and makes slaves therefore helpless. I see these accounts, each in its own way, resonating with the one cited earlier by the slave in Saint Louis: as a Muslim in a Muslim society, there are ways in which one can cease to be a slave and become a “freed” slave. One of them is to be “freed” by your master for any number of reasons prescribed in the Qu’ran. One of them is to purchase your own freedom with the agreement of your master. One of them is not to have someone else declare you “free.” One of them is not to have someone else purchase your freedom. Neither of the latter two options has social or moral currency. Just as indigenous West Africans said of the people of the villages de libérté nearly a century ago: “[they] are not able to be accepted as equal to other ‘free natives’ because for them this ‘act of liberation’ was arbitrary and spoiled.”13 Conclusion This essay—this exploration into history, into colonial conceptions of “freedom” and the process of “freeing,” into African perceptions of both, into African realities as slaves and freed slaves—set out to explain 13 French archival sources from the Tagant region of central Mauritania c.1948–50 make this point as well: “Liberated by the maure (Moor), conforming to “Moorish” custom, the slave (captive) becomes freed, not rupturing what has constituted up until now his normal life. It is thus that the interesting class of haratines [sic] forms itself—modest, frugal, hardworking. Freed by us, to the contrary, the slave contributes nothing of good, probably because he leaves his traditional life.” Clearly this observation is speaking to a different issue than that which concerns us here—that is, to the French concern to keep “good order” and produce “hardworking laborers”; it is also somewhat tainted as an observation precisely because of this not-so-hidden agenda. But that said, there is validity in the administrator’s observation of different understandings of “freeing” and in his recognition that a culturally “freed slave” still had a role in society, while a “free slave,” according to French law, did not. (V. Balesi, 3 August 1950, citing report dated 27 April 1948 Tagant. B1 Esclavage, Rapport des Nations Unis sur l’esclavage, 1950, Archives Nationales, République Islamique de Mauritanie).

THE PRACTICE OF RACHAT IN FRENCH WEST AFRICA

177

three central observations, all articulated in the opening quotations. To be freed is not necessarily to be free, to speak of cultural context is to speak of historical experience, to see Islam only as the property of the master is to miss the essence of being slave and Muslim. Or perhaps I hoped to do slightly more. Perhaps I hoped that echoes might resonate around the morality of “buying slaves,” whatever the end purpose, however Christian the process. The idea that one culture’s vision of what slavery is, and who slaves are, can effectively illuminate another culture’s history; the claim that there is “nothing political” in the act of chronicling, let along “ending,” slavery; the naivety we share in our belief that not only do we know best, we can do no harm—to all of the above, history has a lot to say, if we take the time to listen. Bibliography Books and Articles Bouche, Denise (1968), Les Villages de libérté en Afrique noire française 1887–1910. Paris: Mouton & Co. and Ecole Pratique des Hautes Etudes. Fox-Pitt, Com. T. (1957), “Report of an Investigation in Morocco in October and November, 1956,” The Anti-Slavery Reporter and Aborigines’ Friend 11(1): 19–27. Klein, Martin (1998), Slavery and Colonial Rule in French West Africa. Cambridge: Cambridge University Press. Maugham, Robin (1961), Slaves of Timbuktu. London: Longmans. Green and Co. Inc. McDougall, E. Ann (1988), “Topsy-Turvy World: Slaves and Freed Slaves in the Mauritanian Adrar, 1910–1950,” in R. Roberts and S. Miers (eds.), The Ending of Slavery in Africa (Madison: University of Wisconsin Press), pp. 362–388. McDougall, E. Ann (1998), “Islam: an overview,” in Paul Finkleman and Joseph C. Miller (eds.), Encyclopedia of World Slavery (New York: Macmillan Reference), vol. 1, pp. 434–439. McDougall, E. Ann (2000), “‘A Quest for Honour’: Islam, Slavery and the Contributions of Martin Klein to African History,” in Richard Roberts and Philip Zachernuk, “On Slavery and Islam in African History: A tribute to Martin Klein,” special edition of the Canadian Journal of African Studies 34(3): 546–564. McDougall, E. Ann, Mesky Brhane, and Urs Peter Ruf (2003), “Legacy of Slavery; Promise of Democracy: Mauritania in the 21st Century,” in Malinda Smith (ed.), Globalizing Africa (New Jersey: Africa World Press), pp. 67–88. McDougall, E. Ann, and Mohamed Hassan Mohamed (1999), “Les Maisons Beyrouk et Illigh: Construction d’ identité au Maroc de Sud, 16éme a 18éme siécles,” paper presented at the second Table Ronde de Goulimine, conference jointly sponsored Centre National de Recherche Scientifique, Paris and Institut de Recherche Scientifique, Rabat, Morocco, February 1999 (unpublished).

178

E. ANN MCDOUGALL

Archives Archives Nationales, République Islamique de Mauritanie, Rapport des Nations Unis sur l’esclavage, 1950. B1 Esclavage (Tagant). Archives Nationales du Sénégal, AOF, Rapport des Nations Unis sur l’esclavage, 1950. 2K15 174 Mauritanie.

Oral Interviews Mohamed Said ould Hamody (Hamody’s fourth son): November/December 2000, Washington, D.C.: December 2004/January 2005, Nouakchott RIM. Swaifya mint Hamody [Hamody’s first daughter]: August 1984, Atar RIM; December 2004, Nouakchott RIM. Mohamed Mahmoud ould Hamody [Hamody’s first son]: December 2004, Nouakchott RIM. Mohammed el Haiba ould Hamody [Hamody’s second son]: December 2004, Nouakchott RIM. Ahmed Salem [“Isselmou”] ould Hamody [Hamody’s second son with Medeym]: December 2004/January 2005, Atar RIM. Selka mint M’hamid ould Ismail [Hamody’s daughter-in-law with his second son, Mohammed el-Haiba]: January 2005, Atar RIM. Sidi Mohamed ould Dedi [civil servant, Atar]: January 2005, Atar RIM. Bilal ould Mbarak [hartani of Hamody]: December 2004, Atar RIM.

Web Sites “Coalition Against Slavery in Mauritania and Sudan,” http://members.aol.com/ casmasalc/. “Confusion over Niger Govt Stand on Slavery,” 7 March 2005, http://www.afrol .com/articles/15847. “Niger Govt to Free 7000 Slaves,” 4 March 2005, http://www.afrol.com/ articles/15830. SOS Esclaves Mauritania Reports: 2001, http://sos.iabolish.com/report-2001French.rtf; 2002–03, http://sos.iabolish.com/SOS-Report-2003-French.rtf. “Timidria” [the Nigerian antislavery society], http://www.antislavery.org/ homepage/antislavery/award/timidriabackground2004.htm.

Chapter Epigraphs Chapter epigraphs are taken from Maugham 1961, pp. 204, 204, 171, and 164 respectively.

Chapter Eight The End of Serfdom in Russia—Lessons for Sudan? LISA D. COOK

The literature on slavery in the Americas is often cited when the current problem of slavery in Sudan is considered. But in addition to looking to American slavery as a benchmark, it would be equally instructive to examine the less-compared serf-emancipation schemes developed and executed by tsarist Russia. Emancipation and redemption schemes have been debated and, with some controversy and mixed success, implemented in Sudan. Elaboration of the factors influencing design of an abolition program may shed light on alternatives available to those considering emancipation in Sudan.

The Mechanics of Serfdom in Russia Serfdom comprised a number of forms of limited (by law or by custom) ownership of human beings or legal claims to varying parts of their wealth, income, or services.1 A serf and the serfowner’s land were typically an asset bundle, and title to the land included title to the serf. Data on the evolution of serfdom, on serfs during the eighteenth and nineteenth centuries, and on serfs and serfowners on the eve of emancipation reported in tables 8.1, 8.2, and 8.3 may be useful in understanding the timing, scale, and scope of serfdom in Russia. Blum’s (1961) estimates of the number and proportion of male serfs in the Russian population are given in table 8.2. While most series suggest a decline in the proportion of serfs between 1800 and 1861, estimates vary. The estimates of Hoch and Augustine (2001), for example, are generally lower than Blum’s. They estimate that 50.1 percent of the male population was enserfed in 1811, and that this fell to 36.5 percent by 1857. 1 See Kahan (1973) for a fuller discussion of the definition of serfdom and Troinitskii (1861) for a summary of the rights and responsibilities of serfs and serfowners.

182

LISA D. COOK

TABLE 8.1 Chronology of Russian Serfdom, 1762–1907 1762

1766 1775 1785 1790 1797 1801 1803 1804 1807 1816–19 1825

1829 1837 1842 1847 1849 1855 1856

1857

Catherine the Great succeeds to the throne Secularization of the estates and peasants of the Russian Orthodox Church Catherine’s Instruction to the Legislative Commission alludes to reform of serfdom Provincial reform, including provisions to protect serfs from cruel and exploitative nobles “Charter to the Nobility,” codifying nobles’ rights and privileges Paul succeeds to the throne Labor services banned on Sundays, 3 days of work each on noble and peasant land Alexander I succeeds to the throne Decree on “Free Agriculturalists” permits nobles to free their serfs and sell them land Regulation by law of relations between nobles and serfs in Estonia and Livonia Start of abolition of serfdom in Prussia Landless abolition of serfdom in Estonia, Kurland, and Livonia Nicholas I succeeds to the throne Failed “Decembrist” revolt by army officers to overthrow autocracy and end serfdom Start of Perovskii’s reforms of appanage (imperial) peasantry Start of Kiselev’s reforms of the state peasantry Decree on “Obligated Peasants” permits nobles to regulate their relations with serfs “Auctions” Decree permits serfs to buy their land, limited freedom; effectively repealed in 1849 Decree permits peasants in Livonia to buy land from nobles Alexander II succeeds to the throne Alexander II issues directive to Minister of Internal Affairs to prepare proposals to end serfdom Grand Duchess Elena Paviovna presents plan to free serfs on her estate of Kariovka to Alexander II Decree permits peasants in Estonia to buy land from nobles at prices set by nobles Alexander II sets up “Secret Committee on the Peasant Question,” which proposes gradual reform Alexander II’s Rescript to Governor-General of Lithuania with principles for abolition of serfdom Provincial nobles elect committees to prepare plans for reform based on Lithuanian rescript Continued

THE END OF SERFDOM IN RUSSIA

183

TABLE 8.1 (Continued) 1858

1859 1861

1863

1866 1881

1883–87 1886 1894 1905

1906 1907

Main Committee on the Peasant Question replaces Secret Committee; prepares first abolition program based on the Lithuanian rescript but does not initially guarantee freed serfs land Wave of disturbances in Estonia after 1856 decree (above) published Alexander II tours provinces Alexander II approves creation of the Editing Commissions to draft statutes abolishing serfdom Alexander II signs into law the Proclamation and Statutes abolishing serfdom Massacres of peasants protesting terms of reform in Kandeevka, Penza, Bezdna, Kazan provinces Peace mediators begin to arrive in villages to oversee implementation of reform Formal start of “Temporary Obligation” Polish nationalist revolt, terms of abolition changed by Russian government in favor of peasants Reform of appanage peasants across Russian Empire; could redeem land after 1865 Reform of state peasants; transferred to equivalent of “temporary obligation” Alexander III succeeds to the throne Transfer to redemption required for all peasants still in “temporary obligation” from 1883; redemption payments and compensation to nobles lowered Abolition of poll tax in European provinces of the Russian Empire State peasants permitted to redeem their land Nicholas II succeeds to the throne Outbreak of Revolution Announcement of end of redemption payments, effective January 1, 1907 Peasants granted freedom of movement; peasant households permitted to leave village communes End of Redemption Operation

Source: Moon (2001). Note: Author’s emphasis.

In 1859 and throughout the history of serfdom in Russia generally, pomeshchik serfs dominated the serf population and, among them, peasants assigned to settled lands constituted the majority, or 93 percent, of pomeshchik serfs.2 These peasants were required to work three days a week for the pomeshchiki in order to fulfill their principal duty, which A pomeshchik is a noble landowner. Only hereditary gentry and pomeshchiki could own serfs. This exposition relies heavily on Troinitskii (1861). 2

184

LISA D. COOK

TABLE 8.2 The Serf Population in Russia, Males, Selected Years, 1724–1859 Year 1724 1743–45 1782–83 1796 1811 1817 1835 1851 1858–59

Male Serfs (millions)

Total Male Population (millions)

Male Serfs/Total Population share

3.2 3.4 6.7 8.7 10.4 9.8 11.4 10.7 10.7

5.8 6.6 7.4 12.8 17.9 17.0 21.8 22.3 24.0

55.0 51.5 52.0 49.0 58.0 57.5 52.0 48.0 44.5

Source: Blum (1961). Note: Years selected coincide with census years, and data are derived from census data.

was to cultivate pomeshchik land. In turn, the pomeshchik was obliged to provide peasants with a sufficient amount of land to sustain themselves and their families. Peasants met their obligations to the pomeshchik either by personal work, barshchina, or by direct payment in cash or in kind, obrok. In the 1760s, the annual average cash payment per serf was two rubles; in the 1790s, five rubles (or 3.12 silver rubles); and in the 1840s, 7.5 silver rubles.3 Table 8.4 reports data on the distribution of serfs on obrok and barshchina estates. The share of obrok serfs was higher in nonagricultural areas and lower in agricultural areas. State taxes were paid directly by peasants. State and seigniorial obligations from 1700 to 1799 are summarized in table 8.5. In real terms, the value of serfs’ obligations to the state fell by two-thirds over the period, while the value of their seigniorial obligations roughly doubled over the period. Most dramatically, grain prices simultaneously increased by an order of magnitude. The remaining seven percent of pomeshchik serfs were household serfs, who were employed by the proprietor for work or service but were not allotted land.4 Also unlike serf peasants, household serfs worked continuously, rather than three days a week, and their proprietors paid for food, 3 See Blum (1961, 450–451). Cash, rather than in-kind, payments to landowners became more prevalent as money became more widely adopted as the medium of exchange in Russia. It is estimated that there was little change in the real value of the obrok cash payment after the period covered in table 8.5, since 3.12 silver rubles in the 1790s would purchase 1.25 chetverts, or 7.4 bushels, of rye, and 7.5 rubles would purchase 1.5 chetverts, or 8.9 bushels, of rye in the 1840s. Though not exactly comparable, the Blum and Mironov nominal and real estimates differ slightly with respect to the nominal value of obrok payments. 4 These serfs appear in the entry for “landless gentry” in panel (B) of table 5.3.

185

THE END OF SERFDOM IN RUSSIA

TABLE 8.3 Serf and Serfowner Population in Russia, 1859 (A) Serfs

Total

Pomeshchik serfs Male Female Peasants, various departments Male Female Mill and factory Male Female Total (B) Serfowners

Landless gentry Pomeshchiks holding: up to 21 serfs 21 to 100 serfs 101 to 500 serfs 501 to 1000 serfs
1000 serfs Total

Share

22,284,876 10,858,357 11,426,519 242,156

0.97 0.47 0.50 0.01

127,101 115,055 542,599 259,455 283,144 23,069,631

0.01 0.00 0.02 0.01 0.01 1.00

Total

Number Male Serfs Owned

3,633 100,247 41,016 35,498 19,930 2,421 1,382 103,880

12,045 10,539,137 327,534 1,666,073 3,925,102 1,569,888 3,050,540 10,551,182

(C) Total inhabitants of the Empire Serf population, percent total Serfowner population, percent total Average serf/serfowner ratio, males Average serf/serfowner ratio, all pomeshchik

Share Serfowners by Category 0.03 0.97 0.39 0.34 0.19 0.02 0.01 1.00

64,081,167 34.39 0.16 101.57 222.30

Source: 10th Revision, Russian National Census as reported in Troinitskii (1861); author’s calculations. Note: Data in panel (A) are for European Russia, Siberia, and the Transcaucasian krai. Data in panel (B) are for European Russia.

clothing, and state taxes. Seventy-six percent of pomeshchik serfowners held 100 or fewer serfs. Three percent of pomeshchik serfowners held more than 500 serfs. Regions and provinces varied widely in the proportion of serf population. More than one third of provinces and oblasts in European Russia and the Transcaucasus had serf populations of greater than

186

LISA D. COOK

TABLE 8.4 Serfs on Obrok and Barshchina Estates, Shares, 18th and 19th centuries Latter 18th Century Non-Black Earth Provinces

Mid-19th Century

obrok, share

barshchina, share

obrok, share

barshchina, share

Olonets St. Petersburg Pskov Novgorod Smolensk Tver Iaroslav Kostroma Vologda Vladimir Moscow Kaluga Nizhnii Novgorod

66.0 51.0 21.0 49.0 30.0 46.0 78.0 85.0 83.0 50.0 36.0 58.0 82.0

34.0 49.0 79.0 51.0 70.0 54.0 22.0 15.0 17.0 50.0 64.0 42.0 18.0

72.0 — 23.0 45.5 27.0 40.0 87.4 87.5 84.0 70.0 68.0 55.0 68.0

28.0 — 77.0 54.4 73.0 59.0 12.6 12.5 16.0 30.0 32.0 45.0 32.0

Average

56.5

43.5

60.6

39.3

Black Earth Provinces Orel Tula Riazan Penza Tambov Kursk Voronezh

34.0 8.0 19.0 52.0 22.0 8.0 64.0

66.0 92.0 81.0 48.0 78.0 92.0 36.0

28.0 25.0 38.0 25.0 22.0 24.5 45.0

72.0 75.0 62.0 75.0 78.0 75.5 55.0

Average

29.6

70.4

29.6

70.4

Western Borderlands Belorussia, Lithuania Right-bank Ukraine Left-bank Ukraine

— — —

— — —

7.7 2.6 0.7

92.3 97.4 99.3

Average

—

—

3.7

96.3

Source: Ignatovich in Blum (1961), p. 396 (non-black earth and black earth provinces); Koval’chenko in Moon (2001), p.15 (western borderland). Note: The “black earth” zone is the most fertile region in Russia. “Obrok” implies payment in cash or in kind by serfs, and “barshchina” implies payment with personal labor.

100

259 259 259 259 259 259 259 356

0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.96

Index A

0.27

Roubles

96 115 94 103 72 58 37 32

100

Index B 0.40 0.50 0.70 0.90 1.20 1.60 3.00 4.50 6.00 7.50

Roubles 100 125 175 225 300 400 750 1125 1500 1875

Index A

Dues (obrok)

100 69 67 100 108 159 207 253 219 169

Index B

452 1130 2778

1.50

100

Index A

0.75 1.20

0.60

Area, desiatinii

250

125 200

100

Index B

Value of harvest

Labor (barshchina)

SEIGNIORIAL OBLIGATIONS

100 180 263 226 277 251 362 444 686 1110

Index of Grain Prices

Source: Mironov (1992) in Moon (2001), p. 12. Note: Obligations are means for male serfs. Direct taxes comprise a household tax to 1723 and a poll tax from 1724. Index A reports nominal values. Index B is deflated by the grain-price series. Labor services (barshchina) are calculated using area of land cultivated and value of cultivated output.

1700–09 1710–19 1720–29 1730–39 1740–49 1750–59 1760–69 1770–79 1780–89 1790–99

Years

Direct Taxes

STATE OBLIGATIONS

TABLE 8.5 Serfs’ State and Seigniorial Obligations, Male Serfs, 1700–99

188

LISA D. COOK

half the total population, and three-quarters of provinces and oblasts had serf populations that exceeded one quarter. A second category of serfs comprised peasants belonging to government or state institutions (departments) and non-gentry estates and represented one percent of the total serf population.5 A third category of serfs included those assigned to private mills and, to a lesser extent, to factories.6 The share of female serfs is slightly higher among pomeshchik and mill/factory serfs and lower among department serfs. To summarize the serfs’ laborsupply decision, we see, as in the case of slavery, serfs never received the full value of their marginal product. As aforementioned, serfs primarily worked in agriculture. Rye was the dominant crop in the eighteenth century, while wheat became the dominant crop by the mid-nineteenth century. In general, cereals, flax, and hemp were the most important crops. By the late 1850s, serfs in the 43 provinces held 3.2 desiatinii, or 8.6 acres, of land on average per person, and this included arable land, meadow, and farmyard. Barshchina serf holdings weretypicallysmallerthanobrokserfholdings,becauseseigniorial demesne consumed part of the land.7 Productivity was relatively low in Russian agriculture over time and by international standards. Yields were almost identical in the eighteenth and nineteenth centuries. By the mid-nineteenth century, European Russia had the lowest yields of any European country: 17 bushels per hectare compared to 39.7 bushels in Belgium and Holland; 37.4 bushels in Saxony, Great Britain, Württemberg, and Baden; 26.4 bushels in Sweden; and 21.5 bushels in Norway. Observers attribute low productivity to a number of factors, including poor soil and climate conditions; the incentive problem faced by serfs with respect to improvements in lands that could be repartitioned at any time; neglect, lack of market orientation, and inefficient land management among proprietors; and lack of serf mobility. Low productivity contributed to poverty among serfs and diminished revenue among landlords (Blum 1961, 327–333). While estates were mainly oriented towards the production of food and other products for consumption by the estate owner, estates were, nonetheless, the principal sources of agricultural production. Serfs on obrok estates and state peasants were the second largest sources of agricultural Among these institutions and estates were philanthropic and educational institutions, towns in the western krai, churches and monasteries in the western krai, and the department of the Commander of the St. Petersburg Fortress. A krai is an administrative region comprising several provinces. 6 Serfs assigned to the category known as “mills” were in fact assigned to mines, and among the factories to which serfs were assigned were textile, glass, and salt factories. 7 Estimates of serf landholdings were derived from data gathered by the Editing Commissions, which were charged with preparing the emancipation statutes. See Blum (1961, 529–530). 5

THE END OF SERFDOM IN RUSSIA

189

production, since they could allocate time between cultivation and other activities, unlike the barshchina serfs. For example, it has been estimated that by the mid-nineteenth century, estate owners were responsible for 90 percent of grain supplied to domestic and export markets, and serfs were responsible for 10 percent (Blum 1961, 390–392). Serfs were to perform all tasks, within reason, requested by the estate owner. Many of these duties related to agriculture but they were not limited to this sector. Historically, members of the small merchant class established and participated in large-scale manufacturing, but nobles and peasants began to establish manufacturing firms beginning in the 1760s. By 1813–14, nobles accounted for 64 percent of mining firms, 78 percent of woolen-cloth firms, 60 percent of paper mills, 66 percent of crystal and glass works, and 80 percent of potash works. As partial fulfillment of serfs’ barshchina obligation, serfowners could require serfs to work in these firms. It is believed that the number of serfs owning small factories was large, but relatively few owned large enterprises or became millionaires.8 One of the most striking features of the peasant industrialist was not his or her ability to establish and operate a factory or other business. Of greater interest is the fact that these serfs employed free persons, often to perform menial tasks, and sometimes owned serfs themselves.9

The End-of-Serfdom Debate There was a reinvigorated debate concerning the factors influencing the adoption of slavery and serfdom and factors affecting the end of serfdom in Russia and the end of slavery elsewhere among economic historians in the 1970s and early 1980s. The topics were worthy of debate, because there have been many equilibrium paths to the adoption and elimination of institutionalized forced labor across time and space. Among the theories put forward concerning the abolition of serfdom were political and military raisons d’état, cultural factors, fear of serf revolt, the general “crisis of serfdom,” and the unprofitability of serfdom. Blum (1961) points to several hypotheses, including the spread of humanitarian and liberal ideas among Russian intellectuals and nobility, fear of serf revolt, Russian defeat in the Crimean War, and political raisons d’état. It is the last explanation he finds most credible, or at least the least suspect. Tsar Alexander II was convinced that Russia had begun 8 Blum (1961, 297–298). Landowners engaged in manufacturing remained a minority, however, and most continued in agricultural production. 9 See Blum (1961, 298–301) for a discussion of the Nikita Demidov (a serf) and Sheremetev serfs’ examples.

190

LISA D. COOK

the process of political disintegration after defeat in the Crimean War, he argues, and, unlike Tsar Nicholas, possessed the political will to bring about social revolution, despite the protests of nobles and vested interests.10 Rieber (1966) emphasizes military considerations in explaining the end of serfdom. He also points to the Crimean War of 1853 to 1856 as the catalyst for Alexander II’s embrace of reform. However, unlike Blum, he suggests that the direct implications of defeat—diplomatic isolation and military weakness—were of greater import than the indirect implication—the end of autocracy. Rieber also highlights Alexander II’s resolve and strength of character in bringing an end to serfdom, and, like Blum, appears to implicitly accept a “great-man” hypothesis. That is, by historical accident one person, Alexander II, possessed exceptional qualities and took extraordinary actions that changed the course of history. Field (1976) argues that a cultural explanation is the most defensible, because of its power to explain commitment to reform, which moral arguments can not do. On this view, Russian elites could no longer reconcile their incomes and livelihoods derived from a servile system with the cosmopolitan values of their education and upbringing. Serfdom was an embarrassment for the elites and constrained convergence with civilized nations and their norms. Gershchenkron (1965), in addition to a number of Western and Soviet historians, finds the serf-revolt hypothesis more attractive than does Blum.11 Four serious serf revolts in the seventeenth and eighteenth centuries and increases in the frequency of protests among serfs in the nineteenth century led many to believe that emancipation would diminish the threat of a large-scale serf revolt and the threat to social instability. Indeed, the data presented in figure 8.1 suggest increased violence by serfs between 1826 and 1861. While the estimates differ, each data series implies that peasant “disturbances” increased more than fourfold over the period. Several Marxist historians, including Lyashchenko (1949), Nechkina (1962), Koval’chenko (1967), and Pokrovskii (1924), considered the “crisis of the feudal-serf system” as the principle explanation for the end of serfdom. Among the features of the crisis they cite are increased exploitation of serfs by nobles, declining living standards, and peasant discontent, which was expected in the class struggle as capitalism replaced feudalism.12 See Blum (1961, 612–618), for a more extensive elaboration of these and other views. Moon (2001) reviews the work of a number of Russian and Soviet historians who attempted to describe and quantify peasant unrest and who attempted to establish a causal link between this unrest and the end of serfdom. 12 Moon (2001, 19–28) also includes a rich discussion of the Marxist interpretation of the end of serfdom. Nechkina (1962) was the leader of a group of Soviet historians who began studying the political and social causes of emancipation in 1958. 10 11

191

THE END OF SERFDOM IN RUSSIA

1000

Number of Disturbances

900 800

Ignatovich (1925) Druzhinin series (1961–66) Semeveskii (1888)

700 600 500 400 300 200 100 0 1826–29

1830–34 1835–39 1840–44 1845–49 1850–54 1855–61

Year Figure 8.1 “Peasant disturbances,” Russia, 1826–61. (Source: Moon 2001.)

Pokrovskii (1924) deserves special attention, because the research spawned a particularly lively debate among economic historians in the mid-1980s. He argues, but does not convincingly demonstrate, that serfdom became unprofitable and led to the abolition of serfdom. Historians and political scientists have brought considerable evidence to bear to the contrary. Domar and Machina (1984) subject this hypothesis to econometric tests, and they cannot reject the hypothesis that serfdom was still profitable. Toumanoff (1973) revises and extends their theoretical analysis and concludes that voluntary manumission would have been the appropriate empirical evidence to use to demonstrate that serfdom was unprofitable. While researchers have been able to reject hypotheses, such as the unprofitability of serfdom, the debate is still open with respect to the set of likely determinants, or, more precisely, the weight assigned various likely determinants. This debate will not be settled in this paper, but the paper’s modest goal is to review each major theory. In the following sections, I survey the principal emancipation schemes considered and implementation of the 1861 Emancipation Statutes, compare the Russian and other emancipation schemes, and offer suggestions for contemporary Sudan.

192

LISA D. COOK

The Road to the End of Serfdom As can be seen in table 8.1, it took nearly a century and a half to end the centuries-year-old institution of serfdom in Russia. The chronology is not exhaustive, and there were many events surrounding the development and implementation of the 1861 Emancipation Statutes that were excluded. But the timeline reveals that both emancipation and development of emancipation schemes proceeded unevenly in Russia. The reason most cited among observers of serfdom is the binding constraint of opposition to reform on the part of entrenched and relatively powerful vested interests and nobility. Given this context, two features of Russian emancipation appear relevant to the current discussion: timing and estate owner compensation. Architects of the emancipation statutes designed a program of gradual emancipation. On the whole, the tsarist government implemented and observed the outcomes of limited abolition experiments in the western provinces, that is, the Baltic and right-bank Ukraine provinces. The 1816–19 abolition program initially left serfs in these provinces landless, but the Russian government reversed its position on the land question due to concern for the threat to social order caused by rootless peasants and for serfs’ living standards, which declined after abolition. The 1847–1848 reform program in right-bank Ukraine followed a similar pattern of execution and subsequent partial reversal. It is interesting to note that, 40 years after the Baltic experience, the first empire-wide abolition program was proposed without giving freed serfs access to land, and it, too, was revised prior to 1861. The abolition of serfdom occurred in three general phases in accordance with the Statutes of 1861. First, freed serfs entered a two-year transitional period, during which their land allotments and seigniorial obligations remained fixed.13 Second, in the “temporary obligation” phase, serfs and estate owners were to be bound by regulatory charters, which informed them of the land allocations estate owners were to grant peasants for permanent use and the freed serfs’ obligations to estate owners. The final stage was redemption, during which peasants could buy, or redeem, land allotments from estate owners. This was the stated gradualist abolition program. In fact, practices and institutions emerged to make the actual program more gradual, quantitatively and qualitatively. One set of practices, otrabotka, involved peasants’ working off various types of payments. Among these practices were 13 While the legal status of serfs changed from unfree to free, freedom of movement was not granted until 1906, the year before all abolition-related activities ceased. The move was designed in part to discourage default on redemption payments. See Moon (2001, 80).

THE END OF SERFDOM IN RUSSIA

193

conversion of peasants dues, or obrok, under “temporary obligation” and of payments in arrears into labor on the landowners’ estate; labor demanded from peasants in exchange for access to meadows, pastures, water sources, and land severed from their previous allotments; and lending by landowners in which peasants were obliged to work off interest. As in the American South, the institution of sharecropping also emerged. In Russia, peasants were often rented land by the estate owners and required to make payments in terms of output, typically half, from the land.14 Further and in contrast to the emancipation of slaves in the American South, dependence on estate owners was effectively extended by the continued prohibition on movement by serfs. The upshot of such policies, practices, and institutions is that, rather than being simple and abrupt, the process of manumission was complex and slow. The redemption program was designed to create a broad class of landowning peasants, known as “peasant-proprietors,” who would be independent of the noble landowners. The terms of the redemption program were favorable to the estate owners in order to provide an incentive to opt into the program. In addition to compensating landowners, the program was initially voluntary for the landowner but obligatory for the former serf once the landowner decided to engage in the redemption process. The program proceeded as follows. The peasant and estate owner agreed on a land allotment for purchase by the peasant, a minimum for which was set in the Statutes. Most allotments permitted to be redeemed were lands in permanent use during temporary obligation. The redemption amount did not depend on the market price of land. Instead, it was calculated using the cash dues peasants paid to estate owners for the use of their allotments during temporary obligation. Few serfs could finance the entire redemption payment. Therefore, the Russian government advanced the estate owner 75 to 80 percent of the redemption amount in bank notes and long-term bonds earning annual interest of five percent. Freed serfs would be responsible for paying estate owners the remaining 20 to 25 percent of the redemption amount.15 The peasants’ governmentfinanced mortgage was fixed at a rate of six percent annual interest and for a term of 49 years. Peasants could not contract individually to redeem their allotments. Rather, peasants were obliged to contract jointly with 14 In fact, this system was identical to the institution of obrok payment in kind under serfdom. See Moon (2001) for a fuller discussion of post-abolition practices by which serfdom was effectively extended. 15 The redemption amount was calculated using the stream of income from cash dues peasants owed estate owners following “temporary obligation.” Peasants were to pay all the redemption-related payments to the government rather than directly to estate owners. However, peasants were often required to “work off” redemption payments owed the estate owners as in the cases mentioned above.

194

LISA D. COOK

their village communities, and villagers were jointly responsible for the repayment of the mortgage. Therefore, constraints on peasant migration and on allotment use remained in place during the execution of the redemption program, which lasted until 1907. Between 1864 and 1870, the proportion of freed serfs redeeming their land rose from 10 to 67 percent. Rates of participation, however, were uneven, since noble landowners identified and exploited arbitrage opportunities related to timing and to labor-service–peasant and cash-duespaying-peasant regimes. In 1881, 15 percent of adult male freed serfs were still under the system of temporary obligation. As a result, the tsarist government sought to minimize opportunities for arbitrage by making redemption mandatory from 1883. Many historians claim the peasants’ payments were excessive and onerous. Indeed, the redemption amount was set above the value of land at current market prices (see note 15), but economic historians’ estimates of the real value of redemption payments differ significantly. The Gershchenkron (1965) data imply that serfs were overcharged 37 percent in redemption payments. On the contrary, Gregory (1994) suggests that serfs paid negative prices for redeemed land due to land-price inflation in the two decades following emancipation. Table 8.6 presents the Gershchenkron estimates.16 Thus, the magnitude and direction of financial costs to peasants for their liberation are still an open research question. In general, it is difficult to ascertain whether former serfs were better off once the third and final phase of abolition—redemption—was complete. Early and Marxist researchers observed that significant amounts of land were taken from peasants and that land shortages among peasants increased in the decades following 1861. More recently, these outcomes have been associated more with rapid rural population growth and fundamental change in the Russian economy than with specific elements of the abolition program.

Russia Then and Sudan Now How does the Russian historical experience of ending serfdom inform those seeking to end slavery in Sudan currently? It might be argued that the special features of this (or any) historical economic event are anachronistic. The economic historian’s task is to carefully assess past economic events with a view to informing later economic phenomena. Indeed, both country-specific similarities and differences and more general observations are salient. 16

This discussion relies heavily on Moon (2001, 80–105).

195

THE END OF SERFDOM IN RUSSIA

TABLE 8.6 Redemption Payments and Prices for Land, Loitsky Series

Region Russia and LeftBank Ukraine Non-black earth Black earth Western provinces Total

Value of land at market prices (R, millions)

Land Allotment (000s des.)

1854–58

12186 9841 10141 32168

155 219 170 544

1863–72

Value of land, redemption value (R, millions)

Redemption value, % market prices 1863–72

180 284 184 648

342 342 183 867

190.0 120.0 100.0 136.7

Source: Gershchenkron (1965).

Note: Entry for Total in final column is mean redemption value to market price.

In general, architects of emancipation schemes have assigned more, if not all, weight to the property rights of the owners of serfs or slaves rather than to those of the serfs or slaves. Like most schemes of gradual emancipation over time and around the world, the Russian abolition program required that the direct financial cost of freedom for serfs be borne by the unfree persons themselves. Further, as Engerman (1996) finds in his survey of emancipations in the Americas, Europe, Africa, and Asia, compensation was not paid to serfs and slaves in these emancipation programs. Goldin (1973) also shows complete expropriation of slaveowner property to be missing in emancipation programs in the Western Hemisphere. In striking contrast to the case of the northeastern United States in the late 1700s and early 1800s, the debate among serfowners was neither whether they would be compensated nor who should pay compensation, but rather the magnitude of compensation in Russia that was contentious.17 Another characteristic of emancipation programs is their historical clustering. Engerman (1996) shows that most emancipations around the world happened between the late eighteenth and late nineteenth centuries. There was also clustering in subperiods: 11 between 1801 and 1820, five between 1831 and 1832, 13 between 1841 and 1850, and 4 between 1851 and 1864. By extension, international pressure in favor of abolition increased over the period such that late emancipators became more isolated over time. 17 Freed serfs’ freedom of movement was also a bone of contention. See Engerman and Fogel (1974) for a rich analysis of gradual emancipation in the United States and Great Britain.

196

LISA D. COOK

Given these two dominant features of abolition programs historically, we may attempt to inform policies that may bring about an end to slavery in Sudan. First, the illegality of the means of enslavement and the practice of slavery must be established, acknowledged, and acted upon by the Sudanese government. Slavery and serfdom were legal institutions when abolished over time in other countries and in Sudan itself. Slaveowners, then, can legitimate neither their claim of property rights over slaves nor their claim to compensation for confiscated property. By extension, the slave’s claim of compensation for illegal enslavement and forced labor would be legitimate. A second, more promising strategy stems from the waves of abolition pressure in history. Often, abolition is coupled with other events. Recall that among the leading explanations for the end of serfdom was the diplomatic isolation following Russia’s defeat in the Crimean War. As an agreement to end Africa’s longest-running civil war has been signed, the humanitarian crisis in Darfur intensifies and receives more international attention, and the term “genocide” begins to be used more widely, it may be possible to bring international pressure to bear, to get Sudan’s consent to end the practice of slavery.18 The attempt to end slavery would appear futile without addressing the factors that allow its practice, including the stalled release and return of those persons abducted prior to 2005.19 Finally, the Russian and American experiences with abolition of servile systems raise an important issue concerning the actual act of emancipation. Should freedom of slaves in Sudan be gradual, as in Russia, or immediate? Political theory and direct evidence should be considered in addressing this question. Condorcet ([1788] 1994), for example, argued for abolition and gradualism. On his view, in the American South it was not necessary for the current slaves to be enslaved to execute the tasks required by plantation owners. This was a compromise that would allow freedom but would diminish opposition by the colonists, since free laborers who feel an attachment to the land would not have an incentive to destroy it. In addition, he suggested that a period of transition was necessary for former slaves in order to prepare and educate them for freedom. On the one hand, this view has intuitive appeal, since reintegration into the labor force and into civil society will likely require a period of reorientation. On the other hand, given Sudan’s recent and ancient history 18 Among U.S. efforts have been imposing sanctions in the mid-1990s, putting Sudan on its Trafficking in Persons list for possible sanctions in 2003, and brokering a peace deal between the government and the rebels in southern Sudan in 2002. 19 Further, it might be useful for proponents of abolition to seek pressure from both Arab and non-Arab countries, particularly because of the ethnic dimension through “arabization” that slavery has taken on in Sudan. At the time of writing, no known pressure has been exerted on Sudan by Arab countries.

THE END OF SERFDOM IN RUSSIA

197

with slavery and government indifference to or complicity in the practice, temporary guardianship by individual slaveowners or the state, or gradual emancipation more generally, may not be credible. It may encourage, for example, more intensive use of slaves or trafficking of slaves to areas where forced labor also remains.20 Ceteris paribus, it appears that the more credible policy option is immediate emancipation. Evidence from Russia and the United States suggests that at least one desirable and two undesirable outcomes resulted, regardless of intended speed of reform. The common desirable outcome was that former serfs and slaves were ultimately able to acquire land in significant numbers. In Russia, the amount of land peasants owned rose from 6.5 to 23.5 million desiatinii between 1877 and 1907; and in the United States by the 1890s, nearly 25 percent of southern black farmers owned their land (Moon 2001, 112). With respect to less desirable outcomes, former serfs in Russia were not permitted to leave their land, and, beginning in the 1890s in the United States, blacks faced restrictions on movement, among other things, as segregationist laws and practices were adopted in the South and then nationwide. A second undesirable outcome was the emergence of the sharecropping system in the place of serfdom and slavery. Would one institution replace another in Sudan? This is difficult to know. We would need data on labor-market conditions and former slaves’ opportunities. As a poor, developing country, on the border between medium and low development in the UN Human Development Index with a GDP per capita of $1890, it likely has a relatively high unemployment rate and a high proportion of unskilled workers in the labor market.21 It would be reasonable to infer that the marginal rate of substitution between free and non-free unskilled labor would be high.22 But this assumes a flexible, freely adjusting labor market with profit-maximizing owners.23 This assumption may be too strong, given the limited data we have on owner preferences and behavior and the motives for and implementation of continued slavery in Sudan. 20 Engerman and Fogel (1974) and Goldin (1973) identify the intensification of work by slaves, slave smuggling, and other drawbacks of implementing gradual emancipation around the world and over time. 21 Sudan ranks at the bottom of the “medium development” UN classification, but this ranking is based on data from northern Sudan, which is historically better off than southern Sudan. See UNDP (2004). 22 A high marginal rate of substitution assumes that tasks being performed by forced labor are suitable for unskilled labor. Certainly, this issue is complicated by war. As a result of slaves’ abductions from contested areas, is their proper designation “slaves” or “prisoners of war,” or something else? 23 See Karlan and Krueger, (Chapter 1) and Rogers and Swinnerton (Chapter 2) in this volume for elaboration of labor-market supply and demand conditions.

198

LISA D. COOK

The lessons from the emancipation of Russian serfs may be useful in informing policymakers in considering slavery and emancipation in Sudan. The evidence over time and across countries demonstrates that the enserfed or enslaved typically finance their own freedom. Given the complex and urgent situation in Sudan, economic history might suggest a departure from the experience of Russian serfdom and emancipations generally. While the return of abductees has gone slowly and less than transparently, gradual emancipation may not be a defensible stated policy option for Sudanese policymakers. Further, compensation of slaveowners as a policy is likely infeasible, too, since slavery and similar practices are illegal in Sudan. Because forced labor takes many forms in Sudan, compensation would rely heavily on the scope, quality, and duration of enslavement. While we may have reached the limits of lessons from economic history, more recent agreements, such as the plan of compensation of victims for Nazi Germany’s use of slave labor or Japanese internment during World War II in the United States, might be instructive.24 Bibliography Blum, Jerome (1961), Lord and Peasant in Russia from the Ninth to the Nineteenth Century. Princeton: Princeton University Press. Bok, Francis (2003), Escape from Slavery: The True Story of My Ten Years in Captivity and My Journey to Freedom in America. New York: St. Martin’s Press. Condorcet, Marquis de ([1788] 1994), “Rules for the Society of the Friends of Negroes (1788),” in Iain McLean and Fiona Hewitt (eds.), Condorcet: Foundations of Social Choice and Political Theory (Hants, England: Edward Elgar Publishing Limited). Domar, Evsey D., and Mark L. Machina (1984), “On the Profitability of Russian Serfdom,” Journal of Economic History 44(4): 919–955. Emmons, Terence, ed. (1970), Emancipation of the Serfs in Russia. New York: Holt, Rinehart, and Winston. Engerman, Stanley (1996), “Emancipations in Comparative Perspective,” in Gert Oostinde (ed.), Fifty Years Later: Antislavery, Capitalism and Modernity in the Dutch Orbit (Pittsburgh: University of Pittsburgh Press). Engerman, Stanley, and Robert W. Fogel (1974), “Philanthropy at Bargain Prices: Notes on the Economics of Gradual Emancipation,” Journal of Legal Studies 3(2): 377–401. Field, Daniel (1976), The End of Serfdom: Nobility and Bureaucracy in Russia, 1855–1861. Cambridge, MA: Harvard University Press. 24 The author wishes to thank Laura Beny, Martin Bunzl, and Stanley Engerman for helpful comments and participants in the Princeton Slavery roundtables and the American Economic Association 2005 meetings for helpful conversations. This research was conducted while at the Hoover Institution, Stanford University, and its support is gratefully acknowledged.

THE END OF SERFDOM IN RUSSIA

199

Gershchenkron, Alexander (1962), Economic Backwardness in Historical Perspective. Cambridge, MA: Harvard University Press. Gershchenkron, Alexander (1965), “Agrarian Policies and Industrialization: Russia, 1861–1917,” in M. M. Postan and H. J. Habakkuk (eds), Cambridge Economic History of Europe, vol. 6, part 2, (Cambridge: Cambridge University Press). Goldin, Claudia (1973), “The Economics of Emancipation,” Journal of Economic History 33(1): 66–85. Gregory, Paul (1994), Before Command: An Economic History of Russia from Emancipation to the First Five-Year Plan. Princeton: Princeton University Press. Hoch, S. L., and W. R. Augustine (2001), “The Tax Censuses and the Decline of the Serf Population in Imperial Russia, 1833–1858,” in David Moon, The Abolition of Serfdom in Russia (Harlow: Longman). Kahan, Arcadius (1973), “Notes on Serfdom in Western and Eastern Europe,” Journal of Economic History 33(1): 86–99. Koval’chenko, I. D. (1967), Russkoe krepostnoe krest’ianstvo v pervoi polovine XIX v. (The Russian Serf Peasantry in the First Half of the 19th Century). Moscow: Izdatel’stvo Moskovskogo Universiteta. Lyashchenko, P. I. (1949), History of the National Economy of Russia to the 1917 Revolution, trans. L. H. Herman. New York: Macmillan. Moon, David (2001), The Abolition of Serfdom in Russia. Harlow, England: Longman. Nazer, Mende, and Damien Lewis (2004), Slave. London: Public Affairs. Nechkina, M. V. (1962), Revoliutsionnaia situatsiia v Rossii v 1859–1861 gg. (The Revolutionary Situation in Russia in 1859–1861), trans. Terence Emmons. Moscow: Izdatel’stvo Akademii Nauk SSSR. Pokrovskii, M. N. (1924), Russkaia istoriia s drevneishikh veremen. Tom 4. Izdanie piatoe (Russian History from Earliest Times, vol. 4, 5th ed.), trans. Erica Brendel. Leningrad: Gosudarstvennoe izdaltel’stvo. Rieber, Alfred J. ed. (1966), The Politics of Autocracy: Letters of Alexander II to Prince A. I. Bariatinskii. Paris: Mouton and Company. Rift Valley Institute (2003), “Sudan Abduction and Slavery Project: Abductee Database Report 2003,” www.riftvalley.net/inside/projects.htm. Toumanoff, Peter G. (1973), “The Profitability of Serfdom: A Comment,” Journal of Economic History 45(4): 955–959. Troinitskii, A. (1861), The Serf Population in Russia According to the Tenth National Census, trans. Elaine Herman. St. Petersburg: Karl Vul’f Press. United Nations Development Programme (UNDP). (2004). Human Development Report 2004, hdr.undp.org/reports/global/2004/. U.S. State Department, Bureau of African Affairs (2002), Slavery, Abduction and Forced Servitude in Sudan, Report of the International Eminent Persons Group (2002), www.state.gov/documents/organization/11951.pdf. Vasagar, Jeevan (2004), “Where Straight Hair Is One of Slavery’s Legacies,” Manchester Guardian Weekly, 23–29 July.

Chapter Nine Conflicting Imperatives: Black and White American Abolitionists Debate Slave Redemption MARGARET M. R. KELLOW

As long as slavery persisted in the United States, Americans wrestled with the probity of purchasing the freedom of those held in bondage. The transaction seemed simple enough—monies paid to a master ended his or her claim to the person of the individual who had previously been held in bondage. By virtue of the proceeding, the slave became free. However, in the period between the American Revolution and the Civil War virtually every aspect of this transaction would become problematized. Enslaved African Americans and black and white abolitionists denied a slaveholder’s right to place a price on the freedom of another human being. At the same time other opponents of slavery advanced various schemes to compensate masters as a means of bringing slavery to an end. White antislavery activists denounced compensation schemes as bribery, but contributed readily to funds to purchase the liberty of individual slaves. Fugitive slaves stole themselves out of their masters’ possession, thereby rejecting claims of ownership, and then spent years raising the money to liberate their loved ones. These and other contradictions informed the debate over slave redemption in the United States in the nineteenth century, exposing the tension between the legal and moral condemnation of slavery as an institution and an emotional or humanitarian response to those individuals who were held in bondage. From the very beginning of the colonial period until the Civil War, untold thousands of enslaved African Americans liberated themselves or were liberated by payments to their masters. Francis Payne’s experience was not unusual. In the 1640s in Northampton County on Virginia’s Eastern Shore, Payne bargained with his mistress, Jane Eltonhead, to purchase the freedom of his family. It took Payne thirteen years to raise the 3800 pounds of tobacco his mistress required “for the freedom of himselfe, his wife and children” (as quoted in Breen and Innes 1980, 75). By these and similar means, usually one body at a time, African Americans liberated themselves to form the nucleus of the free black population of the American colonies.

ABOLITIONISTS DEBATE SLAVE REDEMPTION

201

By the time of the American Revolution, difficulty in reconciling slavery with ideals of liberty and equality that Americans espoused began to call into question the institution of slavery, and with it, the validity of a master’s claim to compensation for the loss of those individuals he purported to own. At the same time, however, public discussion on how slavery might be ended repeatedly raised the possibility of compensation to masters. Those who regretted the persistence of slavery and viewed it as an unfortunate legacy of the colonial period suggested that freeing slaves by compensating their masters was a way of solving the problem and doing justice to all concerned. Debates on the Constitution underscored the enormous investment that slavery represented to Southerners, thus early proposals often contained suggestions that African Americans might be emancipated by using the proceeds of the sale of western lands to compensate slaveowners. Recognizing the property right of slaveholders and offering compensation attempted to indemnify them for the losses that emancipation would entail. Supporters of these proposals hoped that compensation would win them the support of at least some slaveowners. Their hopes were disappointed and these early proposals came to nothing as the South rejected the idea out of hand, and white Americans more generally were unwilling to forego the revenue that accrued from the sale of public lands (see Fladeland 1976). In the late eighteenth century, religious concerns about slavery built on the political doubts which arose during the Revolutionary era (see Loveland 1966). This questioning of slavery was often paired with the possibility of compensation for slaveowners in what was understood as a need to do justice to masters and slaves alike. Manumission societies formed in the border states to undertake slave redemption as part of a program to end slavery. Quakers in particular sought to balance the interest of the slave and that of his or her master. Thus they bought and freed slaves in order to live up to their long-standing testimony against slavery. By 1830, for example, North Carolina Quakers had spent nearly $13,000 to ransom slaves. By these means, more than 600 slaves were freed in that state. However, a significant number of the slaves thus freed were elderly or invalids unable to care for themselves, and the Quakers found they had to become guardians to them. More importantly, as racial and economic rationales for the persistence of slavery strengthened in the nineteenth century, southern Quakers became a dissident minority in an environment in which slavery became increasingly difficult to question. Ultimately the efforts of these societies freed very few slaves and did little to undermine slavery as an institution (see Weeks 1896, 224–229). In 1817, the American Colonization Society was founded with a view to ending slavery gradually. By compensating slaveholders who emancipated their slaves and transporting free blacks to Liberia, colonizationists

202

MARGARET M. R. KELLOW

believed that the problem of American slavery could be resolved. The ACS argued that slavery had been established by law and upheld by the state, thus the state had an obligation to compensate those who had legitimately acquired and held slaves. In a letter to the editor of The Liberator (12 July 1834) one supporter of compensation argued that when public opinion, or the law, has sanctioned the investiture of property in any particular way, then it cannot turn round and destroy that system of things, without indemnifying those whom it has encouraged so to invest their property. It is a fundamental principle of law, running through all our constitutions, that no legislature can so legislate as to destroy the tenure of property acquired and held under existing laws without providing for the security of such property. . . . The true question is not between the master and the slave, but between the master and the nation who conferred this right on him. Public discussion over the merits of paying masters to free their slaves became more heated after 1830. The debate emerged from the reaction against gradual plans for emancipation such as that proposed by the ACS. During the 1820s, African Americans such as Samuel Cornish and David Walker, angered by the thinly disguised racism of the colonizationists, denounced ACS plans to deport free blacks to Liberia. “America is more our country, than it is the whites,” proclaimed David Walker’s Appeal in 1829, “we have enriched it with our blood and tears. The greatest riches in all America have arisen from our blood and tears—and will they drive us from our property and homes, which we have earned with our blood?”(1830, 73). In claiming full citizenship for black Americans, these and other black activists rejected understandings of slavery as a regrettable legacy of the past, by means of which slaveholders had acquired a legitimate interest. In their view slavery constituted a moral issue. African American abolitionists argued that slaveholding was inherently unjust. It followed that the putative “owners” of enslaved African Americans had no right to compensation. “Now that which is a man’s right,” William Lloyd Garrison insisted, “there is no law of nature to compel him to buy” (Liberator, 16 June 1832). Persuaded by the arguments of Cornish, Walker, and other black abolitionists, white northerners who opposed slavery distanced themselves from colonizationist proposals involving compensation and from any other proposition that in any way legitimated the property rights of slaveholders. As Garrison’s Boston newspaper The Liberator began arguing in 1831, if the Declaration of Independence meant anything, enslaved men and women should be free, and those who deprived them of their freedom were tyrants. Garrison framed the question in the language of morality. The slaveowner was a sinner, and one did not bribe the sinner to cease his transgression. Moreover, Garrison insisted that to do so

ABOLITIONISTS DEBATE SLAVE REDEMPTION

203

would be to participate in the sin. “[I]t is wrong,” he wrote in an editorial in The Liberator in June of 1832, “and consequently sinful, to give money, or any other pretended equivalent, for ‘the bodies and souls of men,’ under any pretence whatever.” Garrison’s comment suggests that among the many aspects of compensation that troubled him was the perception that when an abolitionist purchased an enslaved person’s freedom, that abolitionist became, however briefly, a slaveowner. A month later, Garrison reiterated, “[it] is the duty of the owners of slaves to liberate their victims immediately—they deserve and should receive no remuneration for giving up stolen property” (Liberator, 14 July 1832). In vitriolic editorials, pamphlets, and speeches, Garrisonian abolitionists denounced compensation and began to call for immediate abolition. When Garrison and his followers met to form the American Antislavery Society (AASS) the following year in Philadelphia, they drafted a Declaration of Sentiments that included the following proclamation (Liberator, 21 August 1857): We maintain, that no compensation should be given to planters emancipating their slaves, because it would be a surrender of the great fundamental principle, that man cannot hold property in man; because slavery is a crime, and therefore [the slave] is not an article to be sold; because the holders of slaves are not the just proprietors of what they claim—freeing the slaves is not depriving them of property, but restoring it to the right owner—it is not wronging the master, but righting the slave, restoring him to himself. The resolve of the new association could scarcely have been more unequivocal. Despite this strengthening conviction, American abolitionists had to reconsider the issue of compensation immediately. In the same year that the American Antislavery Society was founded, the British government legislated the abolition of slavery in its possessions by enacting a program of compensated emancipation. In the British West Indies, slaves were to serve an apprenticeship of five years and then be freed. West Indian planters would receive twenty million pounds in compensation from the British government. Eager to differentiate themselves from their erstwhile colonial masters, American abolitionists criticized compensation ever more fiercely. Garrison, traveling in England at the time the bill was before Parliament, reported to the Liberator in September 1833 that compensation was “justly viewed as money bestowed where no loss can be proved, . . . an abandonment of the high ground of justice.” Comments such as this suggest that American abolitionists envisioned a better, purer form of emancipation in which slaveholders would repent and would require no inducements to let their slaves go free.

204

MARGARET M. R. KELLOW

Despite the firm position of the AASS, the issue came to the fore time and again, most notably in 1847, when English antislavery activists raised money to purchase the freedom of Frederick Douglass, then lecturing in Britain. Douglass had escaped slavery several years previously, so the gesture was largely symbolic, but it provoked strong condemnation in many abolitionist circles. “Seven hundred and fifty dollars have been sacrificed to promote slavery—in clear recognition either of the right of the slaveholder, or the weakness, moral and physical, of the friends of liberty,” thundered the editors of the Boston Chronotype (as quoted in The Liberator, 8 January 1847). Critics cited the AASS denunciation of compensation and pointed out that Douglass needed no “free papers” to make him a free man. Pennsylvania Quaker Lindley Coates made the case against ransoming fugitive slaves most forcefully. He argued in The Liberator of 19 February 1847 that if he paid money to free a slave: I would be increasing the traffic in, and enlarging the market for men, women and children, however I might remonstrate with my tongue against the practice; for I would be giving the slaveholder a stimulus to have such commodities for sale. The slaveholder or slave breeder don’t care, and need not care whether his market is at the South or at the North, whether the planters at the South or the abolitionists at the North, are the purchasers of his slaves, seeing as his object is to make money out of them. Coates went on to say that the antislavery intent of the practice was immaterial. He believed that it was “wrong to buy a slave at all, because, in so doing, we encourage the seller to fill the place of the one sold with another victim, . . . which more than anything else, urges him on to yet further deeds of outrage and wrong towards his fellow-men.” The net effect, Coates concluded, would be to provide an impetus for additional slaves because “as a general rule, it holds good that if a market is created for any kind of goods, it will be supplied.” Black Americans continued to denounce compensation and any other manner of recognizing the property rights of a master. With the majority of their white colleagues in the antislavery movement, black abolitionists agreed that paying to free a slave legitimated the master’s claim to ownership, and they were outspoken in their condemnation of such bargains. Every slave who escaped from bondage denied the legitimacy of his or her master’s financial interest in his or her person. Despite the controversy over his own ransom, Frederick Douglass repudiated compensation emphatically. “[E]very act of purchase enhances the market value of human chattels [sic]; and makes the monsters cling to their property with

ABOLITIONISTS DEBATE SLAVE REDEMPTION

205

a more tenacious grasp,” Douglass wrote in 1849 (The North Star, 13 July): I would simply advise every man, woman and child, whether black, white, or yellow, that so long as they are guiltless of crime, they have a right to freedom. It is theirs. The idea of making them pay for what is their own by the inalienable gift of their Creator, is most absurd, preposterous, and Heaven insulting. Similarly, although in slavery he had worked hard to raise enough money to free himself, once he had escaped from bondage, William Wells Brown scorned a former master’s offer to sell his freedom to Boston abolitionists. “I cannot accept of Mr. Price’s offer to become a purchaser of my body and soul,” he insisted. “God made me as free as he did Enoch Price, and Mr. Price shall never receive a dollar from me or my friends with my consent” (Brown 1849, ix). Henry Bibb, a former fugitive, argued in his Canadian newspaper that the practice of paying for escaped slaves was counterproductive. “A great mistake has been made here in the north, by purchasing the freedom of fugitive slaves . . . it has only served to stimulate the hunt for fugitives.” Bibb went on to suggest that rather than ransoming fugitive slaves, they should be permitted to be returned to the South, where they could instruct their peers about life in the free states and thus disrupt the system from within (Voice of the Fugitive, 5 November 1851). Defiantly, black abolitionists rejected the idea that their bodies were commodities to be bought and sold. The great paradox, however, was that although white and black antislavery activists were for the most part unanimous in their opposition to compensation or to buying the freedom of fugitive slaves, almost everyone in the abolitionist community was involved in redeeming slaves in one way or another. Black churches routinely canvassed their members to ransom the relatives of members of their congregations (see Finkenbine 1993, 180–181). Former slaves published narratives of their lives expressly to raise money to ransom family members. White and black antislavery newspapers carried notices attesting to the honesty of various African Americans attempting to raise money to free themselves or their family members. White abolitionists who belonged to antislavery organizations that explicitly denounced compensation contributed money to help buy individual blacks out of slavery. Henry Ward Beecher staged slave “auctions” in his Brooklyn church to raise money to purchase freedom for a number of young slave women. Even Garrison himself contributed to the “ransom” of Douglass, arguing that the particular circumstances of the case were such that compassion and prudence had to take precedence over principle. “To save a fellow-being,” he wrote in the Liberator in

206

MARGARET M. R. KELLOW

March of 1847, “it is no crime sometimes to comply with even unjust demands.” Within the black community, the inconsistency created few problems. Family reunification overrode other concerns. Even outspoken activists who refused to be redeemed themselves worked hard to raise the funds to free members of their own families. Ties of affection and kinship outweighed other considerations as black men and women went to heroic lengths to free relatives and reunite their families. Edmond Kelley’s efforts to free his family provide a good example of the process. Kelley was born a slave in Maury County, Tennessee, in 1817. He was converted to the Baptist faith in 1838, and married Paralee Walker the following year. He soon felt a call to the ministry and was licensed to preach by the Columbia Baptist Association in 1842. A few years later word reached him that he was about to be sold. The local Baptist association attempted unsuccessfully to raise the money to purchase his freedom. Kelley then tried to work out an arrangement with his mistress, agreeing to pay her $10 a month. Before he could complete the purchase, his mistress, fearing that she would soon become bankrupt, urged Kelley to leave Tennessee and presumably to continue paying her. Kelley took her advice and went first to New York and eventually to New Bedford, Massachusetts, where he was called to be pastor of the Second Baptist Church there in September of 1848. Once in New England, Kelley felt no obligation to continue paying for his own freedom, but he desperately wanted to be reunited with his wife and four children. Kelley entered into negotiations with James Walker, owner of his wife and children, in February of 1850. Walker was extremely reluctant to part with his slaves. Paralee Walker (Kelley) was apparently the daughter of Walker’s old nurse, and he claimed that the ties of affection and his conviction that Kelley could not provide well enough for his family precluded setting them free. However, for $2800, Walker was prepared to think about it. Kelley enlisted the aid of friends and parishioners to help raise the money. The Boston Baptist Association contacted Walker and attempted to persuade him to lower the price. Walker refused to lower the price and hinted that if Kelley did not come up with the money promptly the offer would be retracted. Kelley resigned from his ministry and worked full time on raising the money. Finally, by dint of his own efforts and by canvassing numerous congregations in southern Massachusetts, Kelley managed to raise the money. This did not end his difficulties. As a fugitive, he was not able to return to Tennessee to get his family. An agent had to be hired and paid to conduct the transaction and transport Paralee Walker and her four children to New Bedford. Once they arrived there in May of 1851, James Walker consented to their emancipation (Kelley 1851). Kelley’s experience demonstrates that even when ideolog-

ABOLITIONISTS DEBATE SLAVE REDEMPTION

207

ical issues were not raised, slave redemption was rarely a straightforward proposition. As almost every free black family had relatives still in bondage, thousands of African American families went through experiences such as these. Kelley was fortunate in that he was literate and had friends and acquaintances to help him in his efforts to be reunited with his family. Other fugitives, without friends or money, found it almost impossible to maintain ties with family members in the South. For fugitives, especially following the Fugitive Slave Act of 1850, communicating with a former master in hopes of arranging the purchase of a spouse or child meant revealing their location and thus running the risk of recapture. Returning to the South to fetch family members involved still more dangers. Even if legally emancipated, freed people had few levers with which to pressure former masters to relinquish their kin. The occupational opportunities open to African Americans meant that the accumulation of the necessary funds required enormous efforts. When the money had finally been saved, legal problems and the difficulties of transporting money over long distances remained to be overcome. Since a slave could not purchase another slave, agents had to be employed, and these were not always trustworthy. There were few guarantees that the master would live up to his promise and little recourse if he failed to do so. Yet painstakingly, despite all these obstacles, black families struggled to redeem their loved ones. For those still enslaved and hoping to purchase their liberty, the desire for freedom forced hard choices at almost every turn. Typically, one family member managed to free him or herself and then worked to purchase those still enslaved. Yet self-purchase was particularly fraught with difficulties, as whatever a slave earned or owned legally belonged to his or her master. Arrangements could be struck, but a master might renege, as Moses Grandy learned to his sorrow. Grandy paid his purchase price three times over before finally achieving his freedom (Grandy 1843). Lunsford Lane worked diligently for years to free himself and his family, but he found that local white communities could take offense if a slave was seen to be too successful (Lane 1842). In addition, laws requiring manumitted blacks to leave southern states meant that the pursuit of freedom often entailed years of separation. Even the decision about who should seek freedom first was fraught with difficulties. Arguably it made more sense for a man to purchase his freedom first because he could earn more money and then free the rest of his family more rapidly. Against that, if a black woman remained behind in slavery, any children she might have would be born slaves and thus would also have to be ransomed, raising the amount of money needed considerably. In addition, if an enslaved woman’s husband left the neighborhood, she might be forced into another relationship. Either way, the desire for freedom could lead to

208

MARGARET M. R. KELLOW

years of separation and uncertainty (see Lebsock 1984, 96; Wood 1995, 123–125). Despite their stated objections to paying slaveholders to emancipate their slaves, white abolitionists found their hearts touched by stories such as that of Edmond Kelley’s family. The ties of family and kin were sacred in the culture of antebellum America, and families separated by the machinations of cruel and greedy men were the stock and trade of nineteenth-century sentimental fiction. The figure of the lost child or the faithful spouse inevitably evoked a moral and emotional response. Thus even while denouncing the idea of compensation for slaveowners, white abolitionists acknowledged that every one of their number contributed to these appeals. White congregations gave generously to collections to ransom particular slaves (Shaw 2000). When abolitionist papers exposed as frauds specific African Americans who sought money purportedly to rescue family members, they implied that others were genuine, and thus implicitly endorsed that practice. Garrison, in defending his support for the ransoming of Frederick Douglass, asserted that every one of those who had signed the AASS’s 1833 Declaration condemning compensation “had again and again, and have subsequently, contributed toward ransoming some father or mother, husband or wife, parent or child from slavery without ever dreaming that they were trampling upon moral principle in so doing”(Liberator, 5 March 1847). In a letter to an English friend (Elizabeth Pease, 1 April 1847, in Merrill 1973, 474), Garrison went further, saying that he had never entertained for a moment, that it is wrong to ransom one held in cruel captivity; though I have always maintained, in the case of the slave, that the demand of the slaveholder for compensation was an unjust one. But I see no discrepancy in saying that a certain demand is unjust, and yet being willing to submit to it, in order to save a brother man, if this is clearly made to be the only alternative left to me. At some point, Garrison seemed to be arguing, the need to rescue overcame repugnance toward compensation. When the enslaved individual had a recognizable face, was in some way known or immediate to white abolitionists, they were much more likely to involve themselves in the projected redemption. Garrison’s ambivalence was not unique. White abolitionists had compelling reasons beyond sentiment for supporting the efforts of African Americans to purchase their own freedom and that of their families. As white abolitionists struggled not only against slavery but also against the racist attitudes of their contemporaries, the freedman or woman who went to heroic lengths to free members of his or her family powerfully

ABOLITIONISTS DEBATE SLAVE REDEMPTION

209

contradicted Southern assertions that family ties mattered little to their slaves. Similarly, racist attitudes regarding the sexual mores of African Americans could be rebutted by references to black men and women who labored long and hard to free their spouses and demonstrated fidelity that lasted over years of separation. The slave who freed himself and his family by virtue of his own efforts also contradicted the images of indolence and dependency that attached to freed people in the minds of many northerners. The image of black men toiling resolutely for their own freedom offered a nonthreatening image of black manhood that allayed white fears of racial violence. It also accorded well with the ideal of the self-made man, an ideal espoused by many white abolitionists. Though they may have deplored the notion of compensation in principle, the logic of the antislavery argument itself compelled white antislavery activists to encourage and to contribute to the efforts of the enslaved to emancipate themselves in this way. In turn, such former slaves became powerful advertisements for abolitionist arguments about African-American character and the essential justice of emancipation. White abolitionists were even more willing to contribute money to redeem a slave when the slave in question was female, and especially if the woman was young and light-skinned. In their campaign to build public support for abolition, antislavery writers frequently linked slaveowning with the sexual exploitation of enslaved women. Lurid comparisons between southern plantations and “oriental” harems constituted a powerful tool to mobilize antislavery public opinion. Eager to extract the maximum impact from these images (and for the most part oblivious to the racist implications of the strategy), antislavery writers emphasized that lightskinned or mulatto women were most at risk of sexual exploitation and therefore most in need of rescue. Fictional accounts such as Lydia Maria Child’s short story “The Quadroons” (1842) or slave narratives such as that of Louisa Picquet described the vulnerability and abuse of such women in terms that appealed directly to nineteenth-century ideals of feminine purity. As such they mandated a response. Picquet’s narrative was published in 1861 with the explicit intention of generating funds to redeem her mother. It begins with a description of Picquet as “a little above the medium height, easy and graceful in her manners, of fair complexion and rosy cheeks, with dark eyes, a flowing head of hair with no perceptible inclination to curl, and every appearance, at first view, of an accomplished white lady,” and continues with attestations of her church membership and exemplary character. However, the substance of the narrative consists of Picquet’s efforts to resist the advances of various owners. The account concludes with appeals to assist her in her efforts to ransom her mother (Picquet and Mattison 1861). Picquet’s narrative invoked the twin values of filial piety and female purity, thus questions of the probity of slave

210

MARGARET M. R. KELLOW

redemption could be put aside as white abolitionists were confronted with the spectacle of endangered (almost) white female virtue (Kellow 2007). A more extreme example of this dynamic can be seen in the series of slave “auctions” staged by Henry Ward Beecher, in his fashionable Plymouth Church in Brooklyn, to raise money to ransom a number of enslaved women. Beecher was already a flamboyant figure when he began bringing beautiful young slavewomen before his congregation in order to raise money for their redemption in 1856. As these young women, all of whom were light-skinned, some almost white, stood before the congregation, Beecher would exhort his audience to deliver them from moral as well as physical danger. In one instance, Beecher introduced a particularly beautiful young woman, and an observer recalled that Beecher indicated that the girl was “to be sold by her own father [White] to go South—for what purpose you can imagine when you see her.” At Beecher’s instruction, the girl, “dressed from head to foot in virginal white,” loosened her hair, which “fell in shining waves to the floor.” As she stood silently before the congregation, the plate was passed and the members of the congregation heaped money, “jewelry, diamonds, watches and chains” in the offering to secure her freedom. By playing on the sexual, racial, and emotional sensitivities of his audiences, Beecher raised thousands of dollars to rescue these young women from bondage (as quoted in Shaw 2000, 337). Gender considerations also underscored black ambivalence about paying for freedom. William Wells Brown asserted his free manhood when he rejected his former master’s offer to sell him his liberty. Flight made a clear statement about black men’s attitudes to their masters’ claims to ownership. However, there were good reasons why some chose to pay for their freedom rather than flee. In some instances, black men appear to have negotiated their freedom with their masters in an attempt to compel their masters to recognize them as equals. Repeatedly cheated out of his freedom, Moses Grandy chose to stay and fight rather than to flee. He sued his master to have the bargain the two had entered into recognized. Grandy’s actions demonstrated that he saw himself as his master’s equal and was attempting to force his master to treat him accordingly. Cases like Grandy’s also demonstrate that slaves bought their freedom to put a formal and definitive end to their masters’ control over them. The legal process of emancipation constituted public recognition that the individual was no longer a slave. Freedom papers represented concrete and irrevocable evidence that the master had surrendered his control over the enslaved individual. By the same token, freedom papers offered some protection from patrols and, after the Fugitive Slave Act of 1850, from kidnappers who might try and sell African Americans back into slavery.

ABOLITIONISTS DEBATE SLAVE REDEMPTION

211

Responses to the question of the redemption of enslaved African Americans varied over time and according to the political situation, but also according to whose rescue was being proposed. The aftermath of independence set some Americans to thinking about the conflict between slavery and the values of the American Revolution, and searching for ways in which the interests of slaveowners and the enslaved might be reconciled and slavery ended. In that context, solutions that involved compensation to slaveholders who manumitted their slaves seemed reasonable. However, once the demand for immediate abolition emerged, many antislavery activists rejected compensation. As the founding documents of the AASS and the furor over Douglass’s purchase make clear, by the 1830s and 1840s the formal position of most American abolitionists on the question of compensation to slaveholders had become fairly categorical. Slaveholding was denounced as a crime, and antislavery activists rejected any suggestion that the “criminals” should receive any compensation. Using the language of evangelical righteousness, abolitionists argued publicly and vociferously that it was inappropriate to bribe a sinner to cease his transgressions. Yet as the debate over Douglass’s ransom demonstrated, loud though their protests might be, abolitionists were neither unanimous nor entirely consistent in their objections to slave redemption. Throughout the period black families went to heroic lengths to raise the money to rescue their loved ones from bondage. Moreover, whether they condoned or denounced compensation, black and white antislavery activists actively involved themselves in the redemption of enslaved African Americans. Although they knew very well that there were good reasons not to do it, when the slave in question was someone known, or when the slave’s situation resonated with prevailing cultural norms, abolitionists, black and white, set their principles to one side and ransomed thousands of slaves out of bondage. As Lindley Coates acknowledged when he criticized Garrison’s support for the ransom of Frederick Douglass (Liberator, 12 July 1834): When a self-emancipated man has come soliciting money to ransom his wife and children still in bondage, we have had a hard struggle of doubt, whether it was right to give him assistance or not, but a feeling of humanity for the individual victims to be ransomed has overcome the sense of the wrong thereby done to the mass of the enslaved. When brought face to face with the real suffering engendered by slavery, few abolitionists found they could place righteousness before compassion. As in so many instances, ideological purity and humanity coexisted at best uneasily. Confronted with this dilemma, most abolitionists appear to have opted for humanity.

212

MARGARET M. R. KELLOW

Bibliography The Liberator, 1831–65, Boston. The North Star, 1847–49, Rochester, NY. Voice of the Fugitive, 1851–53, Sandwich, Canada West. Breen, Timothy H., and Steven Innes (1980), “Myne Owne Ground”: Race and Freedom on Virginia’s Eastern Shore, 1640–1676. New York: Oxford University Press. Brown, William Wells (1849), Narrative of William W. Brown, an American Slave. London: C. Gilpin, http://docsouth.unc.edu/brownw/brown.html. Finkenbine, Roy E. (1993), “Boston’s Black Churches: Institutional Centers of the Antislavery Movement,” in Donald M. Jacobs (ed.), Courage and Conscience: Black and White Abolitionists in Boston (Bloomington: Indiana University Press), pp. 169–190. Fladeland, Betty L. (1976), “Compensated Emancipation: A Rejected Alternative,” Journal of Southern History 42: 169–189. Grandy, Moses (1843), Narrative of the Life of Moses Grandy. . . . London: C. Gilpin, http://docsouth.unc.edu/grandy/grandy.html. Kelley, Edmond (1851), A Family Redeemed from Bondage. New Bedford, MA: by the author, http://docsouth.unc.edu/neh/Kelley/Kelley.html. Kellow, Margaret M. R. (2007), “The Oriental Imaginary: Constructions of Female Bondage in Women’s Antislavery Discourse,” in Steven Mintz and John Stauffer (eds.), The Problem of Evil: Slavery, Race, and the Ambiguities of American Reform (Amherst: University of Massachusetts Press). Lane, Lunsford (1842), Narrative of the Life of Lunsford Lane . . . the Redemption by Purchase of Himself and Family from Slavery. . . . Boston: J. G. Torrey. http://docsouth.unc.edu/lanelunsford/lane.html. Lebsock, Suzanne (1984), The Free Women of Petersburg: Status and Culture in a Southern Town, 1784–1860. New York: W. W. Norton and Company. Loveland, Anne C. (1966), “Evangelicalism and Immediate Emancipation in American Antislavery Thought,” Journal of Southern History 32: 172–188. Merrill, Walter, ed. (1973), The Letters of William Lloyd Garrison, vol. 3: No Union with Slave-Holders, 1841–1846. Cambridge, MA: Belknap Press. Picquet, Louisa, and Hiram Mattison (1861), Louisa Picquet, the Octoroon, or Inside Views of Southern Domestic Life. New York: by the author. http:// digilib.nypl.org/dynaweb/digs/wwm97258. Shaw, Wayne (2000), “The Plymouth Pulpit: Henry Ward Beecher’s Auction Block,” American Transcendental Quarterly 14: 335–343. Walker, David (1830), Appeal . . . to the Coloured Citizens of the World. Boston: by the author. Weeks, Stephen B. (1896), Southern Quakers and Slavery: A Study in Institutional History. Baltimore: Johns Hopkins University Press. Wood, Betty (1995), Women’s Work, Men’s Work: The Informal Slave Economies of Lowcountry Georgia. Athens, GA: University of Georgia Press.

Chapter Ten Frederick Douglass and the Politics of Slave Redemptions JOHN STAUFFER

Throughout his career as an abolitionist, Frederick Douglass’s most passionate desire was an immediate end to slavery. He was, like most other “immediatists,” a millennialist: freedom represented a new age and a sharp break from the sins of the past and linear chronology. Emancipation filled Douglass with extraordinary hope and possibility, and he worshipped it with unwavering faith. In fact his swift rise to fame, and his extraordinary productivity during his twenty-year tenure as an abolitionist, depended on his faith that emancipation, and thus a new world, was at hand. But Douglass was also comparatively pragmatic about exactly how emancipation would be achieved. As a member of William Lloyd Garrison’s American Anti-Slavery Society from 1841 to 1849, he espoused Garrison’s policy of nonresistance, which viewed government as corrupt and relied on nonviolence and moral suasion to end slavery. He sometimes argued, following the Society’s policy, that compensating slaveholders to emancipate their slaves amounted to complicity in the sin of slavery. Yet even as a Garrisonian nonresister, Douglass often championed instances of emancipation that involved payment to owners or acts of violence. In 1846 British sympathizers purchased his legal freedom from his master, for which he was grateful. He valorized in speeches and fiction the achievement of Madison Washington, who imitated “George Washington” in gaining liberty through controlled and heroic violence (Douglass 1979–92 [hereafter TFDP] 1:1 [1845], 68). And he continually pointed to his famous fight with Edward Covey as the turning point in his life as slave; it was through this act of violence that he became free “in fact,” if still a slave “in form” (Douglass [1855] 2003, [hereafter My Bondage], 140). After Douglass moved to Rochester, New York, a hotbed of political abolitionism, and became an independent newspaperman, his views of government and the means to emancipation began to change. He accepted the Constitution as an antislavery document, transformed his newspaper into a political abolitionist newspaper, and embraced violence

214

JOHN STAUFFER

as a means to end slavery in the event that peaceful efforts failed. Like most other political abolitionists, he viewed compensated emancipation as a practical alternative to violence. In an important sense, Douglass’s shift from nonresistance to political action reflected his changing conception about the nature of the millennium. As a Garrisonian abolitionist, he separated the sacred and profane aspects of the country; sought “disunion” from the South, slaveholders, and the proslavery government; and viewed the American Anti-Slavery Society as his utopia. “No union with the slaveholder” was his motto, as he noted in an 1845 speech (Douglass 1:1, 1971–72, [1845], 106). But as a political abolitionist, his vision of a sacred and sin-free society encompassed all of America. He acknowledged that the North was complicit in the sin of slavery; and he advocated intervention, whether through force, compensation, or negotiation in slave regions (Stauffer 2002, 19–20). For him and other political abolitionists, “unqualified” and immediate emancipation was not the same thing as “uncompensated” emancipation; and compensated emancipation did not mean complicity with sin, but was part of a larger effort to unify and purify the nation. Even as Douglass denied that humans could be property, he defended the purchase of himself by British sympathizers. While he was on a speaking tour in England in 1845, the Quaker Ellen Richardson arranged to buy his freedom. She raised money, contacted Thomas Auld, his legal owner, and Hugh Auld, Thomas’s brother and the Baltimore master for whom Douglass had labored before fleeing north. Following complicated negotiations, 150 pounds sterling was sent to the United States; Thomas Auld sold Douglass to Hugh Auld for $100; and Hugh, after receiving $711.66, registered a deed of manumission in the Baltimore County courthouse on 12 December 1846 (McFeely 1991, 137, 143–144). A number of Douglass’s antislavery friends were not pleased with this transaction. As Douglass put it, “they thought it a violation of antislavery principles, conceding the right of property in man, and a wasteful expenditure of money.” But Douglass (and Garrison, who defended him) justified the transaction by calling it “ransom,” or “money extorted by a robber”; and since he considered his liberty to be of more value than 150 pounds sterling, he “could not see either a violation of the laws of morality or of economy, in the transaction” (225). The ransom payment could have been avoided had he remained in England, he noted, where he would have lived comparatively free from racism and surrounded by his new British friends. In fact he seriously considered remaining in England, and began making plans to resettle. But he felt duty-bound “to labor and suffer with the oppressed in my native land.” In light of his notoriety, he knew that he would be continually exposed to arrest and recapture in

FREDERICK DOUGLASS AND SLAVE REDEMPTIONS

215

America were he to remain a fugitive. “I think the very best thing was done in letting Master Hugh have the hundred and fifty pounds sterling and leaving me free to return to my appropriate field of labor,” he concluded (225). Douglass acknowledged an uneasiness with the transaction, but it related to its nature, rather than to the principle of compensation. It was ransom money, paid to allay the risk of recapture, not as payment for his person, as he noted. But in one sense, the sum did relate to his person: “Had I been a private person, having no other relations or duties than those of a personal and family nature, I should never have consented to the payment of so large a sum” (225–226). A smaller sum paid for a “private person” was evidently an acceptable form of compensation as well. In his farewell speech to the British people, on 30 March 1847, he said that he would never have solicited friends to purchase his freedom or raise money for his freedom. That would have been improper. But given that it was done “from the prompting or suggestion of their [his friends’] own hearts, entirely independent of myself,” he was deeply grateful (TFDP 1:2 [1847], 43). As a result, he did not have to shoulder “the responsibility of the act,” and so believed that “no right or noble principle” was “sacrificed in the transaction” (TFDP 1:2 [1847], 43–44). Having someone else pay ransom money to purchase your freedom was ethically acceptable; doing it yourself was not. Douglass had a similar problem when ex-slaves purchased their own freedom or those of family members. In 1849 he reprinted in his newspaper the story of an ex-slave, John Douglass, who raised money to purchase the freedom of his mother (for $200 down plus $100 in future payments). In an editorial Douglass chided the man for his actions: “I do not, in any shape or manner, approve or wish to have anything to do with the purchase of slaves,” he wrote, apparently ignoring his own purchase (Douglass 1849). But then he qualified his denunciation of slave purchases: “I cannot allow the right of any human being to claim payment for permitting any other human being to go where he or she desires to go. It is a natural right of every man and woman in God’s universe” (Douglass 1849). It came down to a matter of power; a slave (or ex-slave) was powerless in negotiating for his or his family’s freedom. But friends seeking to change the legal status of a fugitive, who was already free in form, could negotiate from a position of strength. Douglass emphasized the unnatural power imbalance between master and slave, and the natural, God-given equity between people, by stating: “This woman had just as much right” to ask her slave master “for payment that he might be allowed to go where he would, as he had to make this demand upon her.” “Besides,” he added, “every act of purchase enhances the market value of human chattles; and makes the monsters cling to their property with a

216

JOHN STAUFFER

more tenacious grasp” (Douglass 1849). The market value of slaves increased because of the power imbalance between the buyers and sellers. Douglass sought parity in human relationships, and so told slaves to assert what was theirs (Douglass 1849): I would simply advise every man, woman and child, whether black, white, or yellow, that so long as they are guiltless of crime, they have a right to their freedom. It is theirs. The idea of making them pay for what is their own by the inalienable gift of their Creator, is most absurd, preposterous, and Heaven insulting. Let them take what is theirs quietly and peaceably; and God help them to a good use of it. By the 1850s he was encouraging slaves to use force if necessary, even if it meant killing their master in gaining their freedom. “If a slave steals, he takes his own,” he wrote in My Bondage and My Freedom; “if he kills his master, he imitates only the heroes of the revolution,” who fought for freedom and equality (104–105). Douglass unambiguously endorsed compensated emancipation when the money came from people (or an institution) with power. In the same 1847 farewell speech to the British people in which he described his freedom being purchased, he referred to an offer by Arthur Tappan and Gerrit Smith, two wealthy political abolitionists, to contribute $10,000 to help “poor slaveholders” emancipate their slaves (TFDP 1:2 [1847], 41). Smith and Tappan had apparently advertised their offer to slaveholders, and according to Douglass, slaveholders “must have seen this advertisement, for whatever difficulties they have to encounter, they find none in seeing money.” But “was there ever a demand for a single red copper of the whole of those 10,000 dollars? Never; never” (TFDP 1:2 [1847], 41). When people in power, whether Smith and Tappan or statesmen, proposed compensated emancipation, slaveowners rejected the offers out of hand and tenaciously clung to their institution. For masters, owning property in humans was less a matter of money than of identity: owning a slave defined the essence of the master; it made him a “demigod,” to use Douglass’s term, and gave him power, honor, and the means to rise in his world (10, 302). As Douglass knew, and as historians later affirmed, those most vehemently opposed to offers of compensated emancipation were slaveowners themselves. It threatened their identity and way of life (Fladeland 1976, 178, 182). Although Douglass had not yet met Gerrit Smith when he referred to him in his 1847 speech, within a few years the two men would forge a close friendship. Gerrit Smith helped found the Liberty party in 1840 and the National Liberty party in 1848, and was himself a firm believer in compensated emancipation. Smith helped convince Douglass of the efficacy of compensated emancipation, as well as of political abolitionism.

FREDERICK DOUGLASS AND SLAVE REDEMPTIONS

217

By the late 1840s, Douglass began to interpret the Constitution not as a proslavery document, as Garrisonians and slaveowners did, but as a sacred antislavery text. He changed the name of his newspaper to Frederick Douglass’ Paper to reflect his newfound faith in political action; Smith helped fund it and suggested its name. Gerrit Smith was one of the wealthiest men in the country, and gave most of his money to the poor and downtrodden. He purchased the liberty of scores of slaves, paid their way to free states, and helped them resettle (often in his village of Peterboro, New York) and find employment. Through such acts he transformed his farm community of Madison County, New York, into a model interracial community. And in 1846 he undertook one of the largest philanthropic acts in the nineteenth century: he gave away 120,000 acres to some 3000 poor, landless New York State blacks, amounting to forty acres a person, which represented a rehearsal for Reconstruction and anticipated the plea among freedmen and women for “40 acres and a mule” to start a new life (Stauffer 2002, 144). Smith’s land was in the Adirondacks, in Franklin and Essex counties, and partly because neither Smith nor the recipients of his gift had the cash to buy mules and supplies to start life anew as independent farmers, only about eighty families settled at “Timbucto,” as they called their settlement (Stauffer 2002, 141). But John Brown settled there with his family; it was there where he conceived his scheme to invade the South; and he considered Timbucto his permanent home from 1848 until his death at Harpers Ferry. Douglass lauded Smith’s gift and urged recipients to save their money for the move. For both men, Timbucto was a model interracial community and a form of compensated emancipation—for oppressed and poor blacks, rather than racist oppressors. In 1855, Smith and Douglass applied this vision of compensation for the oppressed to a new political party they helped found: the Radical Abolition party, which advocated the redistribution of land so that “no man” would be “rich” and “no man” “poor,” as Smith put it (Stauffer 2002, 137). They viewed such redistribution of land as part of their larger efforts to end all evil; in fact the platform of the Radical Abolition party advocated an immediate end to slavery, through negotiation if possible, force if necessary, as well as land for all ex-slaves. Throughout the 1850s, Douglass and Smith understood the need to negotiate with the South from a position of force—whether through monetary enticements to end slavery, or violent intervention. Smith’s numerous purchases of slaves and his gifts of land to poor blacks led to more formal proposals for compensated emancipation. With Douglass stumping for him, he was elected to Congress in 1852, and in his April 1854 speech on the Kansas-Nebraska bill, which repealed the Missouri Compromise and opened northern territories to slavery, he

218

JOHN STAUFFER

lauded England’s Emancipation Act of 1834, which freed 800,000 slaves in the British West Indies and paid over 20 million pounds sterling (roughly $100 million) in compensation to slaveholders. Smith (1856, 206) suggested a similar form of compensated emancipation in the United States: The English people gave to the masters of eight hundred thousand slaves a hundred millions of dollars. I would, that the American people, after they shall have abolished American slavery, might give to the masters of four times that number of slaves four times the hundred of millions of dollars; and far more, would I that they should provide liberally for the humbler and cheaper, but infinitely more sacred, needs of the emancipated. Douglass reprinted Smith’s speech in its entirety in his newspaper, and pronounced it “the mightiest and grandest production ever before delivered in the House or Senate of this nation” (Douglass 1950, 2: 281). Three years later, in 1857, Smith proposed a more detailed plan of compensated emancipation, which he hoped would preempt the use of violence to end slavery. At the National Compensation Convention in Cleveland, organized by the pacifist Elihu Burritt, Smith suggested a formula for peaceful and immediate emancipation: each master would receive for each slave $150 from the national treasury and $75 from the emancipating state; each slave would receive $25 plus a plot of land. Smith assumed that the $525 million needed for his scheme to work could be raised by having the federal government sell off its remaining public lands. He attacked Burritt’s plan (drawn from earlier proposals by Rufus King and Daniel Webster), which appropriated public land to pay off slaveholders but provided no remuneration to slaves. “I am a landreformer, and I hold that to the landless belongs the vacant land,” Smith asserted (Smith 1857). For Smith, ending land monopoly was “the only sure way to abolish and prevent the return of” slavery. His plan was in one respect analogous to his 1846 gift of land at Timbucto, but extended throughout the country: in each case he sought to secure blacks’ freedom through small plots of land (Stauffer 2002, 143). Douglass wanted to attend the convention, “simply as a listener and reporter,” as he told Smith, but with some of his employees away he was shorthanded and needed to remain at Rochester to get out his paper (Douglass 1857, August 18). He reprinted reports of the National Compensation Convention, emphasizing that the South universally ridiculed the idea of compensated emancipation, while also noting the impracticality of such a plan. The South, he knew, would never go for it. Douglass also worried that focusing on compensated emancipation might shift the question of emancipation “too much from the conscience to the counter” (TFDP 1:3 [1860], 332). But he lauded Smith’s efforts at the convention,

FREDERICK DOUGLASS AND SLAVE REDEMPTIONS

219

and urged him to reply to William Lloyd Garrison’s derisive attack of Smith’s proposal. “Do write a reply to [Garrison’s] most unfair treatment of your speech, and send it to the Liberator,” Douglass urged his friend (Douglass 1857, October 13). Douglass’s greatest support for the idea of compensated emancipation was expressed in his speeches commemorating British West Indian emancipation. American blacks began celebrating West Indian emancipation on August 1, 1834, the day legal slavery ended in the British West Indies; and soon August 1 became a sacred day that rivaled the Fourth of July (TFDP 1:3 [1858], 227). By the 1840s, August 1 celebrations had become integrated affairs, though typically with more blacks than whites in attendance (Jeffrey 2006). From 1847 until his death in 1895 Douglass gave over twenty speeches celebrating West Indian emancipation. For most of his life he viewed the event as the “greatest and grandest of the nineteenth century,” as he put it in 1847 (TFDP 1:2 [1847], 69). Ten years later he said much the same thing: West Indian emancipation was “of vast and sublime significance, surpassing all power of exaggeration” (TFDP 1:3 [1857], 191). It was “the most interesting and sublime event of the nineteenth century” (194): It was the triumph of a great moral principle, a decisive victory, after a severe and protracted struggle, of freedom over slavery; of justice and mercy against a grim and bloody system of devilish brutality. It was an acknowledgement by a great nation of the sacredness of humanity, as against the claims of power and cupidity. He saw in British emancipation a typology or foreshadowing of the American millennium, a kind of Second Coming on a national scale. “There was something Godlike in this decree of the British nation. It was the spirit of the Son of God commanding the devil of slavery to go out of the British West Indies” (TFDP 1:3 [1857], 190). Nowhere was Douglass so optimistic about the new age than in his August 1 speeches, which are among his best: “We live in stirring times, and amid thrilling events,” he said in his celebration speech of 1848, the same year that France abolished slavery in the French West Indies (TFDP 1:2 [1848], 135): There is no telling what a day may bring forth. The human mind is everywhere filled with expectation. The moral sky is studded with signs and wonder. . . . Liberty rides as on a chariot of fire. . . . The grand conflict of the angel Liberty with the monster Slavery has at last come. The globe shakes with the contest. I thank God that I am permitted, with you, to live in these days, and to participate humbly, in this struggle.

220

JOHN STAUFFER

And nowhere, until after the Civil War, did Douglass criticize England for compensating slaveowners as part of its Emancipation Act (TFDP 1:4 [1869], 222). In fact, from his first August 1 speech until 1860, Douglass sought to deflect any criticism against England’s “sublime” act. In his August 1 speech in 1857, he referred to objections to August 1 celebrations by his friend James McCune Smith, the black physician and intellectual. McCune Smith opposed Englands’ compensated emancipation and argued that African Americans should “do something ourselves worthy of celebration, and not be everlastingly celebrating the deeds of a race by which we are despised” (TFDP 1:3 [1857], 200). But blacks were partly responsible for British emancipation, Douglass countered: “a share of the credit of the result falls justly to the slaves themselves. ‘Though slaves, they were rebellious slaves.’ They bore themselves well. They did not hug their chains, but according to their opportunities, swelled the general protest against oppression.” While white abolitionists sought to achieve emancipation through their eloquence and petitions, the slaves endeavored to gain their freedom “by outbreaks and violence. The combined action of one and the other wrought out the final result” (TFDP 1:3 [1857], 207). The first time Douglass referred to England compensating slaveowners was in 1860, and he described the payment as part of England’s sublime humanitarianism. “The British public, though weighed down and staggering under a heavy weight of taxation, bore, without a murmur, the additional burden of twenty millions sterling. If there was any complaint at all, it was that the masters got it instead of the slaves” (TFDP 1:3 [1860], 369). Douglass’s actual complaint, and “the contrast in respect to slavery,” was “between England and America, the mother and the daughter” (TFDP 1:3 [1860], 369). While England was “sublime” in emancipating its slaves, America was steeped in materialism and immorality. This, not the means by which emancipation occurred, is the key issue in Douglass’s August 1 speeches. In the face of unprecedented material prosperity in America, moral decay was everywhere. These material achievements “sink to nothingness” compared to Britain’s emancipation decree (TFDP 1:3 [1857], 193). But no “such responsive note of rejoicing” over Britain’s example has occurred in America, “except from a part of the colored people and their few white friends. As a nation, we are deaf, dumb, and blind to the moral beauty, and transcendent sublimity of West India Emancipation” (TFDP 1:3 [1857], 196). Why? Because “out of the fullness of our dollar-loving hearts, we have asked with owl-like wisdom, will it pay? Will it increase the growth of sugar? Will it cheapen tobacco? Will it increase the imports and exports of the Islands?” (TFDP 1:3 [1857], 196–197). In America, Douglass concluded, “money is the measure of morality, and the success

FREDERICK DOUGLASS AND SLAVE REDEMPTIONS

221

or failure of slavery, as a money-making system, determines with many whether the thing is virtuous or villainous, and whether it should be maintained or abolished” (TFDP 1:3 [1857], 198). Douglass knew as early as the late 1840s that some form of compensation would be needed before slavery could end in America. It would take the form either of ransom money, or of blood money, as he later said. In England, emancipation was bought “with money,” in America, “with blood” (TFDP 1:5 [1885], 196). War was politics by other means, which required different forms of compensation. But Douglass believed that in both events, the main impetus for emancipation came not from offers of compensation, but from God, with abolitionists bearing witness to God’s truth. As Douglass put it: “had slavery been abolished simply by the sword, it would have revived as soon as the sword was returned to its scappard” (TFDP 1:5 [1885], 199). What he never acknowledged was that a peaceful, compensated emancipation would have been far preferable to the apocalypse of war that brought immediate emancipation, even though southern states consistently rejected such proposals. After losing the war to end slavery, white southerners redeemed themselves by retaliating against blacks and preserving the old hierarchy. The century of horrible racism and racial oppression following the war stemmed in part from the violence that brought slavery to an end. Far more than their counterparts in the British West Indies, American freedmen and women lived in a state of “war” and were continually subjected to “terror, intimidation, and violence” (Litwack 1998, xi, xiii). Slavery was revived, though in extralegal forms, “as soon as the sword was returned to its scappard.” Bibliography Douglass, Frederick ([1845] 2003), Narrative of the Life of Frederick Douglass, An American Slave, Written by Himself, edited by David W. Blight. Boston: Bedford/ St. Martins. Douglass, Frederick (1849), “Buying a Mother’s Freedom,” North Star, July 13. Douglass, Frederick ([1855] 2003), My Bondage and My Freedom, edited by John Stauffer. New York: The Modern Library. Douglass, Frederick (1857), Letters, Frederick Douglass to Gerrit Smith, 18 August and 13 October 1857. The Gerrit Smith Papers, Syracuse University (and on microfilm). Douglass, Frederick (1950), The Life and Writings of Frederick Douglass, vols. 1–5, edited by Philip S. Foner. New York: International Publishers. Douglass, Frederick (1979–92), The Frederick Douglass Papers (TFDP), series 1, vols. 1–5, edited by John Blassingame. New Haven: Yale University Press. Fladeland, Betty L. (1976), “Compensated Emancipation: A Rejected Alternative,” Journal of Southern History, 42(2): 169–186.

222

JOHN STAUFFER

Jeffrey, Julie Roy (2006), “‘No Occurrence in Human History Is More Deserving of Commemoration Than This’: Abolitionist Celebrations of Freedom,” in Prophets of Protest: Reconsidering the History of American Abolitionism, edited by Timothy Patrick McCarthy and John Stauffer. New York: The New Press, 200–19. Litwack, Leon F. (1998), Trouble in Mind: Black Southerners in the Age of Jim Crow. New York: Alfred A. Knopf. McCarthy and John Stauffer (eds.), Prophets of Protest: New Essays on American Abolitionism (New York: The New Press), pp. 209–219. McFeely, William S. (1991), Frederick Douglass. New York: W.W. Norton. Smith, Gerrit (1856), Speeches of Gerrit Smith in Congress. New York: Mason Brothers. Smith, Gerrit (1857), Compensated Emancipation. A Speech by Gerrit Smith, In the National Compensation Convention, held in Cleveland, Ohio, August 25, 26, and 27, 1857. Broadside. The Gerrit Smith Papers, Syracuse University (and on microfilm). Stauffer, John (2002), The Black Hearts of Men: Radical Abolitionists and the Transformation of Race. Cambridge, MA: Harvard University Press.

Chapter Eleven The Moral Quandary of Slave Redemption HOWARD M C GARY

Slave redemption programs invoke feelings of discomfort in some people. And this discomfort sometimes continues to exist even in people who ultimately endorse such efforts. Some of these people are troubled because they question whether the good outweighs the bad. A number of papers in this volume address this concern. However, the source of this discomfort can also be attributed to the belief that it is wrong to always order the overall value of the states of affairs produced by alternative actions, and on the basis of these results, make a determination about what we ought to do. These critics, who have been labeled ethical deontologists, believe that it is morally wrong, at least in some instances, to bring about the best overall states of affairs based on prior rankings of value. According to the deontologist, there are certain duties and correlated rights that are absolute and should not be violated even if doing so produces the most overall good. The philosopher Immanuel Kant is most often associated with this way of thinking. The strict deontological way of looking at things draws a sharp distinction between the right and the good, and gives priority to the right over the good. Consequentialists or teleologists, on the other hand, derive their conceptions of right from the good. For them, the right thing to do is determined by promoting or maximizing overall good. For the deontologist, it is our motives for acting rather than the outcomes of our actions that should determine our duties. Doing things for the right reason is crucial for the deontologist. In the debate over slave redemption programs, the critics of such efforts don’t deny that in some cases the good is achieved by purchasing slaves. Nonetheless, these critics believe that by purchasing slaves the redeemers are giving incorrect priority to the good over the right. For them there are some things that should not be done, and by failing to stand against these things, we cannot be correctly described as moral persons. Is this a viable conclusion? I have given two reasons why some people object to slave redemption efforts. The first reason focuses on the doubts that some people have about whether such efforts will actually do more harm than good. The second reason relies on the idea that the buying and

226

HOWARD M C GARY

selling of human beings is always wrong, even if doing so in some instances promotes the greater good. It is the second reason that shall occupy our attention in the remainder of this paper.

Deontological Objections to Slave Redemption Efforts Above I said that the deontological cast of mind requires that people should stand for or against certain things. For the deontologist, when people are reasoning from the proper perspective (neutral, impartial, fair), they will see that we have a categorical duty not to do certain things. Kant ([1788] 1949) famously argued that telling a lie was never morally permissible. Is purchasing the freedom of another human being categorically impermissible? At first glance, it does not appear to be so. But perhaps if we dig deeper, we will be able to identify a principle that supports such a strong categorical stance. In his celebrated account of justice, John Rawls (1971) tells us that rational, fair, and impartial people would not agree to principles of justice that justified slavery even if the greater good were achieved by doing so. Rawls reaches this conclusion by constructing a decision procedure that he claims rational and fair people would submit to for making important moral judgments under conditions of uncertainty. Rawls does not say that slavery is absolutely wrong in a transcendental sense, but he does say that given a choice between his principles of justice which prevent slavery from being just, and consequentialist principles like the principle of utility that justify slavery under certain conditions, his principles would emerge as the principles of justice. Rawls’s argument relies on the viability of his decision procedure. The pure procedure justice (a procedure which says that we don’t know in advance what justice is, but we have a fair and impartial procedure that will produce it if the procedure is followed) that Rawls adopts is controversial, and, as a consequence, everyone does not accept his categorical rejection of slavery. Some people would still argue that the numbers should count when we decide whether we have a duty to reject the practice of slavery. However, even if we accept Rawls’s view that slavery is always unjust, this does not mean that it is unjust for people to purchase the freedom of slaves. Given the strong moral prohibition against slavery, it would seem that purchasing the freedom of slaves would be the just thing to do absent more viable ways of freeing them. The people who are buying slaves are not purchasing them to hold them as property, but to give them something that is rightfully theirs—their freedom. But at the same time there is another deontological consideration that pulls in the opposite direction.

THE MORAL QUANDARY OF SLAVE REDEMPTION

227

To free slaves by purchasing them is to participate in the commodifying of human beings, which deontological principles plausibly prohibit. But that would be to embrace an extraordinarily strong prohibition that might rule out cases in which we believe that it can be morally appropriate and even required to calculate how much a person’s life or limbs are worth because of an accident or negligence. The model of slavery that is often referred to when people discuss slavery is American chattel slavery (McGary and Lawson 1992, 8–9). However, this form of slavery differs from other forms of slavery (e.g., slavery in ancient Greece) in an important way. Chattel slaves were viewed as objects or things that could be bought or sold or used in any other way that their slaveowners saw fit. The slave owners thought that these slaves had no right to have their desires and wishes acknowledged, let alone respected, by them. Any limitations placed on the slaveowner’s treatment of slaves was not because of any rights the slaves possessed, but because of how such treatment was perceived by slave owners to effect their interests. Viewing chattel slaves as property is quite different from simply placing a value or a monetary value on human life. Some tort lawyers make handsome livings by determining how much compensation should be given for the loss of life and limb. However, in doing so, they are not viewing their clients as property or things without rights or valuable interests. In a letter to Henry C. Wright in 1846, Frederick Douglass (1999, 51) gives the following characterization of the slave redemption efforts of fellow abolitionists: Every man has a natural and inalienable right to himself. The inference from this is, “that man cannot hold property in man”—and as man cannot hold property in man, neither can Hugh Auld nor the United States have any right of property in me, they have no right to sell me— and, having no right to sell me, no one has a right to buy me. According to Douglass, the best reason for thinking that abolitionists should not purchase the freedom of slaves is that you cannot rightfully sell something that you don’t own. And, given that slaves are human beings, slaveowners cannot sell something they don’t rightfully own. On the other side of the equation, abolitionists cannot rightfully purchase slaves, because they cannot buy something from someone who does not have the right to sell it. Douglass has given us a convincing argument for the conclusion that purchasing slaves should not be seen as a market exchange. These considerations force us to re-conceptualize our problem. Is it right to think of redemption as involving slaves being bought? Perhaps a better way of describing what the slave redeemers are doing would be ransoming slaves rather than purchasing them. Suppose we view the slave redemption efforts in the Sudan as a kind of ransom. Should this

228

HOWARD M C GARY

change the deontologist’s thinking about the moral permissibility of paying money to redeem slaves when doing so promotes the greater good? Remember that the strong deontologist believes that we should never do so. But if we view the redemption as a ransom, we are not purchasing human beings; we are paying a ransom to free them from their unjust conditions. But ransom is a form of extortion. As such, those who oppose coercion have a good reason for thinking that ransoms are generally impermissible. From the consequentialist perspective, this is true because most of the time paying ransoms bring about bad consequences. However, in cases where this is not so, the consequentialist has no objection to ransoms. Does our deontologist have a good argument for opposing slave redemptions as a type of ransom that brings about good consequences?

A Deontological Argument against Slave Ransoms A prime candidate for a principle that a deontologist might employ to reject slave ransoms is the Never Reward Someone for Engaging in Intentional Wrongful Actions (NRW) principle. A deontologist might claim that the NRW principle should be followed without appeal to the consequences of doing so. The deontologist can admit that violating the principle might, in some instances, promote better consequences than not doing so for the individual or society at large, but deny that this option should ever be taken. It is important to note that the deontologist is not going to be persuaded by the numbers. Or, put in another way, the deontologist denies that showing things such as that ransoming of slaves relieves pain and suffering is sufficient for overriding their commitment to this principle. But why should there be strict adherence to the NRW principle? There are obvious advantages that result from preventing people from benefiting from their wrongful actions. Doing otherwise would allow individuals to have a material reason for not doing the right thing. And, if we generalize this point, we will see that many people could embrace this motive. Clearly our current legal system and our folk morality want to discourage such a motive. This is why both of these systems embrace compensation and punishment principles as appropriate responses to wrongdoing. An important function of both of these principles is to take the reward out of illegality and wrongdoing. On the level of legal and moral psychology, legal and moral systems also want to encourage emotions like contempt, guilt, and shame as appropriate reactions to illegality and wrongdoing. Absent these reactions, people who might be tempted to do wrong may have a greater inclination to do so.

THE MORAL QUANDARY OF SLAVE REDEMPTION

229

The intuition that I am drawing on here is the belief that it is never morally permissible to reward people for intentional wrongdoing. Of course, there are often good consequentialist reasons for following this principle. However, my point is not to give those reasons. My intuition is that we have reason to follow this principle even when the consequences of not doing so outweigh those for doing so. Obviously, some people disagree. Even amongst people who believe that certain things (like lying) should be forbidden, there is strong disagreement over the method or principles that should be used to identify an actual list of duties and rights. A very common method is one that I have employed. It relies on the moral intuitions of the parties involved in the moral disputes. Moral intuitions are clearly valuable, but this method for deciding what we ought to do has not withstood close scrutiny. The discomfort with relying on moral intuitions has not led philosophers to reject them, but instead to try to find a way to refine them. John Rawls developed a method to specify the content of a deontological theory by refining our moral intuitions. He called it the method of reflective equilibrium (1971, 48–51). The method of reflective equilibrium involves working back and forth between our considered moral judgments about a particular case, the principles that apply to the case, and the reasons that bear on accepting these principles, and revising our cases and principles whenever necessary to achieve coherence. This, of course, is only one method that has been offered for refining our moral intuitions. Kant, of course, provided us with another method. However, the methods offered by Kant and Rawls have not convinced consequentialists. Slave redemptions don’t simply allow wrongdoers to avoid punishment and the payment of compensation for their wrongful acts; they also allow wrongdoers to be paid for their vices! This fact provides potential wrongdoers with an additional motive for failing to do what they ought to do. We could attempt to avoid these outcomes by embracing the Platonic view that “justice is its own reward.” But, unfortunately, this response is not available to the consequentialist because she is more likely than not to adopt the incentives or sanctions models of moral motivation. On these models, people do what they ought to do because they want to be rewarded for doing so or they want to avoid the negative consequences of their moral failings. Since this is the case, consequentialists must be concerned about any program that provides people with additional motives for doing wrong. One might think that the NRW principle should never be violated because it always respects rights. And since deontological moral systems define rights in nonderivative terms, we have a nonconsequentialist reason for not violating the principle. However, with cases of slave redemptions

230

HOWARD M C GARY

in the Sudan, the person who pays to redeem slaves does not violate anyone’s rights. The slaveowners don’t have a right to hold slaves, and the person who pays the ransoms is not violating anyone’s rights in doing so. This is quite different from a case where I steal someone’s money to redeem slaves or I spend the money that I owe to others without their permission in order to buy the freedom of slaves. Our case in the Sudan is also quite different from a case where a person has placed herself voluntarily under harsh conditions of servitude for money and I defraud her master by manipulating him into accepting less for her contract than it is worth. Our case is also very different from Frederick Douglass’s manumission. In Douglass’s case his slaverowner, Thomas Auld, had a property right in Douglass. Auld had a legal right to do what was morally wrong. In the Sudan, slavery is illegal, but the government does not have the will or resources to prevent it. Slaveowners in the Sudan don’t have a moral or legal right to hold slaves. In Douglass’s case, he felt that he had to show why the people who purchased his freedom should not be condemned for rewarding slaveowners for conduct that was very immoral but not illegal. Douglass (1999, 52) said: Take another case—I have had dealings with a man. I have owed him one hundred dollars, and I have paid it; I have lost the receipt. He comes upon me the second time for the money. I know, and he knows, he has no right to it; but he is a villain, and has me in his power. The law is with him, and against me. I must pay or be dragged to jail. I choose to pay the bill for a second time. To say I sanctioned his right to rob me, because I preferred to pay rather than go to jail, is to utter an absurdity, to which no sane man would give heed. The people who purchased Douglass’s freedom had no other viable legal options because of the legality of American chattel slavery. The people who purchase the freedom of slaves in the Sudan are not in the same predicament. They can petition the government and take other legal actions to get the Sudanese government to do what it is legally obligated to do. And, by not pressing the government to live up to its responsibilities, these people, although well intentioned, are not showing the appropriate regard for justice and the law. So maybe this is a reason why a deontologist should not purchase slaves in the Sudan. Are there other reasons why a deontologist might oppose this practice? Ethical theorists who embrace the deontological perspective usually regard impartiality as an essential component of morality. They recognize, however, that it is virtually impossible to act impartially when it comes to helping or hurting each person within a relevantly specified group. Given this recognition, theorists have concluded that in order to be a moral agent one must be impartial when it comes to following moral rules.

THE MORAL QUANDARY OF SLAVE REDEMPTION

231

If one does not violate these rules, then one follows them impartially in regard to everyone in the relevant group. How does this apply to the issue of slave redemption in the Sudan? A deontologist could argue that given that there is not enough money to buy the freedom of all the persons held as slaves, showing partiality by purchasing some slaves violates the principle of impartiality and the agent-neutral perspective held so dearly by deontologists. There are good reasons for thinking that the slave redeemers in the Sudan show partiality. For example, it has been alleged that Christian groups are more likely to redeem persons who are sympathetic to their religious views. And clearly, in the case of family members who redeem members of their family, there is partiality. Of course, one could try to argue that family loyalties and commitments don’t violate the principles of impartiality. Frederick Douglass raises this concern in the letter that I quoted from above. Douglass rejected the idea that he should purchase his freedom, but he thought it was permissible for friends to do so. Douglass (1999, 53) said: I am free to say, that, had I possessed one hundred and fifty pounds, I would have seen Hugh Auld kicking, before I would have given it to him. I would have waited until till the emergency came, and only given up the money when nothing else would do. But my friends thought it best to provide against the contingency; they acted on their own responsibility, and I am not disturbed about the results. In John Stauffer’s paper in this volume, he discusses Douglass’s concern with the belief that it is all right to have someone else pay ransom money to purchase your freedom, but unacceptable to do so yourself. According to Stauffer, Frederick Douglass also chided another ex-slave, John Douglass, for purchasing the freedom of his mother. The only sense that I can make of Frederick Douglass’s reluctance, or in some cases, straight-out rejection, of slaves purchasing their own freedom or the freedom of members of their own families, is something like an impartiality principle. In the Sudan, then, we want to insure that the people who are being purchased are not beneficiaries of partial treatment. In practice that might be harder to do than in theory. But, at least, it is not clear that purchasing slaves in the Sudan must violate the impartiality requirement. If it can be done in a way that does not, then we can’t use this as a general reason to forbid it, especially when it promotes good consequences.

Virtue Ethics and the Problem of Slave Redemptions Although I don’t have an apriori argument in hand, in the light of the above considerations, when all is said and done, I doubt that we can give

232

HOWARD M C GARY

convincing arguments in favor of taking a categorical stance against carefully crafted slave redemption programs in the Sudan. I don’t know what such a categorical duty would look like. However, I don’t think this fact would entail that the consequentialist approach is the right way to go. Perhaps there is another moral perspective that might be able to capture the discomfort people feel about slave redemption programs. So far we have considered the issue of slave redemption from the ethical perspective of duty and obligation. From this perspective, the dispute between deontologists and consequentialists is over how we should characterize a duty-filled life. But by looking at the ethical life in terms of duties and obligation, maybe both approaches are incapable of capturing what is really right or wrong with slave redemption programs. Elizabeth Anscombe (1958), Alasdair MacIntyre (1984), and Bernard Williams (1973) are well known for their challenges to the rights and duties model of moral thinking. Anscombe, for example, criticizes modern moral philosophy’s preoccupation with a “law conception of ethics.” By this, she means a conception of ethics that focuses mostly on duty and universally applicable principles. Anscombe urges moral philosophers to return to a way of thinking about moral philosophy that is found in the work of Aristotle. The philosophers who have taken up her challenge have been described as doing virtue ethics because virtue is given a central place in their understanding of morality. Even though this alternative approach has been labeled virtue ethics, it is misleading to say that all who describe themselves as virtue ethicists are spending their time clarifying and illustrating the role that virtues play in our moral lives. However, there is a substantial group of virtue ethicists who do believe in a type of ethical judgment that does not focus on duty and obligation. These judgments are called eudaimonistic judgments. Instead of duty and obligation, they focus on things like human excellence and on characterizing the good life. Philosophers who take this approach have been described as doing virtue ethics because they don’t take a duty or an obligation to follow certain moral rules as primary. The virtue ethicist focuses on the agent rather than actions or rules. This has led many people to see this approach as one that studies human character as a way of understanding a good life. Since a good person is not measured in terms of how faithful she is at following certain rules, we have to know a lot more about the agents in question before we can render a moral judgment. This is why many people who take this approach study paradigmatic individuals as a way of illustrating the nature of a good life. Of course, all who do virtue ethics don’t focus on character in a narrow way as something mental and internal and separate from its external manifestations. Some of them also believe that a thorough eudaimonist must do a critical evaluation of character as well as personal achievement.

THE MORAL QUANDARY OF SLAVE REDEMPTION

233

Virtues are paramount in this ethical perspective, but not the exclusive focus of someone doing virtue ethics. The virtues are usually seen as stable character traits (such as honesty, generosity, and courage) that generate good lives. The virtues are important, but the person doing virtue ethics is also concerned about understanding how vices and other shortcomings affect our lives. Although virtues are the important focus, we need to be clear here. Deontologists and consequentialists can also focus on virtues, but when they do they see them as stable dispositions to act according to duty. A true eudaimonist theory does not see dutifulness as a necessary condition for a good life. But as I alluded to above, there are varied accounts of virtue ethics. Some of these accounts are not eudaimonist. Perhaps the following example will help to illustrate the difference between theories that make duties primary and ones that do not. A common complaint today is that doctors know very little about their patients and that too many don’t care enough about their patients and are not prone to make sacrifices for them. Some people attribute the present state of the doctor-patient relationship to the understanding of this relationship in terms of rights and duties. The critics believe that seeing things in terms of rights and duties will encourage people to fall short of being the best that they can be. Of course, the supporters of understanding the doctor-patient relationship in terms of rights and duties would argue there are good reasons why this model has become so prevalent. The model is seen as necessary in order to protect both parties from abuse. Even doctors who see themselves as acting in the interest of their patients can inadvertently cause abuse. The issue of medical paternalism is taken to be an illustration of this fact and as a good reason for preserving the rights and duties model of the doctor-patient relationship. However, one group of virtue ethicists would respond that the rights and duties model in the case above fails to give a central place to virtues like care and concern for others, patience, and self-sacrifice. Writers like Annette Baier (1985, especially chapters 6 and 12) have claimed that there is a tendency to ignore these virtues, and by doing so, we are unable to capture a central aspect of what it means to be moral. Baier’s criticism is not restricted to theorists who focus on rights and duties. She is well aware that even some virtue ethicists may omit these virtues. They may do so because these virtues are usually exemplified by women and people without power and influence. The paternalism counterexample fails to appreciate that in the real world people are very often not in relationships where personal autonomy can be given the highest priority. Is there an analogous case for how our concerns about the ethics of slave redemption in the Sudan might best be understood? Would we fare better by not looking at this issue in terms of rights and duties?

234

HOWARD McGARY

What would it mean to do so? How would a person who exemplified the virtues act in such cases? Would these paradigmatic individuals purchase slaves? Is the purchasing of slaves the best way to demonstrate concern and respect for the less fortunate? Are there alternative ways to liberate slaves that are morally superior to purchasing their freedom, but we are only willing to take them because we are unwilling to make greater sacrifices? These are interesting questions that are clearly open for discussion. I think they force sensible people to rethink the dominant ways of thinking about moral issues in western societies. Maybe there are better ways to address these problems. Perhaps one of the ethical perspectives discussed above reduces to the other, or maybe it is wrong to think that either is primary. Answers to these questions, I believe, will bring us closer to feeling comfortable about what we should do or don’t do in the case of slave redemption in the Sudan.

Bibliography Anscombe, G.E.M. (1958), “Modern Moral Philosophy,” Philosophy 33: 1–19. Baier, Annette (1985), Postures of the Mind: Essays on Mind and Morals. Minneapolis: University of Minnesota Press. Douglass, Frederick (1999), “A Letter to Henry C. Wright, December 22, 1846,” in Frederick Douglass Selected Speeches and Writings, edited by Yuval Taylor and Philip S. Foner (Chicago: Lawrence Hill Books), pp. 49–54. Kant, Immanuel ([1788] 1949), “A Supposed Right to Lie from Altrustic Motives,” in The Critique of Practical Reason and Other Writings in Moral Philosophy, translated by Lewis White Beck (Chicago: University of Chicago Press), pp. 346–350. MacIntyre, Alasdair (1984), After Virtue. Notre Dame, IN: University of Notre Dame Press. McGary, Howard, and Bill E. Lawson (1992), Between Slavery and Freedom: Philosophy and American Slavery. Bloomington: Indiana University Press. Rawls, John (1971), A Theory of Justice. Cambridge, MA: Harvard University Press. Williams, Bernard (1973), Problems of the Self. Cambridge: Cambridge University Press.

Chapter Twelve The Next Best Thing MARTIN BUNZL

Let us assume that the arguments of Karlan and Krueger that appear in this volume are correct. That is to say, let us assume that redeeming the freedom of one slave (under most circumstances) does no harm and likely does at least some good. That is to say, let us assume that if I buy the freedom of 10 slaves, no more than 10 people will be newly enslaved and likely less than 10 will be enslaved. So, all other things being equal, netnet, let us assume that I do more good than harm. (All of these assumptions bracket deontic objections taken up by others in this volume.) Well, so what, Peter Singer might say—you can bring about a much greater good by giving your money to feed starving people. (In fact that is just what Singer has said [in conversation] about this issue.) I want to examine the force of this prescription in what follows. In other words, I am interested in the question of to what degree our actions in helping others should be constrained by the fact that we might be able to do more or better. I am going to be especially interested in focusing on the force of the phrase “might be able to” and in contrasting it to what we are likely to do. The core question I am interested in, specific to the subject of this volume, is this: at least when it comes to doing good, as in the act of freeing a slave, even if there are alternatives that would produce greater good, what kinds of factors might undermine the likelihood of those alternatives being realized? That is a quick and dirty version of the question that I am interested in. The more proper version is comparative: at least when it comes to doing good, as in the act of freeing a slave, even if there are alternatives that would produce greater good, what kind of factors might undermine the comparative likelihood of those alternatives being realized over the likelihood of freeing a slave? One more twist, for we ought to worry not just that the best alternative is less likely to be realized than the second best, but that that comparative likelihood is relatively hard to change: at least when it comes to doing good, as in the act of freeing a slave, even if there are alternatives that would produce greater good, how robust are the factors that might diminish the comparative likelihood of those alternatives being realized over the likelihood of freeing a slave?

236

MARTIN BUNZL

My motivation for these concerns is derived from purely anecdotal considerations: reports about the enthusiasm and interest in campaigns to free slaves. Being a philosopher, that is enough to get me worrying about the issue, since I am not shy about treating the concern as a hypothetical. Suppose people are more likely to be moved to give (or give more) to help free a slave than feed a starving person (notwithstanding the greater benefit of doing the latter over the former). You might wonder why these prescriptive worries arise in the first place. If buying the freedom of a slave is morally permissible, is it morally obligatory? If not, isn’t it just supererogatory, and if so, aren’t I free to exercise my supererogatory inclinations at will? Not necessarily, in three different ways. First, even if buying the freedom of a slave is just supererogatory, it does not follow that I am under no moral constraints. Consider a big shot who sets up a scholarship fund for whites only. Isn’t he free to give his money as he wishes? Not necessarily. Doing so may produce such profound resentment among those shut out, especially given the reasons why they are shut out, that his gift does more harm than good. Here we assumed that freeing the slave does no more harm than good. So this worry is ruled out. But that does not undermine the claim that such actions are not free of at least some of the constraints that arise from moral obligations. Suppose Peter Singer sets out to improve the world by prescribing, to those that wish to listen, courses of action that will produce the greatest good. Singer himself, and with him, many others, would say that helping is not necessarily a supererogatory act. It is not when it comes to our obligation to jump into a pond we walk by to save a drowning child. But further, if distance is not morally relevant, we are equally obligated to help children far away as well as those close by. That may not be practicable when it comes to saving the distant drowning. But put generally, we have duties of intervention and those are not sensitive to distance. Assume that freeing a slave produces at least as much good as harm and, all other things being equal, counts as one of those obligations. Still, whether it is so in a world of scarce resources and opportunity costs depends on whether there are other alternatives that are also obligatory that are mutually exclusive. If so, we are in business. There is a third, middle ground between the obligations to act and the supererogatory. As my colleague Larry Temkin would argue (2005), I may not be obliged to act to free a slave (at least one that I do not myself own) as a consequence of considerations of justice, but there still may be moral reasons that make it more than merely supererogatory. If there are such reasons, then we are in business this way as well via the conflicts that arise for mutually incompatible prima facie moral reasons that I ought to attend to. So on any of the three alternatives, you might say that, if I am to do good, that I ought to do the greatest good has some pull.

THE NEXT BEST THING

237

Let us assume that to be the case without worrying too much why it might be the case. Any of the three routes to that conclusion I have suggested above will do. But to stop here is to stop far short of the finishing line if our goal is to achieve action that instantiates such a “greatest good” general principle. Let us consider some of the ways in which a smooth and seamless transition from a greatest good prescription to action may break down. If we view decisions as being made on a case-by-case basis, the most obvious of these are epistemic and temporal. I may not be able to know which is the best alternative, and even if I were able to know it in principle, the time it would take to find out carries costs that need to be offset against the benefits. These are the standard objections that motivate the transformation of a greatest good principle from being a moral case-by-case guide into a moral guide for formulating rules to govern classes of cases—like, in general, greater good will come from feeding the starving than freeing slaves. But, on either variant, there is a traditional worry about whether a greatest good principle will in fact produce the greatest good that goes all the way back to Sidgwick. We would not need a moral code in the first place if we were not prone to act immorally or at least amorally. One who advocates a moral code against this background on the grounds that it will produce the greatest good, if followed, has to offset these benefits with the costs of implementing such a policy—costs which may be high depending on the state of nature that forms its backdrop. But that assumes success in transforming us whatever the costs. Sidgwick’s concern was what the consequences are if no full transformation is possible, if some of our natural inclinations are intractable and not fully reformable (Sidgwick 1907). Sidgwick’s worries arise from the tension between those we are close to and strangers. By his lights, promoting the general good and the relations of “affection” we have to those close to us will come into conflict. You might think that in response, “Utilitarianism must therefore prescribe such a culture of feelings as will, so far as possible, counteract this tendency” (Sidgwick 1907, 434). But the problem is whether we can do so without transforming such feelings into “a watery kindness” (as Sidwick points out, the phrase comes from Aristotle) when it is expressed in its universalized form. Better, Sidgwick thinks, to pay the price of giving up a degree of impartiality to harness feelings for the other in their full strength at least at the parochial level. I think this picture of thinking and feeling is at best only part of the story. Of course we have special feelings for those close to us as compared to strangers. But it is not as if strangers don’t evoke feelings in us as well. It is not that we lack a culture of feeling toward them. Worse, ignoring that emotional dimension of our relations to others feeds the illusion that

238

MARTIN BUNZL

thinking alone is all we need to worry about. Instead, we need to wonder about how and where feelings intrude; whether in doing so they act as a help or a hindrance; and if the latter, whether they are reformable. I want to consider these concerns for three aspects of our emotional lives that I think come into play in potentially subversive ways if our goal is to help the neediest with a view to maximizing good—namely: familiarity, distress, and empathy. You are to choose between giving $100 to free two slaves or feed four people for a year. Singer whispers in your ear, “Feed the hungry, you will do more good.” Set aside complex arguments about the facts—what will happen to the starving after your donation runs out? What is the probability of the slave being self-supporting after redemption? And so on. Instead, let us worry about how feelings might intrude. There are three ways. First, as an adjunct to reasoning; second, as a substitute for reasoning; and third, in the implementation of reasoning. For now let us just worry about the first two of these. Part of what drives my anecdotally based concerns about the widespread enthusiasm for slave redemption programs is the intuition that it is easier to “feel for” the enslaved than it is for the starving. And moreover, at least at first blush, these feelings bear no obvious relationship to desert. But whether my intuition is correct or not in this particular case does not really matter because the point I want to make is a much more general one: when it comes to choosing between the objects of our moral concern, it is possible to characterize at least some influences to which such feelings are subject. That is especially the case when it comes to familiarity and distress. As we will see, familiarity operates to increase feeling and distress operates to decrease it, yet neither bear a direct relationship to desert. Indeed, if anything, in the case of the latter it is even worse: the more distress-inducing case will likely be the more needy. But, be that as it may, in both the case of familiarity and distress, feelings can play a differentiating role in cases involving not just family and community versus strangers but cases involving choices between strangers alone. If that is obvious in the case of distress, it is quite surprising in the case of familiarity, indeed even seemingly contradictory, in the sense that strangers are surely those we are unfamiliar with by definition! However, William Kunst-Wilson and Robert Zajonc (1980) have shown than familiarity of the most innocuous nature can produce a preference bias. Subjects were exposed for one millisecond to each member of a set of 10 irregular octagons. The one-millisecond exposure length had previously been established to be too short to produce anything better than chance results in a recognition memory test in which the subjects were asked to discriminate the previously exposed stimuli from novel ones. However, when subjects were asked to make pair-wise preference judg-

THE NEXT BEST THING

239

ments between random pairings of the previously exposed stimuli and novel ones (both of which were irregular octagons), a statistically significant preference for the previously exposed stimuli was demonstrated. (Sixteen of 24 subjects preferred the old stimuli to the novel ones but only five of the 24 recognized them.) Kunst-Wilson and Zajonc used these data to argue that we have the capacity to make judgments noncognitively, and their research has become central to a debate about whether cognitive representation is a necessary condition for emotional responses to stimuli. The conclusion I want to draw is different and twofold: first, familiarity affects preferences. But more importantly, familiarity starts much further from home than we might think—even in the domain of what is, for all intents and purposes, the unfamiliar. As a result, any pairwise choice situation may be directly subject to preference effects without our even knowing it. Still, someone might object as follows—we are discussing familiarity as an adjunct to moral reasoning. Surely when I ask you to choose who is most deserving, A or B, that you may prefer A over B, even without knowing it, is not necessarily to say that your moral choice will be affected. But even if mere familiarity did not have such direct effects, it can also have indirect effects in a circumstance of moral choice as well in virtue of its effects on empathy. For it is a truism that empathy is shaped (in part) by preferences. And if mere familiarity can drive preferences, then it will do so as well when it comes to empathy. The news here is not that, but rather that if the effects of familiarity are so fine-grained that they can be exercised on choices between strangers, so too will they differentially affect our empathy for strangers. Of course feeling empathy for a stranger would be a gain to the extent that we view empathy as an unalloyed virtue. But, at least as an adjunct to moral reasoning, judgments of fairness turn out to be biased by empathy whether or not familiarity and its associated preferences bias such judgments directly. Traditionally, empathy has been viewed as a necessary condition on prosocial behavior. You walk down the street and see a stranger who needs your help. Standard social psychology analyzes this sort of prima facie other-directed behavior as a function of the perceived opportunity costs you face and the degree of empathy you feel. (For example, Sears et al. 1985.) It is for that reason (among others) that educational efforts to motivate and increase empathic responses are taken to serve a social good. But Batson and his colleagues (Batson et al. 1995) have shown that inducing empathy is a double-edged sword when it comes to determining whom to help as opposed to whether they will be helped. Batson and his colleagues deceived student subjects into thinking they were participating in an experiment to assess the consequences of positive versus negative consequences on workers (among other things). Students believed they

240

MARTIN BUNZL

were being randomly assigned to either a worker or supervisor role when in fact they were all being assigned the “supervisor” role. They were then led to believe that their job was to assign the “workers” to one of two groups—those receiving positive reinforcement (a gift certificate) for successful completion of an experimental task and those receiving negative reinforcement (mild electric shock) for unsuccessful completion of the same task. Subjects were instructed that “most supervisors feel that flipping a coin is the fairest way to assign workers to the task, but it is entirely up to you” (1995, 1044), and a coin was provided to them. However, two-thirds of the subjects were told that they would be receiving a confidential communication written by one of the people they were to assign. The content of the communication was unrelated to the assignment or task but rather talked about something sad that had recently happened in the writer’s life. Of these, half were told to read it but take an objective stance. The other half were instructed to read it and to try to “imagine how this student feels” (1995, 1044). So in summary, this experiment examined subjects’ assignments of others to positive and negative consequence groups with no communication, with communication with instructions designed to elicit low empathy, and with communication with instructions designed to elicit high empathy. Subjects in the low and high empathy groups were tested to establish that empathy had been elicited as intended. Results confirmed that it had. Comparing the assignments made and the methods used for the three groups: all of the “no communication” subjects used random assignments. In the “low empathy” group, 17 of the 20 subjects used random assignments, while 3 did not, favoring the person from whom they had received the communication. However, in the high empathy group, half the subjects favored the writer. (All results reported were statistically significant.) Interestingly, nearly all participants in the study viewed random assignment as the fair way to make assignments. Moreover, those who acted partially in making the assignments viewed themselves as having acted less morally than if they had used a random assignment process. These findings underscore how, short of universal empathy, such feelings can undercut a fair decision procedure even for subjects that are selfconsciously engaged in trying to make such decisions. Choosing between outcomes in terms of greatest good is not necessarily the same as deciding what is fair, but it too involves a deliberative process, and there is no reason to think Batson’s results would not apply here, too. Batson’s setup also nicely illustrates how easy it is to trigger empathy in a stranger and that might encourage the view that with some “juicing up” an expectation of universal empathy is an ideal to encourage which would result in something stronger than the weak “watery” kindness Sidgwick worried about. The problem, however, is not that general empathy

THE NEXT BEST THING

241

is realizable but weak, but rather that it is subject to catastrophic failure under circumstances when it would be most needed if it were to be relied on as an adjunct to moral reasoning. Before proceeding, readers may wish to go to: http://fas-philosophy .rutgers.edu/~bunzl/slavery.html. You will find some pictures at that site. Please view them in the context of a thought experiment in which you receive a direct mail solicitation or a solicitation on the street to support efforts to combat hunger in country x. You are told that these pictures illustrate the kind of population that will be served. Now click through to the pictures, viewing set 1 first and then set 2. In each case gauge your reactions. If you have decided to proceed as outlined above, don’t read on until you have done so. A word to those readers who elected not to go to the website above. The site contains two sets of pictures of undernourished children. The first set are of children who are clearly underweight, gaunt, and sadlooking. The second set are of children who are more dramatically emaciated and who appear to be near death. If you did look at the site, how did you react and how did you feel? Here is how I suspect you responded. The first set of pictures probably evoked feelings of sadness in you, and if not empathy (since the pictures are of children and you [presumably] are an adult), sympathy. That empathy or sympathy may have been followed by a feeling that you wished you could do something to help these sorts of children. In the case of the second set of pictures, I suspect you also felt sadness, even extreme sadness and sorrow. But at the same time, you probably felt anxiety and some distress. You may have literally sucked in air and physically recoiled from the screen, possibly averting your eyes. If you forced your gaze to linger, your feelings of distress likely increased and you may have been aware of a feeling of physical disgust. You may have had the same kind of feeling if you have ever seen pictures of a decaying corpse. If you had that feeling of disgust, you most likely will have broken your gaze and averted your eyes. Whether you felt distress and stopped looking or your feeling reached a level of disgust immediately or after looking for a while, I suspect you felt less empathy or sympathy for the second set of children than you did for the first even as you may have thought they were more deserving of your sympathy. If I am right, you also did not have a feeling that you wished you could do something to help these sorts of children— your level of distress and (if it happened) disgust reaction trumped that. If you had such a reaction, note how it distanced you from your connection to the pictures. You may not only have jolted back from the screen, but psychologically, you may have caught yourself thinking that you did not want to see these images, or thinking that I had a lot of nerve to have you look at them, or having a myriad of other self-focused thoughts.

242

MARTIN BUNZL

Distress and disgust functioned to break your connection with the needs of the children, replacing it with your own needs. Well, maybe you didn’t experience what I just outlined at all! (The author invites responses to [email protected].) Here are two possible reasons why you didn’t: seeing the first set of images, you may have thought that if the first were deserving of your help then so were the second, irrespective of your feelings. You may also have had an inkling about what was coming from earlier passages in this paper and been primed for the images. Such priming has been shown to produce a compensation effect (Berkowitz et al. 2000)—it is the same reason your dentist tells you a procedure is really going to hurt. But then again, my retrodictions may just be plain false! But if all of this makes empathy seem a dubious ally for moral reasoning, let alone a substitute for such reasoning, it isn’t as though we can ignore its vagaries—if only as a source of interference with reasoning. And if the traditional worry (going back to Sidgwick) is that such feelings intrude only close to home, here we have shown that its reach is much further than we might think. Now one way to react to this is to argue for a traditional rationalist line that we should eschew a Humean conception of morality as a slave of the passions and boldly assert the preeminence of reasoning over feeling where the two conflict. Of course that assumes we can discipline our feelings at least when it comes to reason. But we’d better not hope for too much success if we are interested in the probability of the conclusions of such reasoning being implemented at least on the standard model of altruism outlined above. For on that model empathy is a necessary condition for us to act altruistically as opposed to merely reasoning about it. That creates a new opportunity for empathy to undermine the outcome of reasoning—not directly, but instead by the selectivity by which we are responsive to its prescriptions. Let us assume this model of altruism is correct, without worrying too much about the alleged distinction between empathy and sympathy. How deferential should we be to empathy’s dictates? Should the moral prescriptivist hold his or her ground in the face empathically driven insurrection or settle for the next best thing? Even if we assume the model of altruism we are relying on is correct, we can’t begin to answer these questions without a sense of how reformable empathy is. I think there is reason to be pessimistic about the prospects for reforming empathy, given the kinds of factors that we have seen that can affect it. When it comes to disgust, we have a wealth of day-to-day experience that demonstrates how firmly entrenched such responses are and how unyielding they are to day-to-day exposure. Most of our responses don’t have this characteristic. Repeated exposure to the same stimulus engenders indifference over time. Not so when it comes to disgust-provoking

THE NEXT BEST THING

243

stimuli. It is hard to discuss this without some indelicacy—but think of your own experience when it comes to defecating. Like most people you probably examine the results of your efforts on a daily basis. And if you are like most people, your nose wrinkles every time you do so. One would think that after so many years of following this procedure, you would have gotten used to it. But if you are like me, the wrinkle of your nose, the recoil of your body is as robust today as it was when you were seven years old. And it is not just so with fecal matter. Other disgust reactions are remarkable for their hold on us even in the face of countervailing considerations of rationality. Take spit. When you kiss another, you don’t mind exchanging spit. Yet the idea of drinking a teaspoon of that same person’s spit is profoundly off-putting. These are examples of the hold that so-called basic emotions have on us. Maybe there is no such thing as a basic emotion, but even if there isn’t, disgust prompted by these sorts of emotions lacks much of the cultural overlay that disgust at the idea of having sex with a sibling presumably has. Disgust with such a cultural overlay also shows remarkable persistence. I hesitate to recommend you try sleeping with your siblings to see if it gets any less off-putting with repetition. Try this instead: find a dead cockroach and sterilize it. Now put it in your mouth. Repeat the effort once a day for a week and see how far you get. Or, if you prefer, go to your local hospital supply store and buy a new bedpan. Unwrap it and pour some wine into it. Drink the wine if you can. Repeat the procedure every day for a week and see if you feel any less reluctant about the exercise. (Most of the above is inspired by Paul Rozin’s work [see for example Rozin and Nemeroff 1990].) Even imagining some disgusting acts has quite robust consequences. Want to diminish your desire for bagels? Take a bagel out of your fridge and now just imagine a dead cockroach lying between the two halves. Now try to eat it and see how you feel. (Do you think you will feel an overwhelming need to check that there is no cockroach there before you eat it? And even if there is none there, will a feeling of nausea at the idea of its being there persist?) Of course some people do not have these reactions. Whether they don’t as a matter of accommodation or preexisting disposition is an open question. Perhaps they were never put off by these things in the first place. Or more likely, unlike me (and most likely, you) they are capable of becoming desensitized over the long run by a disposition that they have and we lack. But, be that as it may, our concern is with the role of disgust response for most of us, and most of us have disgust responses that are remarkable in their strength and persistence over time. When it comes to familiarity, things are much more complicated and murky. If the incidence of disgust is easy to identify but our reactions are hard to purge, the situation of familiarity at best runs in the other

244

MARTIN BUNZL

direction. Familiarity may be easier to override than disgust reactions (even if in doing so we overcompensate), but if Zajonc is right, the low threshold for it to be triggered means that it is not only pervasive in our lives but below our awareness most of the time, thus making it hard to reform. Moreover, there is one more turn of the screw to be made when it comes to certain kinds of unfamiliarity: both Hart et al. (2000) and Phelps et al. (2002) have found quite significant differences in the reactivity of the amygdalas of subjects presented with images of strangers differentiated by race. The amygdala is a part of the brain which is strongly implicated in many “emotional” reactions closely associated with fear. Thus it might not be surprising that a variant of such a reaction would be implicated when one is presented with images of strangers. All the more so, since, as Joseph Ledoux (1996) has argued, fear reactions in the amygdala are directly linked to the sensory thalamus as well as being linked via cortical areas. That allows for two routes to the amygdala—one faster than the other. The quicker route allows us to have a “reflexive” reaction to some stimuli which slower processing either amplifies or tamps down. Phelps et al. (2002) found significantly more amygdala activation when white subjects were presented with pictures of unfamiliar black faces (taken from a high school yearbook) as compared to unfamiliar white faces (from the same source). Interestingly, the difference is not found with images of famous people. There is no reason to think that this finding is not symmetric—that is, that black subjects would react to non-black versus black strangers in a similarly differential way. In fact, Hart et al. (2000) scanned the brains of both black and white subjects. If the amygdala reacts to the image of a stranger, it also eventually becomes desensitized to that image and its reactivity diminishes. That fact allows observers to measure the differential rate at which such desensitization takes place. Hart et al. found a slower rate of amygdala deactivation for subjects viewing different-race stranger images as opposed to same-race stranger images. Is this an instance of a deeply ingrained racial sensitivity? More likely, it is a deeply ingrained “us versus them” function, acquired through natural selection, that is cashed in in culturally local terms that, in our case, run along racial lines. But whatever its etiology, this phenomenon underscores the fact that familiarity operates in a much more complex way than a simple model of more exposure for familiarity might suggest, and in fact may be subject to quite deep-seated categorical distinctions that are just as hard to unlearn as disgust reactions are. (Mary Wheeler and Susan Fiske [2005] have reported results consistent with the above when subjects are viewing images with an assigned task of social categorization. However, they report the results evaporate

THE NEXT BEST THING

245

when subjects are viewing the images with an assigned task that calls for social individuation.) If the truly needy are unfamiliar (in the way discussed above) and disgust-provoking in their very neediness, then empathy comes under attack from both directions and its chances of prevailing in provoking action will suffer. But even bracketing these external disruptions to our empathic capacities, empathy itself contains the seeds of its own destruction. If you were like me in your reaction to the second set of images of starving children discussed above, both distress and disgust functioned to break your connection with the needs of the children, replacing it with your own needs. So far we have only discussed the disgust component. If altruism has empathy as a necessary condition, it is not just any empathy, but empathy for or with the distress of the other. I experience what you feel. My feelings match yours. Whether I respond to those feelings on your behalf or mine is a second question. On Martin Hoffman’s view (2000) my capacity to have a sympathetic reaction to such empathic distress eventually gets replaced by a self-directed reaction when the empathic distress reaches a certain level (which is what the images were designed to elicit). Empathy draws us to others, but if we get too drawn to them and their condition is too extreme, our other-directed response turns to a focus on ourselves and withdrawal from the other. Thus the greater your distress, and my empathic response to it, the greater the chance that I will react on my behalf, not yours. If that is right then, if the neediest are the most deserving (because helping them will do the greatest good), they will be more likely to prompt a self-directed response as compared to the less deserving. For distress reactions, unlike disgust reactions, there is no reason not to think that the tipping point from concern for the other to concern for oneself is quite varied and subject to many determinants. Be that as it may, with the exception of Mother Teresa, for most of us, moral prescription and empathic responses will part ways at some point. Here a moral prescriptivist would do well to think like a preacher. A preacher, at least one interested in producing good, ought to worry about whether his or her prescriptions fall within the realm of psychological possibility for his or her charges. That is not to say that there are not religious orders that render us always wanting. But a religion that makes the task of moral improvement hopelessly unattainable runs the risk of provoking the embrace of sin—after all, if we are doomed to be sinners, why not at least have fun along the way? Of course the danger, for a preacher or prescriptivist, or temptation (depending on your view) is to misjudge psychological possibility. If aiming too high makes us all and only sinners, aiming too low makes it too easy for us to be saints.

246

MARTIN BUNZL

Yet should we be concerned about possibility, or would we do better to rather think about probability? Not, obviously, the probability of the course of action I am likeliest to engage in, but the product of the probability of such an action and the good it produces. Should we be concerned with the fact that people are more likely to free a slave than feed a starving person even if the latter would do more good? Or should we worry about whether people are capable of being induced to do the latter over the former? That is to say, whether people’s probabilities of action for the outcome that will produce the greatest good can be increased? If you are interested in producing the greatest good, this choice turns out to be quite simple—but only if you have all the facts in hand. You or Singer will want to know the effect of your prescription on increasing the probability of my engaging in one action rather than another, along with the good resulting from those actions. But this is surely madness. We have reduced the role of a moral prescriptivist to a calculational opportunist. Instead of advocating what Singer believes to be the best course of action, we want him to advocate the best course of action in the light of what we are likely to do in the light of his advice—even if it is second best. It is one thing for the prescriptivist not to advocate a course of action that I am not capable of following, but another to specify a course of action for me based on the belief that specifying so will produce the best outcome in the long run. But why? The answer is that the second loses sight of my agency and treats me merely as an extension of your or Singer’s agency. On the other hand, prescribing with an eye on the range of my psychological capacities makes sense because those capacities act as limits on my agency. If that sounds like a plea for “ought” to respect the limits of “can,” there will be detractors because of some famous arguments against this view. But before we even worry about such arguments, we should dwell on the idea of capacity here. It is one thing to say, there is no point in advocating a course of action that would require me to be in two places at once because that is physically impossible. But why is it another to say there is no point in advocating a course of action that would require me to sacrifice my child to save a strange child? Or at least to toss a coin? Psychological possibility comes in two grades—my capacities as they are now, given who I am, and the range of capacities I could have. Given who I am now, I would find it impossible (psychologically) to do many things that stronger people might be able to do and that I might be able to do if I were to become a stronger person. Assume that second set of possibility has limits that are perhaps species-specific. Do these even come close to the sense of physical impossibility that grounds arguments about ought implying could? But notice that even if they do, nothing discussed here so far (including disgust reac-

THE NEXT BEST THING

247

tions) comes close to this sort of impossibility. For even in the case of disgust, we have allowed that some people are capable of overriding their disgust reactions (most notably, Catherine of Siena; see William Miller 1997). So, am I driven back to calculational opportunism as the only reason to advocate advocating second best? Is that the only reason to prescribe freeing slaves over feeding the hungry? To do so is to accept the picture as one that sees the matter as respecting our agency or behaving opportunistically. But that seems to me to set up a false dichotomy. The tradition of moral reasoning begun by Piaget and developed by Kohlberg treats (nearly) all of us as capable of moral reasoning with varying degrees of sophistication. But in addition to its disinterest in the consequences of such reasoning for moral action, it treats moral reasoning as a process that is isolated from the effects of other psychological process. The whole enterprise is a purely internal matter. We tend to treat agency as an artifact of this reasoning process, and emotion as an interference— what Elster terms “sand in the machine.” That picture feeds the idea that with discipline we can filter out that interference. But what if that is an illusion? Some (like Peter Railton 1984) would turn this into a virtue, arguing that it is a hopeless enterprise for us to hope to follow the dictates of consequentialism directly and explicitly. Instead, we should aim to harness our psychology. “A sophisticated act-consequentialist should realize that certain goods are reliably attainable—or attainable at all—only if people have well developed characters; that the human psyche is capable of only so much self-regulation and refinement; and that human perception and reasoning are liable to a host of biases and errors. Therefore individuals may be more likely to act rightly if they possess certain enduring motivational patterns, character traits, or prima facie commitments to rules in addition to what whatever commitment they have to act for the best” (1984, 158). But, whether you view agency expansively as encompassing these emotions, or narrowly but as always operating with them in the background, understanding their vagaries, plasticity, and educability is unavoidable if we are interested in understanding our capacity for moral action. Only then can we know if the next best thing is the most we can reasonably be expected to do.

Bibliography Batson, C. Daniel, Tricia Klein, Lori Highberger, and Laura Shaw (1995), “Immorality From Empathy-Induced Altruism: When Compassion and Justice Conflict,” Journal of Personality and Social Psychology 68(6): 1042–1054.

248

MARTIN BUNZL

Berkowitz, Leonard, Sara Jaffee, Eunkyung Jo, and Bartholomeu Troccoli (2000), “On the Correction of Feeling-Induced Judgment Biases,” in Joseph Forgas (ed.), Feeling and Thinking (Cambridge: Cambridge University Press), pp. 131–152. Hart, Allen, Paul Whalen, Lisa Shin, Sean McInerney, Hakan Fischer, and Scott Rauch (2000), “Differential Response in the Human Amygdala to racial outgroup vs ingroup face stimuli,” Neuroreport 11 (11): 2351–2355. Hoffman, Martin (2000), Empathy and Moral Development. Cambridge: Cambridge University Press. Kunst-Wilson, William, and Robert Zajonc (1980), “Affective Discrimination of Stimuli That Cannot Be Recognized,” Science 207: 557–558. Ledoux, Joseph (1996), The Emotional Brain. New York: Simon and Schuster. Miller, William (1997), The Anatomy of Disgust. Cambridge, MA: Harvard. Phelps, Elizabeth, Kevin O’Connor, William Cunningham, E. Sumie Funayama, Christopher Gatenby, John Gore, and Mahzarin Banaji (2002), “Performance on Indirect Measures of Race Evaluation Predicts Amygdala Activation,” in John Cacioppo et al. (eds.), Foundations in Social Neuroscience (Cambridge: MIT Press), pp. 614–627. Railton, Peter (1984), “Alienation, Consequentialism, and the Demands of Morality,” Philosophy and Public Affairs 13(2): 134–171. Rozin, P., and Carol Nemeroff (1990), “The Laws of Sympathetic Magic: A Psychological Analysis of Similarity and Contagion,” in J. Stigler, G. Herdt, and R. A. Shweder (eds.), Cultural Psychology: Essays on Comparative Human Development (Cambridge: Cambridge University Press), pp. 205–232. Sears, David, Jonathan Freedman, and Lelita Peplau (1985), Social Psychology. New York: Prentice Hall. Sidgwick, Henry (1907), The Methods of Ethics. London: Macmillan. Temkin, Larry (2005). “Thinking about the Needy, Justice, and International Organizations,” Journal of Ethics 8(4): 349–395. Wheeler, Mary, and Susan Fiske (2005), “Controlling Racial Prejudice: Social Cognitive Goals Affect Amygdala and Stereotype Activation,” Psychological Science 16(1): 56–63.

Chapter Thirteen What’s Wrong with Slavery? KWAME ANTHONY APPIAH

I grew up in Ghana’s Ashanti region, in its capital, Kumasi, which was, for more than two centuries, beginning in the late 1600s, the center of the flourishing Asante empire. At some point in my education (I don’t remember how early) I learned that the kingdom of Asante had been the center of a great trading system, with roads radiating out from the capital in every direction, connecting us, for example, with the coast and the Atlantic trading system, and with the north and the trans-Saharan trade. Gold, everyone knew, was one of the commodities we exported: the empire of Asante covered most of what was once called the Gold Coast (and became the British Gold Coast Colony and then, at independence in 1957, Ghana). And we imported weapons, I knew, beginning with guns supplied through the Dutch and Danish and European castles on the coast, including Christiansborg Castle in Accra, finished in 1661 by the Danes, just a few years before the first Asante king, Osei Tutu, began his rise to power. (The ancient guns now carried by Asante soldiers on ceremonial occasions today are called “Dane guns” in much of Englishspeaking West Africa.) What I don’t remember hearing much about was the centrality of the slave trade to the growth of the Asante empire in the eighteenth century. As plantation slavery expanded in the New World through the eighteenth century the demand for slave labor grew enormously. At the start of the eighteenth century as much as two-thirds of the value of Asante trade was in slaves. By the late eighteenth century 100,000 slaves a year were leaving Africa’s western coasts for the Americas, many of them through the forty or so British, Danish, and Dutch forts along the Gold Coast. When the British banned the North Atlantic slave trade in the early nineteenth century, Asante expanded its trade in kola nuts (whose role in West African social life was somewhere between that of coffee and that of alcohol in the modern United States). But the suppression of the slave trade began the period of imperial decline, which was to end with final conquest by the British at the start of the twentieth century. Ironically, however, while the slave trade played an important role in the consolidation of Asante power, it was probably the importation of

250

KWAME ANTHONY APPIAH

slaves from farther east, sold to Asante—and some of the Akan states that preceded it—by the Portuguese starting in the fifteenth century, that made the rise of Asante possible at all. For it is likely that, in West Africa, as in the New World, the rise of settled states depended on the rise of plantation agriculture; and the rise of plantation agriculture depended on involuntary labor. It was, of course, dependent on the development of the Atlantic trading system as well, since some of the major food crops— cassava and maize, most prominently—were brought to the region by the Portuguese. Both rubber and oil palms, which were important commercial crops, were indigenous plants. Cocoa, which became a major export commodity in the Gold Coast in the late nineteenth century, was brought back from a trip to Equatorial Guinea by a Ga blacksmith named Tetteh Quarshie (as every Ghanaian child learns in primary school). There are other reasons why slavery was important to the history of the society in which I was raised. Systems of land tenure in the region did not allow private land ownership. Heads of families could acquire for their lineages the right to farm land, but the land itself could not be sold to others, since the land itself was held by the chiefs and belonged to the community. The result was that one means of the acquisition of wealth— the capture or purchase of land—was unavailable. Slaves—and especially trained slaves—were (along with gold and other precious metals) an important alternative way of accumulating wealth. And displays of one’s wealth in slaves and in gold were an important means of establishing social status. To speak of slavery in these ways—as an economic and political phenomenon, a response to the question how to accumulate resources and stabilize governments—is, of course, to ignore the central moral questions about the practice. And, once we do turn to those moral questions, we can see this abstract economic or political approach is also, in at least one respect, potentially highly misleading. For to speak of slavery in this generalizing way is to ignore the fact that the people who entered the Atlantic slave trade from the fifteenth century onwards were available for purchase for a variety of different kinds of reasons. To say only that they were slaves already is to lose track of the fact that, in the societies from which they came, they did not all have the same status. In Asante, for example, there were people of subordinate status who worked for others and were part of their households, but who could not be sold. These people—the Twi word is akoa—were what we should probably call servants. But the word akoa was used more widely in connection with other forms of subordination, so that, for example, any subject could be called the akoa of his chief, and an inferior chief could be referred to as his superior’s akoa. Akosua Perbi, the leading scholar on slavery in the Ghanaian past, treats these extended usages as figurative

WHAT’S WRONG WITH SLAVERY?

251

(Perbi 2004, 3). After all, a chief might refer to his subjects as his children, as well: in each case, however, the point was to focus on a reciprocal, though unequal, relationship, in which the superior had responsibilities for the welfare of the subordinate as well as a right of command, and the subordinate had duties of obedience to the master or mistress, but also rights against him or her to maintenance—to food, clothing, and shelter. An akoa would also have had other rights: in particular, to marriage and to life. In every Akan community only a chief had the right to kill someone, and then only for an offense; and a master who engaged in adultery with his akoa’s wife was (at least in theory) as liable to the penalties for adultery as anyone else. An akoa was not a chattel, a possession. While you could, no doubt, send servants to work for someone else for a period, you could not transfer them to anyone else without their consent (or that of their families). There was another kind of subordinate person in many households, an awowa, whose presence was the result of an economic transaction: an awowa was a member of someone else’s family who was sent to work for you in return for a loan, and who remained until the loan was repaid. (The usual translation of awowa is “pawn.”) But a pawn was not strictly a possession either. You couldn’t sell her because, of course, she was due to be returned to her family once the debt was repaid. There were two other kinds of subordinates in many Asante households, who might have been called “slaves.” One was the dommum or war captive. Such captives might, if they were lucky, be redeemed by the payment of a ransom. But if they were not redeemed, they were obliged to work for and obey their captors: and, in the end, they might be sold. The other was the donko, whom you would most naturally call a slave. The donko had been bought, or could be sold, in a slave market. Some had become slaves under the law or custom of some other state and had been bought by Asantes; some were the children of existing slaves; and some had been captured in warfare or raiding and not been redeemed. A final sort of subordinate, who could be found at work especially in the households of chiefs, was the akyere. This was someone who had been found guilty of a capital offense and condemned to death. (Since there were no prisons in Asante—though there were regions of towns set aside for the residence of such condemned people—the main forms of punishment were fines and corporal and capital punishment.) Such people were not executed immediately on sentencing. Rather they were compelled to do what they were ordered to do through the authority of the chief who had condemned them until such time as it was deemed religiously auspicious to kill them. When a king died, for example, many of these condemned people would be executed to “accompany” him to the next life. Or they might be executed as an offering to a god to insure its

252

KWAME ANTHONY APPIAH

support in warfare or some other major state enterprise. The result was that many people who had been condemned to death lived on for many years under sentence. These distinctions in status between the akoa, the awowa, the dommum, the donko, and the akyere did not correspond to differences in the tasks that a person would be asked to perform. They differed in how they entered the status—by birth, capture, pawning, or capital sentencing— and in whether or not they could be sold to others inside or outside the kingdom. And all of them had rights against their masters or mistresses.1 When I was a child, people used to visit us regularly from a village called Nuadom, in the forest to the south of Kumasi. The village belonged, I was told, to my father’s family, and he inherited responsibility for it when he became the head of the family at my great-grand-uncle’s death. Being head of our family gave him the right to appoint chiefs and queen mothers for the village, and allowed him to exercise rights over land of the sort that chiefs characteristically exercise in other parts of the country. The reason for this anomalous situation, as I learned from my father’s autobiography, was that Akroma-Ampim, an Asante general who was an ancestor of his, had settled the area with numerous war captives a couple of hundred years earlier, and received permission from the king to settle these slaves there and to develop a plantation. (Those slaves then had the status of a dommum.) The name of the place comes from the word nuadewa, which is the name in Asante-Twi for the garden eggplant, and family tradition has it that that was a major crop. The visitors who came to ask my father to settle disputes brought us fruits and coffee that they had grown, as well as the occasional chicken, turkey, or sheep. But though my father met and talked to them, he tried to persuade them always that they had to settle their disputes for themselves. He was willing to help them, he told them, but they no longer belonged to him. The people of Nuadom are, by now, Asantes if they are anything (as well, of course, as being uncontroversially Ghanaian citizens). Their ancestors were almost certainly not, and their status as the descendants of captives was one of hereditary inferiority to free Asantes. And that still matters, even though slavery hasn’t been legal in Asante for roughly a century. (The final rules for abolition were made in 1908: they allowed for redemption for a fixed fee, obliged men to emancipate female slaves with whom they had children, made cruelty the basis for emancipation, and declared that children born to slaves after a certain date would no longer be slaves themselves; Perbi 2004, 193). On several occasions, a man who worked for my great-grand-uncle Yao Antony, and who often, as a result, had dealings with Nuadom, has insisted to me that he is not 1

See Perbi 2004, which is my source for this description.

WHAT’S WRONG WITH SLAVERY?

253

from that town. It is only recently that it has occurred to me that this matters to him, not just because he is, no doubt, concerned, as Asantes tend to be, with a proper accounting of his origins, but also because his family, unlike most of those in the village, never belonged to anyone. The low status of these slave ancestors still matters, then, generations after slavery has gone. Even though the sale of slaves is illegal and demands for unpaid work are unenforceable, there are still people, descended probably from either a dommum or a donko, who work in the households of prosperous Asantes without remuneration. (There are also some who are sent by families that cannot afford to feed them, people who are properly servants, nkoa, though their remuneration is not monetary, and the families to which they go tend to treat them like the poor relation in a Victorian novel.)2 Their status, so it seems to me, is like that of children. They are in the care of the families they work for, which have obligations to maintain and support them, but their labor and their lives are pretty much governed by the wills of the people in whose households they live. They are a kind of fictive kin, if you like, but they are not social equals. In principle, they are free to leave whenever they choose. In practice, they often have nowhere to go. Never having been paid, they have no savings. Often they have only the most basic education, so their only skill is domestic work, a market in which there is a great deal of competition. Once, when I was a child, I asked my father, in a room full of people, how we were related to a woman whom I knew and liked, who lived in one of the family houses. I thought of her, in fact, as one of my “aunts.” My father brushed the question angrily aside. Only later, when we were alone, did he say that one should never inquire after people’s ancestry in public. She was, as it turned out, the descendant of a family slave. Everybody in the family knew this, and that meant that she was of lower status than the rest of us. As children we were required to be courteous to all adults, even those of lower status. But that didn’t mean they weren’t inferior. I am not here reporting my father’s thoughts or mine about this woman’s status. My father was trying to avoid embarrassing her; he almost certainly did not think her ancestry an embarrassment himself. But she was often treated by people I knew in ways that reflected a conception of her as having an inferior status. And I suspect that the most important truth here is that this is how she thought of herself. She could not be sold, like her ancestors; she could not be separated against her will from her children; she was free to work wherever she could and live wherever it pleased her. But she would never think of herself as equal to 2

Nkoa is the plural of akoa.

254

KWAME ANTHONY APPIAH

us: and the safest place for her, if she could not find a husband to provide for her, was with the family to which her ancestors had belonged. A prosperous husband was unlikely to marry her, because of her status; and the status of her children, unless they were adopted by someone else, would be her status, because we are a matrilineal society. So she stayed. This woman is not a slave. But she carries on something crucial to slavery as it was in Asante: beyond the possibility of being sold away and the impossibility of making your own decisions about how to work, slavery meant that certain people were hereditarily inferior. You can abandon the slave markets and demand that everyone who works is paid for their labor and is free to leave it; but the last work of emancipation is to remove that stigma. Out of these fragments of experience of life with one kind of African slavery, I ask myself what it is about slavery that is wrong. European and American abolitionists in the nineteenth century tended to focus, reasonably enough, perhaps, on the cruelty of plantation slavery. Frederick Douglass, in his autobiographies, pointed to the violence, the whippings, the mutilations, and the unpunished murder of slaves. Harriet Jacobs spoke of the way in which white men felt free to have sex with their female slaves—to rape them, we should now say; about the cruelty of white women that their knowledge of their husband’s infidelities could lead to. Both spoke of the broken families, children wrenched from their mothers, wives from their husbands, sold “down the river,” for profit or as punishment. But the reply of slavery’s defenders to these charges was that these were abuses of the slave system. By the end of plantation slavery in the American South an ideology that Eugene Genovese has rightly called paternalist had developed according to which the slave master was the paterfamilias of a great extended family, and his black slaves, incapable (like his wife and his children) of managing their own lives, were taken care of and assigned their tasks by him. Some people believed this, no doubt. Many, we can be sure—not all of them black—did not. But the defense, such as it was, amounted to this: cruelty is indeed wrong, but slavery need not be cruel. If you have a problem with slavery itself, it must be something else. This answer strikes me as being onto something, even if it also surely misses as much as it gets right. For, of course, if slavery gives the master great power over the slaves, it is predictable, given the way we humans are, that it will often be abused. To say that there is nothing wrong with the system, in these circumstances, is like leaving a child alone with a butterfly and then expressing shocked surprise when the creature’s wings are torn off. Still, the paternalist response to the accusation that slavery is cruel gets something right: physical cruelty is, indeed, not intrinsic to the definition

WHAT’S WRONG WITH SLAVERY?

255

of slavery, and, especially when slaves are expensive, it is not likely to be common. But to say this is not to defend the institution (as the paternalists wrongly believed). For the lives of the happiest slaves, men and women well treated by their masters, are still diminished by their status. And at the heart of the matter, so it seems to me, are two issues: hereditary social inferiority and heteronomous personal life. I want in the rest of this essay to explain something of the character of these harms, and argue that freeing someone from the legal status of a slave does not guarantee that you will solve either of them. Emancipation is important, but undoing the harms of slavery is, I think, slow work. The differential treatment of blacks and whites in the United States before emancipation simply presupposed the inferior status of black people. Certainly slaves were legally inferior in social status: they were expected to display deference to others, and if they failed to do so—if they were “uppity”—they could be punished for it. But it was because they were black that they could be enslaved, and even free blacks were usually treated as inferior to white people of the same wealth or accomplishments. White people could be Mr. and Mrs. Smith; black people were Jim and Jemima. Black people were expected to let white people precede them through doorways, to step out of their way on the sidewalk, to walk past them with heads bowed, to avoid looking them in the eye. These rules largely applied, as I say, to free blacks as well, before the Civil War. One way of understanding this fact is to suppose that even free blacks were stigmatized because, though they were legally free, they belonged, unlike white people, to a kind of people that could be enslaved. There was, in other words, an intimate connection between racial disrespect and slave status—a nexus nicely captured in the Fifteenth Amendment to the United States Constitution that says that voting rights may not be abridged on account of “race, color, or previous condition of servitude.” Being a Negro, having a dark skin, being a slave, were easily conflated in the American system. More than this, because slaves were inferior, their work was stigmatized, too: agricultural labor, associated with deferential, dependent slaves, was something that respectable people did not do. It is noticeable that that association has survived. White people who do the sort of work the slaves used to do, working all day in the hot sun, get sunburn on their necks. Their necks are red; and “redneck” is not a term of respect. Notice that the law played a relatively small role in defining the patterns of deference required of the slave. An owner’s right to punish a slave was a matter of law; treating a white person, even a social inferior, this way would, from a legal point of view, have been assault and battery. Deference was enforced not so much by legal punishment as by a social

256

KWAME ANTHONY APPIAH

understanding. As a result, when, after emancipation, the legal apparatus of slavery was disestablished, the patterns of deference could be, and, as we know, often were, continued. This is true about post-emancipation societies everywhere. Much of the labor that had been done on farms in central Ghana by slaves in the nineteenth century was done by the midtwentieth century, by migrant laborers from further north (from northern Ghana and countries of the Sahel). These people often had homes in places where many of the slaves had come from in the past. They were doing work not suitable in Akan eyes for free people. Similar patterns of contempt for the work that the slaves once did can be found in the Caribbean and in Latin America. Slavery in other parts of the world and at other times was also associated with one particular ethnic or racial group. The etymology of the English word “slave” reflects the large-scale enslavement of Slavs in the European Middle Ages; and in modern Arabic (in Iraq, for example) the word ’abd—a classical word for a slave—is used to refer to darkskinned people. These associations are not, I think, very surprising. If low social status is part of the conceptual content of slavery, then it is natural that people who own slaves should want to think of themselves as belonging to a different kind. This can provide a rationalization for the bad treatment that slaves receive—the work they are assigned, the rights that are denied to them—soothing those pangs of conscience that this abusive treatment might otherwise give rise to. Aristotle argued, famously, at the beginning of the Politics, that certain people were deficient in the capacities for rational deliberation required to manage their own lives and that these people were slaves “by nature.” (He also believed that, where a person who was capable of rational deliberation was treated by law or custom as a slave, this was wrong.) Here the low status and poor treatment of the slave has been rationalized as a proper response to a preexisting deficiency. The point is simply this: once you think of slaves as belonging to a kind—a race, a tribe, a class—that is suited to enslavement, then the abandonment of the formal institutions of slavery is not likely to change this fundamental attitude. As I have already said, the Fifteenth Amendment presciently anticipated that the vote might be denied to people on the basis of “race, color, or previous condition of servitude.” But the vote was only one element of what the ex-slaves discovered might be denied to them. For what they really ended up without was not just the respect of the white majority but what John Rawls called the “social bases of selfrespect.” Rawls argued that it was essential to a life of dignity that one should have confidence that one had a life whose aims were worth pursuing and that one was competent to manage that life. To have this confidence was

WHAT’S WRONG WITH SLAVERY?

257

to have self-respect; and to maintain this confidence, he thought, it was necessary that the institutions of the state should reflect back to one an image of oneself as a dignified person deserving of respect. This psychological claim is extremely plausible. It is one of the great wrongs that racism, sexism, and homophobia have all inflicted that they have created a society that reflects back to black people, women, and lesbians and gays a demeaning image of themselves that is an obstacle to self-respect. As Charles Taylor has insisted, it is precisely this active debasement of people on the basis of their race, nationality, sexual orientation, class, or gender, that has fueled the “politics of recognition” that has come to be so dominant in modern societies. People who are treated with social contempt recognize that they are being wronged and demand the fundamental respect whose denial they understand has made self-respect difficult for them. Slavery as a social or legal institution has built into it, in other words, a denial of the social bases of self-respect: it defines the slave as lower in status precisely by denying that she has a plan of life worthy of consideration and rejecting the enslaved person’s claim to manage her own life. Some slaves, as Frederick Douglass demonstrated vividly, sustain self-respect in a society that denies them respect. (So too, women in sexist societies may know that they live lives worth living and have plans that they are entitled to pursue, even when most people around them think that their lives need managing for them.) But the experience of enslavement and servitude makes this very difficult. And placing this obstacle to selfrespect in a person’s way does her a very great wrong. It does her wrong, of course, whether it is done by the state or by society without legal sanction. And that is why, when slavery as a legal institution is abolished, there is no guarantee that the wrong will cease. We have seen this over and over again in post-emancipation societies historically, as I have already pointed out. And so it is not surprising that the end of slavery is only the beginning of real freedom. So much for an outline of the first of slavery’s major offenses against the enslaved: its imposition of a low status with the consequent denial of the social bases of self-respect. The second, associated wrong, as I say, is that slavery imposes heteronomy on people. Instead of allowing them to manage their own lives, it allows the slaveholder to shape the slave’s life entirely without regard to her concerns. The denial of the social bases of respect undermines one’s belief in one’s capacity to run one’s own life; the imposition of heteronomy denies one the exercise of that capacity. It is central to human flourishing that each of us should be master of our own fate; we have ultimate responsibility for our own lives and the right to make the major decisions. Those decisions, properly made, will acknowledge the moral interests of others, so that mastering my fate

258

KWAME ANTHONY APPIAH

cannot mean ignoring the fates of others. But even acknowledging the rights and interests of others is something that I must take responsibility for myself. As John Stuart Mill wrote in one of the most famous lines of On Liberty: “If a person possesses any tolerable amount of common sense and experience, his own mode of laying out his existence is the best, not because it is the best in itself, but because it is his own mode” ([1859] 1977, 270). Mill—and Humboldt before him—placed the right to be in command of one’s own fate at the heart of the liberal tradition. And it is not, when one thinks about it, surprising that it is that right that the slave is denied. After all, the Latin word “liberalis” means “worthy of a free man.” To buy and free a slave is to remove legal or social recognition from the slaveholder’s claim to manage the slave’s life for her. That is worth doing. But you cannot thereby command respect for her or grant her self-respect. Neither is within the power of the market. Nor can you guarantee that someone who has experienced only slavery will be prepared to manage a life alone. It is possible that the recently enslaved are likely to be quicker to recover command of their lives, so with them we may achieve more than when we seek to liberate people born into slavery. But in thinking about what we are doing when we send money to buy people out of contemporary slavery, we should recognize that their escape from the harms of slavery is not yet achieved: emancipation is only the beginning of freedom.

Bibliography Perbi, Akosua Adoma (2004), A History of Indigenous Slavery in Ghana: From the 15th to the 19th Century. Accra: Sub-Saharan Publishers. Mill, John Stuart ([1859] 1977), On Liberty, The Collected Works of John Stuart Mill, vol. 18, edited by John Robson. Toronto: University of Toronto Press.

Appendix: “They Call Us Animals,” Testimonies of Abductees and Slaves in Sudan JOK MADUT JOK

Introduction The resurgence and practice of slavery in Sudan between 1983 and 2001, although built upon the old racial, social, and economic foundations of nineteenth-century slavery, was triggered this time around by the ongoing north-south conflict (Idris 2001; Collins 1992; Johnson 2002). The government used it to achieve two goals at the same time. It allowed the government to arm and train the militias that were also used to fight the war by proxy. It also functioned as a way to appease the Arab tribes of Darfur and Kordofan, which would have otherwise been hostile to the government in Khartoum for its neglect of these areas in terms of social services. The result has been that southern women and children have been abducted and taken into slavery in the north and subjected to various fates. Women are used for physical labor in farming, as domestic servants, and as sexual slaves. Large numbers of women have been found to have performed vital labor for Arab farmers and herdsmen over this period. Women have also been useful to the slaveowners biologically as well as ideologically. Biologically, their reproductive potential has proven one of the reasons for the persistence of this practice, as many Arab families make claims to children born to these women during their captivity. Ideologically, these slave women serve as the means with which to reconfigure race relations in order to forge a homogeneous society based on Arab and Islamic identity. Southern women are to be found throughout Darfur and Kordofan, as forced and free laborers, as concubines and wives, and as internally displaced persons and migrant workers. Children, both abductees and those born in captivity, are also found in these conditions in the north. This appendix attempts to describe what life looks like for these slave women and children. This is done through testimonies from some of the women and young adults who were once enslaved but managed to escape or were released otherwise. Years ago, when I was doing research for a book on slavery in Sudan (Jok 2001), I traveled all over Bahr el-Ghazal, the southern Sudan’s region most affected by slave raiding. Since 1999, I have undertaken ten trips to the region, six of them related to a project aimed at making a registry of

260

JOK MADUT JOK

people who have been abducted from this region by northern Arab tribesmen in the employ of the Sudan’s central government during the course of the ongoing war between north and south.1 During all of these trips, I was able to interview a large number of families that had lost their loved ones to this raiding, many taken into slavery. I also interviewed numerous ex-slaves who had escaped or had been, for one reason or another, released by their captors or masters, or who had been freed through slave redemption. From these interviews, we are able to get a picture of what life is like for the abductees, especially those who have been enslaved or made to work for a meager wage under horrendous circumstances in a deeply racially and religiously divided society. While it is hard to be certain about how typical the views of the interviewees are for all the enslaved, they represent a few of the various fates that await the southern abductees in the north. The testimonies represent the various forms of abuse that are experienced by abductees, some sexual, some in the form of forced labor, others as an effort to change their religious beliefs, and others as physical punishment for failure to accept the slave status or for refusal to perform the tasks assigned. In the Words of Former Slaves Our interviews and many NGO reports contain lengthy and detailed testimonies of women and children abducted and kept in these circumstances, who regained freedom only by escaping or through ransom. The captors are often referred to as PDF, Murahileen militia, or sometimes, even as government soldiers. It is therefore beyond doubt that regular troops also take part in or facilitate the raids. According to certain accounts, the perpetrators were said to be wearing uniforms, whereas Murahileen and other militias usually wear plainclothes. Although an auxiliary force, the PDF are directly under the control of the Sudanese authorities, sometimes conducting raids jointly with the regular Sudanese troops. Moreover, the militias are armed by the Sudanese authorities and receive ammunition from the Sudanese army. Raid preparations follow established patterns, leaving little doubt as to the Murahileen’s intentions of capturing women and children for slavery practices. The testimonies of the abductees contain descriptions of the ill treatment and forced work to which many abductees are subjected, usually involving 1 As the codirector of the Rift Valley Institute project on abductee registry, I was able to travel widely in northern Bahr el-Ghazal between 2001 and 2004. The project was funded by Open Society Institute, the Kaplan Fund, and the UK Department for International Development. On behalf of my colleagues John Ryle and Fergus Boyle, I extend a word of thanks to these agencies.

APPENDIX: “THEY CALL US ANIMALS”

261

cooking, cultivation, tending animals in the desert, collecting firewood, fetching water from the wells, washing cloths, and other domestic chores. Domestic servants report being made to sleep in the kitchen, while the cattle tenders are made to sleep in the same space as the cattle. They also report various other forms of abuse such as being addressed with demeaning terms like Abd, “slave.” Women’s and girls’ testimonies cite rape, forced “marriage,” and other sexual abuses amounting, in certain cases, to sexual slavery. Many of those who were freed were either pregnant or gave birth to children fathered by their captors. In some cases, women were freed and instructed to return to the South but leaving the children behind with the Arab men who fathered them. Atem Garang Akol, who was abducted at a very young age, and had worked for his captor Muhammed Juma, spoke of witnessing many children being sold to the Kababish who live further up in North and West Darfur. He also spoke in detail about his duties during the many years of his captivity. He had been promised by his master that he would not be treated like a slave and that he would be paid for grazing duties he was performing. His payment was one cow paid to him annually. He was told to keep his cows together with the master’s and that he would have the exclusive use of their dairy products, but he was prevented from selling the cows or taking them away, which really means that they were not his cattle, and therefore in reality, he was not being paid for his work. Speaking about his master, he said: When I was first turned over to him by the people who had captured me, he had told me that I should not fear him and that I would be treated just like one of his children. He said that being asked to do things around the house or to graze cattle was not to enslave me, but rather to share the tasks with the other boys, his sons, and this seemed to have been the case for the first ten days, after which everybody else in his house began to be really mean and calling me names such as abd or jengai, the black one, the donkey, or the dog. This is when everyone began to order me around. Even the other kids whom I had been told would do the same chores as myself were now sending me to serve them. When I would go grazing with them, they would go and sit under the shade of the brush and shout at me that I should go and round up the cattle. . . . They threatened me with violence. . . . They said they could simply kill me right in the forest and go home to tell their family that I had escaped and nobody would question them. I was always terrified. I had seen other boys and girls subjected to terrible cruelty and I dared not risk becoming like those children. I did everything as I was told and that is how I was able to survive. In fact I think he began to think that I had forgotten where I had come from. But the

262

JOK MADUT JOK

scar of everything I have witnessed throughout my life since the day of that raid in which I was seized is not something anyone can ever forget no matter what the Arab master gives you . . . of course there were some days when no one hurled insults or beat you, and you forget those nice days. But the really dark days you cannot forget. I had never stopped dreaming of the day when I would return home and I was extremely overjoyed when these CEAWAC men came around and told me it was time to go home. I had no idea that I would come here only to find that I have no family left and nobody is interested in me. All the dreams of being finally in a place where I looked like everybody else were shattered when I found out that there was no food, no school, and no home to go to. I might have to return to the north. . . . Of course I hated being a slave but at least I was able to eat. . . . If I return to the north, I will definitely not go back to where I was enslaved. Here in the south, being a free man with nothing to eat, I do not know which existence is better. If anyone was really interested in helping us, they would have given someone like me the legal power to get my cows from this man who promised me payment and then refused to respect it. Atem was abducted together with his brother Akol; Akol is still missing to this day and Atem does not know his whereabouts. He still speaks Dinka, but a broken form of it. He cannot remember the name of his village. He is only able to recall the names of some of his relatives and their clan, but no one in Alek has been able to recognize the names of his people. Atem also spoke of the kinds of atrocities that are usually committed by the captors when they discover the captives are planning to run away or to take away their cattle payment. He himself got his teeth knocked out with a bamboo club one day by his master for having allowed the goats to go astray. One boy he had known very well, and with whom he had met frequently on the pastures, one day stole the camels and drove them all the way to Amkabish between Mujalid and Fulla. He was found by other Rezeigat a branch of the Baggara Arabs and they captured him, took the camels away from him, and enslaved him again. Atem said that his master had told him this story frequently as a reminder that any attempts to escape would end in the same way as this boy. Another boy who had attempted to escape was punished in the most severe manner. He got his Achilles tendons severed with a machete and was disabled forever. He was made to spend his days on the back of a grazing camel. Every morning the camel kneeled and he would climb on; upon return in the evening, he would climb down and crawl into the barn where he slept. Atem describes the misery of this boy as “unsightly.” He said that “His knees and palms were all scabbed from crawling without any

APPENDIX: “THEY CALL US ANIMALS”

263

clothes; he wore rags, and his face was always so downcast that he made me not want to look at his eyes when he spoke during the few times that I would go visit him in his barn. . . . He made me feel so bad about him. At times I even felt much better off despite my own miserable futureless existence. . . . It is so strange how one becomes immune to pain when you have not known much else all your childhood.” Another returned abductee I interviewed, a young woman whose master had renamed her Zeinab, had the following to say: The man in whose house I had worked for a few months following my abduction was nice at first, but when he decided one day to force himself onto me he would not stop anymore and there was nothing I could do as he told me that he would tell his other friends to rape me if I did not comply with his demands. He demanded that I become his wife, but he had no courage to tell his real wife and did not want his neighbors to know about it. His wife still found out and she became furious but aimed her anger at me instead. She would not stop yelling all kinds of insults at me and threatened to kill me, at which point the husband decided to let me go because it was going to bring disgrace to him. At this time I was already two months pregnant. I moved to a displaced persons camp in el-Daein where I met up with other Dinka people from my home area. I remained there for nearly five years, and that is where the CEAWAC people found me and told me they could take me home if I wanted to. Although my child is an Arab child, I consider him my compensation. I am blessed that my child was not taken away like all the other women who were forced to have children with Arab men and were forced again to leave the children behind. Many similar stories abound: children who have reported serious physical abuse to the police only to end up in police detention instead of their cases being investigated; girls who have been sexually abused by a father and his sons; children who feel forced to remain in captivity because they do not know where they would go if they escape. Other stories are about escape from bondage, the journey back to the south, hunger and thirst along the rough terrain between towns in Kordofan and Darfur and the river Kir. The picture that emerges from these stories is one of total despair; but one that does not diminish the thirst and quest for freedom. And freedom, whatever disappointments await one at the end of the escape tunnel, is worth every bit of sacrifice, the slaves say. Akuol Majok, a 27-year-old mother of a four-year-old girl, was abducted from the village of Mabok Tong in northern Bahr el-Ghazal in 1998. Her daughter was fathered by one of her captors. She was now a free woman, as she had been released through her family’s efforts to redeem her. She was pregnant with her second child when we interviewed

264

JOK MADUT JOK

her in the summer of 2003. She is not certain about the father of her unborn child, as she said she had been repeatedly the victim of sexual assault while in captivity. She recalled three particular men, including her main master, a man by the name of Babikir al-Maali, who had forced their way into the kitchen hut where she slept anytime of the night they desired. Akuol said that the worst time was when I was breastfeeding my first child. Something that we Dinka cannot do is to have a man touch you while the child is still suckling. . . . I tried resisting at first and I got beaten and really badly injured. Sometimes I would wake up in the morning with my entire face completely swollen from the beating the night before. Some sympathetic people would suggest that I try to inform the authorities. But which authorities? Every woman abductee has gone through the same and none of them that I know has actually gotten any police to do so much as ask these animals who treat you like dirt. . . . I knew two women who complained to the police and they got laughed at and rebuked and told to return to work. I was so happy when they let me go, but horrified at the thought of leaving my little girl behind, as my master wanted to keep the child. . . . He allowed my girl to leave with me after she cried, wailed and went through a fit. The Umda [chief] of the area sympathized with me and ordered that I be allowed to go with my kid . . . that was a rotten place where rotten men behave as if they are not afraid of God. . . . Ironically, even people who are dirt poor, just like me, believe that they are by far better than them and have their reasons to scorn them. In a way, being a slave adds onto the filth of poverty, and you are pushed even deeper than the bottom of anyone’s shoes. Mawien Garang Lual, a young man in his twenties, was abducted from Aweil East County during a raid on the biggest Dinka-Arab border market, called Warawar. Like many others who were deemed unfit for enslavement, Mawien was not held in captivity, but instead, he was dumped in South Darfur. After about a year of wandering in the area from one displaced persons camp to the next, he was able to settle in the biggest of the southern IDP camps, el-Daein. He spent another year in the camp before he was to find a group of Dinka men and women who had decided to venture back to Dinka territory, across the same rough terrain through which they had been dragged at the time of the abduction. When we interviewed him in Malwal Kon in the summer of 2003, he described his ordeal and that of hundreds of thousands more like him: IDP camps in Darfur have two lives, one at night and another during daytime. In daytime, aid workers come to feed the displaced, distribute

APPENDIX: “THEY CALL US ANIMALS”

265

grain, provide health care, and interview people about their future aspirations regarding a possibility of return home. Everything would look as normal as a refugee could be. At night, however, it is a different story. The camps are swarmed by security agents to interrogate and arrest anyone they would have noticed talking to the foreigners during the daytime. The Arab Baggara would also send groups of armed men to confiscate the relief supplies distributed earlier to the refugees and abductees. All the food, especially grain, is looted and people beaten. This food gets used to hire some of us to work for the Arabs to clear planting fields and to serve in their homes as servants. The supplies that were supposed to go to us are confiscated and are used to enslave us with. They sing songs about these activities. One song is about how to use the Khawaja [whites] food to hire the Abid [slaves]. Makuc Chol Bol, a 17-year-old boy, had just escaped and returned to northern Bahr el-Ghazal when we met him in Tuic County in 1999. He had been in captivity for some four years. He described the incident in which he was abducted: When the attack came and my whole village ran in disarray, I took to running with my older brother and my uncle. The sounds of gunfire, dogs barking, people shouting and children crying was all just maddening, and no one knew which direction was safe to run towards. The horseback Arab attackers had actually come from the southern end of the village, where they would be least expected to come from. When we ran, some people managed to hide behind the bush or jumped in the river to cross to the other side. Some people were shot dead right then and there. One of the raiders caught me, and with a number of others, we were chained and marched north, through the forest with no water and no food on a three-day journey. As soon as we got into the Arab territory, all the different raiders took any number of abductees they desired and went whichever direction their home village was. I was taken to the town of Meiram and sold to one Yussif Ibrahim, for whom I became a slave for four years. For the first year I wore the same rags I had on when I was taken until they fell apart completely, and that was when Yussif gave me his old gown to tear it apart and pieced it together into a pair of shorts and that was the only piece of cloth I would have for the next several months. I had doubts that I would ever leave that place and to see my family again. While I was there, for nearly a year, I woke up every morning hoping that it would be different day and something or someone would free me from the bondage. I worked so hard everyday hoping that Yussif would either like me for my hard work and treat me like his son, or would let me go. But the harder I worked, the more reliant on me he became and he would treat

266

JOK MADUT JOK

me like the sturdiest donkey he has ever had. You know how Arabs love donkeys; they would groom them and pat them only to put the heaviest load on their backs. Ariath Ariath Wol, a tall thin Dinka man in his forties, was not abducted, but his wife and children were taken during a raid upon his village of Matur in Aweil North County, and he has not been able to get them back. For one, he was shot in the lower back during the raid and is paralyzed from the waist down. For another, he had heard stories about his family from returning slaves. He said that “They tell me that my daughter has been forced to marry an Arab. They circumcised her as they do their girls and I can’t imagine anything worst than that for a young girl to go through. My wife is also said to have born more children by different Arabs. Now I cannot go looking for my family, I can’t avenge them, and I can’t even take care of myself. Sometimes I wish they had just killed me.” Conclusion This appendix has examined the use of slave raids by the government of Sudan to weaken the south-based armed opposition by destabilizing the civilian population. I have also attempted to show how, although the destruction is universal, the experience of enslavement is most taxing for women and children, who are subjected to horrendous working conditions while in captivity. The cumulative effect of military activities, the actions of slave-takers, and the attitudes of final owners of the slaves is a state of chronic physical abuse of the abductees and the enslaved. This in turn has led to a chronic population drain from the south towards the transition zone (southern Kordofan and Darfur), and further north to Khartoum. The northwards movement of the displaced has created other types of humanitarian problems. Thus, for example, the displaced in the transition zone are often exposed to economic exploitation by local landowners. It is also here that women are most susceptible to rape and labor exploitation, and children are subject to recruitment into the army to fight against their own people or blackmailed into the Muslim faith. It is here that the cultural front in Sudan’s conflict is most active. The good news is that raiding has stopped since 2001 due in part to progress made towards a peace deal between the north and the south. But the bad news is that there are still thousands of southern women and children that remain under conditions of enslavement in the north. As the peace agreement was being finalized, there were many voices within the civil society movements calling for a mechanism to be built into the agreement that will insure accountability for lives lost in the course of slave raiding. Human rights groups have also called for the establishment

APPENDIX: “THEY CALL US ANIMALS”

267

of an instrument that would allow the victims and their families to seek justice, and for a declaration that slavery be considered a war crime and a crime against humanity. This is because Sudan’s slavery was practiced in the course of a civil war and as a war tactic. These voices were not heeded by the mediators and the warring parties alike partly because negotiated settlements hardly ever end up restoring justice to the victims, as the mediators cannot insist on establishment of such mechanisms lest they risk losing the negotiating momentum. The parties to the conflict were also reluctant to build into the agreement a legal instrument that might be used against them. The result is that Sudan’s slaves might not see justice after all. An international criminal court for Sudan was suggested by the European Union and human rights organizations to look into the various violations in Sudan’s wars, including slavery, but the proposal was opposed by the United States, and Sudan’s military authorities, responsible for some of the gravest violations, will probably go free (see Human Rights Watch 2005).

Bibliography Collins, Robert (1992), “Nilotic Slavery,” in Elizabeth Savage (ed.), Human Commodity (London: Frank Cass). Human Rights Watch (2005), US Thwarts Justice for Darfur: For Bush Administration, Blocking ICC Trumps Protecting Civilians. Brussels: Human Rights Watch. Idris, Amir (2001), Sudan’s Civil War: Slavery, Race and Formational Identities. Lewiston, NY: Edwin Mellen Press. Johnson, Douglas (2002), The Root Causes of Sudan’s Civil Wars. Oxford: James Currey. Jok, Jok Madut (2001), War and Slavery in Sudan. Philadelphia: University of Pennsylvania Press.

Contributors

Kwame Anthony Appiah is Laurance S. Rockefeller University Professor of Philosophy at Princeton University. Kevin Bales is Professor of Sociology at Roehampton University London and president of Free the Slaves, Washington, DC. Arnab K. Basu is associate professor of economics and public policy at the College of William and Mary and senior fellow at the Center for Development Research (ZEF), University of Bonn. Martin Bunzl is professor of philosophy at Rutgers University. Nancy H. Chau is associate professor of applied economics and management at Cornell University and senior fellow at the Center for Development Research (ZEF), University of Bonn. Jonathan Conning is associate professor of economics at Hunter College and the Graduate Center of the City University of New York. Lisa D. Cook is assistant professor of economics at Michigan State University. Stanley Engerman is John H. Munro Professor of Economics and Professor of History at the University of Rochester. Jok Madut Jok is associate professor of history at Loyola Marymount University. Dean S. Karlan is assistant professor of economics at Yale University. Margaret M. R. Kellow is associate professor of history at University of Western Ontario. Michael Kevane is associate professor of economics at Santa Clara University. Alan B. Krueger is Bendheim Professor of Economics and Public Affairs at Princeton University. E. Ann McDougall is professor of history and classics at University of Alberta. Howard McGary is professor of philosophy at Rutgers University.

270

CONTRIBUTORS

Carol Ann Rogers is associate professor of economics at Georgetown University. John Stauffer is professor of English and American literature and language, and chair of the History of American Civilization at Harvard University. Kenneth A. Swinnerton is a senior research economist in the Bureau of International Labor Affairs, U.S. Department of Labor.

Index

abduction, in Sudan, 149, 152, 153, 154–57, 260, 265. See also slavery abolitionism, 200–211, 213, 217, 227, 254 Adrar region, 166 affection, relations of, 237. See also empathy Africa, 37, 38, 45, 84, 92, 95, 97 African Americans: after emancipation, 88, 89, 90–91; inferior status of, 255; movement north by, 90; payments for freedom by, 200; and redemption, 3; and redemption debate, 201, 202, 204, 205, 206–9; and slavery, 78, 81, 83–84 age, 44, 46, 66, 83 agency, 246, 247 agriculture: in Africa, 92; child labor in, 38, 42, 49, 62, 64, 66, 67, 68, 70; after emancipation, 87, 88, 89, 90, 91; forced labor in, 259; and involuntary contracts, 113; and model of labor and tenancy, 114–22; and output collapse following emancipation, 130; and Russian serfs, 188–89; stigmatization of labor in, 255; in Sudan, 148; in United States, 88, 89. See also plantations akoa/nkoa, 250–51, 252, 253 Akol, Atem Garang, 261–63 akyere, 251–52 Albania, 66 Alexander II, 189–90 Al-Maali, Babikir, 264 Alpers, Edward, 97 altruism, 242, 245 American Anti-Slavery Group (AASG), 144, 145, 146 American Anti-Slavery Society (AASS), 203, 208, 211, 213, 214 American Civil War, 200 American Colonization Society, 201–2 American Revolution, 200, 201, 211 Americas, 86, 88, 91, 94, 97, 103, 113, 119, 122, 181, 195, 249, 250 Amnesty International, 144 Anscombe, Elizabeth, 232 Antigua, 87

Anti-Slavery Society of France, 162, 163 Antony, Yao, 252 Arabs, 147, 149, 151, 160, 256, 260, 265 Argentina, 45 Aristotle, 232, 237, 256 Asante/Ashanti, 101, 249–54 Asia, 37, 38, 45, 95, 97, 130 asset inequality, 128–33, 134 asylum, 43, 61, 73 Atlantic trade, 249, 250 Augustine, W. R., 181 Auld, Hugh, 214, 227, 231 Auld, Thomas, 214, 230 Australia, 45 awowa, 251, 252 Aztecs, 97 Baggara, 34, 147–48, 149, 265 Bahr el-Ghazal, 149, 150, 152, 154, 155, 259 Baier, Annette, 233 Baland, Jean Marie, 110, 135 Bales, Kevin, 15 Banaji, D. R., 96–97 Bangladesh, 45, 66 Barbados, 87 bargaining power: and child labor, 43, 63, 67, 70, 71, 73; and slavery, 24, 215. See also labor unions Barrow, R. H., 96 Basic Human Functional Capabilities, 78 Basu, Arnab K., 39, 42, 61, 64, 70 Basu, Kaushik, 110, 135 Batson, C. Daniel, 239 Bauer, Arnold J., 113 Baum, Gerhardt, 144 Beecher, Henry Ward, 205, 210 Belgium, 39 Belize, 45 Benin, 66 Berkowitz, Leonard, 242 Bhutan, 45, 66 Bibb, Henry, 205 Bîro, Casper, 144 Black Codes, 127–28

272 black non-Arabs, 160 black people, 257. See also African Americans; skin color Blackstone, William, 99 Blum, Jerome, 181, 188, 189 Bodin, Jean, 98 Bok, Francis, 145 Bol, Makuc Chol, 265–66 Bolivia, 45, 66 Boston Baptist Association, 206 brain, 244 Brazil, 66, 86 Brenner, Robert, 114 bride price, 164, 167 British and Foreign Anti-Slavery Society, 159, 167 British Guiana, 87 British West Indies, 89, 93, 95, 100, 203, 218, 219–20, 221 Brown, John, 217 Brown, William Wells, 205, 210 Burkina Faso, 66 Burma (Myanmar), 66 Burritt, Elihu, 218 Bush, George W., 146 Bush, M. L., 113 buyers: of child labor, 43; and Douglass, 216; redeemers as, 25–27; of slaves, 22, 24; and voluntary slavery, 80. See also redemption; slave owners Cambodia, 45, 66 Cameroon, 66 Campbell, Gwyn, 97 capability, 79, 82, 93; defined, 77–78 capacity, 82, 246–47 Caribbean, 256 Catholic missions, 162–65 Central America, 45 character, 232, 233 Chau, Nancy H., 42, 61, 64, 70 Child, Lydia Marie, “The Quadroons,” 209 child labor, 37–73; bans on, 110–11, 135; and debt compensation, 114; distribution of, 38; in French West Africa, 165; history of, 39; incidence of, 37, 42, 49, 71; and law, 44; and parents, 40, 96–97, 98, 99, 112, 114; sources of, 43; supply of, 42; and World Bank, xi child pornography, 44, 51

INDEX

children: abandonment of, 95; of African Americans after emancipation, 90, 91; and American slavery, 84, 207, 208; in ancient Greece, 96; in armed conflicts, 37; of Asante, 252; in French West Africa, 163, 164, 172, 173; as gift to church, 95; and infanticide, 95, 96; inheritance of slave status by, 101; and labor policy, 34; legal definitions of, 44; in Mauritania, 166, 167; and parents, 82; plans for, 102; and poverty, 79; and rational choice, 83; rights of, 83, 96; sale of, 16, 96–97, 98, 99, 100; in sex trade/prostitution, 15, 33n, 34, 37, 40, 41, 44, 50, 51, 72; and slavery, 254; in Sudan, 149, 150, 151, 153, 154, 155, 259, 260, 261–63, 266; survival of, 96; trafficking in, 32, 40, 41, 44–61 China, 88, 97 choice, 79, 81–83, 84, 94, 238–39 Christian groups, 146, 147, 152, 231 Christianity, 163, 177 Christian Solidarity International (CSI), x, 27–28, 31, 143, 144, 145, 156 Clendinnen, Inga, 97 Coates, Lindley, 204, 211 Cochran, Johnnie, 145 coercion: and child trafficking, 49; and dependency, 82; and freedom, 82; of labor, 112, 113, 117–22, 130, 133; opportunity cost of, 110; by slave owners, 79. See also violence cognition, 238–39, 244 collective bargaining, 65, 71, 73. See also bargaining power Colombia, 45, 66 colonialism, 84, 159–77 colonizationists, American, 103, 201–2 compensation: and British abolition of slavery, 203, 218; as complicity in sin of slavery, 213, 214; debate over in United States, 200–211; and Douglass, 214, 215–16, 218–21; forms of, 103–4; justice of, viii; and Madison, 18; as ransom, 214, 221; of serfowners, 192; to serfs, 195; to slave owners, viii, 18, 86, 103–4, 198, 200–211, 216; to slaves, 195, 196; in Sudan, 198 compensation effect, 242 concubinage, 172, 173–74, 259. See also marriage

INDEX

Condorcet, Marquis de, 99–100, 196 Congo, 66 consequentialism, 1, 225, 226, 228, 229, 232, 233, 247 contracts: and competition, 109, 110, 115, 116, 118, 120, 122, 123, 124, 125; freedom of, 135–36; and income, 128–30; involuntary, 113, 127; and labor, 113; labor-service tenancy, 114, 124; between landlords and ex-slaves, 2; linked, 109, 110; and moral hazard, 125; regulation of, 108–9, 110–11, 112, 113, 122, 127, 130, 131, 132, 135–36; structuring of, 110; tied, 109, 110, 122–28, 132, 135; voluntary, 108–37 Cooper, Frederick, 97 Cornish, Samuel, 202 cost: emotional, 15–16; of experience of slavery, 15–16, 20, 29; of redemption, 30; of slave acquisition, 10, 11, 12, 15, 16, 17, 23, 24, 25, 27, 29; to slave holders, 28; of supporting slaves, 23 Côte d’Ivoire, 45, 66 Covey, Edward, 213 Cox, Caroline, 145 credit: advance of, 126–27; for African Americans after emancipation, 90; and child labor, 42, 43; and debt bondage, 66, 67, 70, 71; post-emancipation access to, 122 credit market, 135 Crimean War, 189, 190, 196 Crowley, Peter, 1 cruelty, 254–55 Cuba, 86, 87 Darfur, 147, 151, 196, 259, 261, 263, 266 Deardorff, Alan, 134 debt: and child labor, 40, 41, 114; in French West Africa, 165; payment of, 43; and sale of children, 96; slavery for satisfaction of, 15 debt bondage, 61–73; after abolition of slavery, 122; in agriculture, 113; in ancient Greece, 96; and child labor, 43; defined, 41–42, 61; in French West Africa, 165; liberation from, 61, 62 decision procedure, 226 Declaration of Independence, 202

273 deference, 125–26, 255–56 demand: and child trafficking, 44, 45, 49, 54, 56–57, 58–59, 60, 61; in French West Africa, 162; for redemption, 21, 22; in slavery, 2, 10–11, 12, 13–14, 16, 18, 21, 29; and trafficking, 46 Denmark, 86 Dennis, Michael J., 44 deontology, 1, 225, 226–31, 232, 233 dependency, 49, 82, 209 developed economy, 38, 45 developed world, 40 developing countries, 39, 45, 67 dignity, 256–57 Dinka Chiefs Committee, 154, 156 Dinkas, 9, 30, 34, 145, 147–48, 151, 152, 154, 156, 262, 263, 264 disgust, 241, 242–43, 244, 245, 246–47 distress, 241, 245 distributive inequality, 40–41 Domar, Evsey D., 109, 115, 119, 120, 191; “The Causes of Slavery or Serfdom,” 118 domestic service, 38, 45, 90–91, 259, 260–61. See also labor Dominican Republic, 66, 87 dommum, 251, 252, 253 donko, 251, 252, 253 Douglass, Frederick, 3, 103, 204–5, 208, 211, 213–21, 227, 230, 231, 254, 257; My Bondage and My Freedom, 216 Douglass, John, 231 Du Bois, W.E.B., 93 duty, 225, 226, 232, 233 East Africa, 152, 162 economic development, 69, 70, 77–78, 80–81 economic status, 85, 95 economy: of British West Indies, 89; and child labor, 38; and child trafficking, 54, 55; after emancipation, 86–87, 88, 108, 115, 122, 130, 132–33; growth in, 82, 83, 112; of United States, 88 Ecuador, 45 Edmonds, Eric, 112 education, 34, 39, 45, 62, 90, 102 Egypt, 66 El Salvador, 45 Elster, Jon, 247 Eltonhead, Jane, 200

274 emancipation: in Asante, 252; and contracts, 110; contracts with landlords after, 2, 134; and Douglass, 213; economic consequences of, 86–87, 88, 108, 115, 122, 130, 132–33; experiences after, xi, 3, 85–93, 127–28; forms of, 86; and harms of slavery, 255; immediate vs. gradual, 196; and inferiority, 254; labor after, 256; occupations after, 89, 90–91; output collapse following, 130, 132–33; periods of, 195; respect after, 257, 258; by self-purchase, 103, 175–76, 200, 205, 207, 210, 215; in Sudan, 181; in United States, 88, 89, 193. See also redemption empathy, 238, 239–42, 245 Engerman, Stanley, 32, 108, 118, 133, 195 equality: and American slavery, 210; and Douglass, 216; and emancipation, 86; and land ownership, 128–33, 134; and relationships of women and men, 83 Eradication of Abduction of Women and Children (CEAWAC). See under Sudan Estonia, 45 Ethiopia, 66 ethnicity, 34, 256 eudaimonism, 232, 233 Europe, 37, 45, 46, 95 European Union, 62 exchange, freedom to, 82 exports, 38, 50, 67, 68, 85, 89. See also trade ex-slave(s). See emancipation fairness, 239, 240 familiarity, 237, 238–45 family: African American, 90, 91, 92, 205, 206–9, 211; in Asante, 253; and awowa, 251; cost of slavery to, 15–16; and debt bondage, 63; desire to remain with, 84, 100; disruption of, 149, 155, 156, 254; educational subsidies to, 34; after emancipation, 88, 90, 91, 92; and freedom, 83; in French West Africa, 165, 172, 174; loss of, 30; and partiality, 231; purchase by, 215; redemption initiated by, 2, 151, 215, 263; sale of adults by, 100; and slavery, 254; in Sudan, 2, 149, 151, 155, 156, 260, 263, 266. See also household(s)

INDEX

fear, 244 feeling, 237, 241 female genital mutilation, 83 fertility, 45, 90, 91 Field, Daniel, 190 Fiske, Susan, 244–45 Fogel, Robert William, 132–33 food, 82 Fox-Pitt, T., 167–71 France, 39, 86, 161 Franco, Leonardo, 144 fraud/deceit: in redemption, 14, 146–47; in trafficking, 32, 44, 49; and unfree labor, 112 Frazier, E. Franklin, 93 Frederick Douglass’ Paper, 217 freedom: benefits associated with, 82; capability as, 77, 78; and choice, 79, 81–83, 94; and constraints, 81–82; costs of, 100–101; and development, 78, 80–81; and economic growth, 83; and economic needs, 78; after emancipation, 89; and exchange, 82; as expanding capabilities, 79; and family, 83; in French West Africa, 159, 161, 163–64, 169, 170, 171, 172, 173–74, 175, 176, 177; and functioning, 77; and income, 82, 94; for individual vs. group, 81; legal, 84; legal and social constraints on, 81; meaning of to ex-slaves, 86; and mobility, 89; monetary value on, 144; payment for in United States, 200; purchase of, 200; refusal to accept, 84, 100; and respect, 257, 258; right to, 205; and Russian serfdom, 189–94; self-finance of, 198; to sell labor, 108, 134; and society, 80, 83; and survival, 79; to transact, 82; and voluntary slavery, 94–95, 101; as welfare improvement, 102–3; and women, 79 Free Liberation Army (Morocco), 168 French African Army (tirailleurs), 162, 167 French Sudan, 164 French West Africa (AOF), 158–77 Friebel, Guido, 42, 60 frontier, 113, 118, 119, 122, 127, 133, 135 Fugitive Slave Act of 1850, 207, 210 functioning, 77

INDEX

Gabon, 45 Gambia, 45 Garrison, William Lloyd, viii, 202, 205–6, 208, 211, 213, 219 Garrisonians, 214, 217 gays, 257 gender, 79, 83, 84, 91, 92, 93. See also women General System of Preference (GSP), 62 Genicot, Garance, 110, 135 genocide, 149, 196 Genovese, Eugene D., 101, 254 Georgia, 45 Germany, 46 Gershchenkron, Alexander, 190 Ghana, 66, 249, 250, 256 Gold Coast, 249 Goldin, Claudia, 195 good: alternative, 235, 236, 237, 246, 247; greater, 226, 228, 235; greatest, 237, 240, 246; probability of, 246; and right, 225 good life, 232 Goulimine, 167, 168, 169, 170, 172 Grandy, Moses, 207, 210 Great Britain: abolitionists in, 204; abolition of slavery by, 86, 203, 219–20, 221, 249; child labor in, 39; and Douglass, 214, 215; Emancipation Act of 1834, 218; Factories Act of 1802, 39; and labor, 95; and legal vs. economic status, 95; and Sudan, 160 Greece, 46 Greece, ancient, 96 gross domestic product, 67, 68 gross national product, 50 Grotius, Hugo, 98, 102 Guatemala, 45, 62 Guinea, 66, 174 Guriev, Sergei, 42, 60 Gutman, Herbert, 93 Haiti, 66, 87, 103 Hamilton, Alexander, 9 Hammurabi Code, 96 Hamody, 172 Hart, Allen, 244 Hart, G., 114 health, 77, 79, 80, 94 Hellie, Richard, 98 heteronomy, 257

275 Hobbes,Thomas, 98, 99 Hoch, S. L., 181 Hoffman, Martin, 245 homophobia, 257 Honduras, 45 host countries, for trafficking. See under trafficking household(s): of African Americans after emancipation, 91; and child labor, 39, 40, 43, 67; female-headed, 92; income to from sale of family members, 15, 16; single-parent, 93; two-parent under slavery, 93. See also family hub countries, for trafficking. See under trafficking humanity, common, 80 human rights, 18, 40, 41, 43, 46, 144, 145, 150, 153, 168. See also morality human rights groups, 266–67 Human Rights Watch, 1, 9, 14, 144 humans: commodification of, 1, 227; and excellence, 232; as property, 214, 215; purchase and sale of as wrong, 225–26; self-responsibility of, 257–58; value of, 227. See also morality Humboldt, Wilhelm von, 258 Hume, David, 242 Humphries, Jane, 39 impartiality, 230–31, 237 income: of African Americans after emancipation, 89, 90, 92; and bonded labor, 110–11; and child labor, 39, 110–11; and child trafficking, 45, 54–55; with competitive market vs. tied contracts, 128–30; and debt bondage, 67, 68, 70; and freedom, 82, 94; in free market, 121; to household from sale of family members, 15, 16; and land, 129; low levels of, 79; and model of labor and tenancy, 118; and slave economy, 85; subsistence-level, 82; and voluntary slavery, 94, 95, 96, 101; and welfare, 81. See also wages indentured labor. See labor India, 45, 62, 64, 65, 66, 72, 88, 95, 96–97, 100; Bonded Labor System (Abolition) Act of 1976, 63 Indonesia, 66 infanticide, 95, 96 inferiority, 31, 99, 252, 253, 254, 255

276 internally displaced persons, 151, 154, 155, 156, 259, 266. See also refugees internally displaced persons camps, 151, 155, 263, 264–65 International Labor Organization (ILO), 37, 62, 112; Convention 29 on Forced Labour (1930), 41, 51; Convention 138 on Minimum Age for Admission to Employment (1973), 39; Convention 105 on the Abolition of Forced Labour (1957), 41, 51, 52; Convention 182 on the Elimination of the Worst Forms of Child Labour (1999), 39, 41, 51, 52, 60, 61 Islam, 143, 145, 148, 152, 166, 167, 173, 174, 175–76, 177, 259. See also Muslims Islamic tax (zakat), 174 Italy, 46 Jacobs, Harriet, 254 Jamaica, 87 Japan, 37, 38 Jok, Jok Madut, 27–28, 31 Juman, Muhammad, 261 justice, 226, 236 Kanbur, Ravi, 40 Kansas-Nebraska bill, 217 Kant, Immanuel, 226, 229 Kaufmann, Daniel, 53 Kazakhstan, 45 Kelley, Edmond, 206–7, 208 Kenya, 66, 152 Khartoum, ix, 143, 146, 148, 149, 153, 154, 156, 259, 266 King, Rufus, 218 Kir River, 148, 263 Kliuchevsky, V. O., 118, 127 Knight, Alan, 113 Kohlberg, Lawrence, 247 Kolchin, Peter, 118 Kopytoff, Igor, 97 Kordofan, 147, 151, 259, 263, 266 Koval’chenko, I. D., 190 Kraay, Aart, 53 Kristof, Nicholas, 15 Krueger, Alan, 67 Krugman, Paul, 112 Kunst-Wilson, William, 238 Kyrgyzstan, 45

INDEX

labor: of African Americans after emancipation, 92; Black Code restrictions on, 127–28; bonded, 109, 110–11, 122, 130, 135; bonded child, 37, 40, 43, 61, 62, 63, 64, 65, 70; and child trafficking, 44, 45, 49, 50; choice in, 89; coercion of, 112, 113, 117–22, 130, 133; and competition, 135; and emancipation, 92, 196; forced, 17, 37, 38, 40, 41, 49, 94, 95, 111, 112, 198, 259, 260–61, 265–66; free, 120, 121; freedom to sell, 108, 134; and freed Russian serfs, 192–93; free vs. slave, 95; in French West Africa, 162, 165, 169, 173; gang, 85, 87, 88, 89, 90; in Great Britain, 95; indentured, 15, 100, 109, 111, 113, 122; involuntary, 114, 130, 132; and land, 115, 116, 117, 118, 119, 121, 132, 134; and landlords’ collusion, 124; landlords’ refusal to enslave, 110; and liberty villages, 164; migrant, 45; mobility of, 80, 110, 118, 119, 122, 126, 127–28, 130, 135; and model of labor and tenancy, 116, 117, 118, 119, 120, 121, 124; monopsony rents from, 122, 131, 132, 133; opportunity restrictions on, 110, 125–26, 131; post-emancipation productivity of, 87, 88, 130, 132–33; regulation of, 34, 130, 132; rights of, 66–72; scarcity of, 117–22; and serfdom, 122, 125, 183–84, 186, 187, 188, 189; servile, 114; shortage of, 127; slave, 21, 62; standards for, 66, 67; stigmatization of, 255, 256; in Sudan, 198, 259, 260–61, 265–66; and tenancy, 109, 113–14, 126; tied, 113, 122–28, 129, 130; unfree, 108, 111–14; voluntary, 41, 109, 110, 114, 122, 132; and voluntary contracts, 108–37; and voluntary slavery, 101; of women in slavery, 92. See also market labor unions, 63, 64, 65, 70, 71, 73 land: and African Americans, 90; Black Code restrictions on, 127; and contracts, 115; and contract structuring, 110; and freed Russian serfs, 192, 193, 197; in Ghana, 250; and income, 129; and labor and tenancy model, 115, 116, 117; landlords’ encroachment on, 133; post-emancipation ownership of, 87,

INDEX

89, 90, 122, 123, 218; property rights over, 119; ratio of to labor, 115, 116, 117, 118, 119, 121, 132, 134; restricted leasing of, 131, 132; and Russian serfdom, 181, 184, 194, 195; and Smith, 217; unequal ownership of, 128–33, 134. See also market; property landlord(s): and child labor, 41, 43; collusion by, 109, 110, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137; competition between, 126; and consent to hire outsiders, 126; and contracts with ex-slaves, 2; and debt bondage, 42, 62; encroachment on peasant land by, 133; and labor and tenancy model, 115, 116, 118, 119, 120, 121; and land-to-labor ratio, 134; and moral hazard, 125, 126; refusal to enslave labor by, 110; restricted leasing of land by, 131; and restriction of labor opportunities, 110; and serfdom, 122, 125; and tied labor, 122, 129, 130, 137; violence by, 119 Lane, Lunsford, 207 Laos, 45, 66 latifundia, 130 Latin America, 37, 38, 113, 130, 133, 256 Lavigerie, Cardinal, 162 law(s): and American slavery, 202, 207, 210, 230, 255–56; and child labor, 39, 41, 61, 73; and child trafficking, 44, 46, 51–53, 58–59, 60–61; and debt bondage, 63–64, 65, 66, 67–68, 71; and deference, 255; and Douglass, 230; and emancipation, 88; enforcement of, 53, 60, 61, 73; and ethics, 232; and forced labor, 94; and freedom, 81, 84; in French West Africa, 159–60; Islamic customary, 166; in Mauritania, 166, 167; against prostitution, 60, 61; in Russian serfdom, 122; and slave owners, 196; and slavery, 102, 103, 196, 202, 207, 210, 230, 255–56; status under, 52, 60, 83, 85, 95; in Sudan, 196, 198, 230; and women and child trafficking, 51–53, 58–59, 60–61; and women’s rights, 83; and wrongdoing, 228 League of Nations, 164 Ledoux, Joseph, 244

277 Leon XIII, 162 lesbians, 257 Levitt, S. D., 60 The Liberator, 202–3, 204, 205–6, 211 Liberia, 62, 66, 201–2 Liberty party, 216 liberty villages (villages de liberte), 161–65, 166, 176 Libya, 66 life expectancy, 77, 80, 82, 89 Livingstone, David, 162 Locke, John, 99 Lovejoy, Paul E., 97 Lual, Mawien Garang, 264–65 Lyashchenko, P. I., 190 Machina, Mark L., 191 MacIntyre, Alasdair, 232 Madison, James, 18 Madison, Joe, 144–45 Madison County, New York, 217 Maine, Henry Sumner, 78n5 Majok, Akuol, 263–64 Malawi, 45, 66 Maldives, 62 Mali, 66, 160, 161, 167, 174 manumission. See emancipation manumission societies, in United States, 201 market: and American slavery, 204–5; child labor, 40, 42, 43, 49, 50, 62, 63, 73; constraint imposed by, 81; and Douglass, 227; financial, 67, 70; free, 121; labor, 62, 109, 110, 114, 116, 117, 118, 119, 120, 121, 122, 123, 125, 126, 128–29, 130, 131, 132, 133, 135, 137, 163, 197, 253; land, 109, 114, 115, 116, 117, 118, 121, 122, 130, 131, 133, 137, 193, 194, 195; manipulation of, 110; slave, 11, 12, 15, 16, 17, 18, 20, 21, 22, 23, 25, 27, 28, 143, 159, 165, 204, 215, 216, 227, 251, 254; tenancy, 132, 133 market imperfections, 2, 39, 43, 66–70, 73 marriage: abusive, 94; of African Americans after emancipation, 90–91; forced, 261, 263; in French West Africa, 172; in Mauritania, 166, 167; rights in, 83. See also concubinage; gender Marx, Karl, 109, 119n Marxism, 80, 130, 190, 194

278 materialism, 220–21 Maugham, Robin, 161, 167, 175; Slaves of Timbuktu, 159–60 Mauritania, 2, 66, 160, 161–77 Mayans, 97 Mecca, pilgrimage to, 174, 175 Mendelsohn, Isaac, 96 Mexico, 45 Middle East, 37, 38, 96 Miers, Suzanne, 97 migrants, 42, 44, 46, 49, 51, 52 Mill, John Stuart, On Liberty, 258 Miller, William, 247 Missouri Compromise, 217 mobility. See labor Montesquieu, 100 Moon, David, 190, 197 moral hazard, 125, 126 moral intuition, 229 morality: and child labor, 40–41; Douglass on, 220–21; in French West Africa, 164; and Garrison, 213; and impartiality, 230–31; and passion, 242; of redemption, 225–34, 235–47; and sale of family members, 15; and slavery, 3, 202–3, 204, 208, 210, 211, 249–58; in Sudan, 230; and wrongdoing, 228. See also human rights; humans; slavery, redemption as promoting moral motivation, 229 moral prescription, 3, 235, 236, 237, 242, 245, 246, 247 moral reasoning, 238, 239, 241, 242, 247 Morocco, 66, 160, 161, 167–70 Moynihan Report, 92 Mozambique, 45, 66 multinational firms, 40 Murahileen militia, 149, 150, 259, 260 Muslims, 31, 159, 160, 173, 174, 175–76. See also Islam Myanmar, 45, 62 Namibia, 66 Nash bargaining, 24 National Compensation Convention, 218–19 National Liberty party, 216 Near East, ancient, 96 Nechkina, M. V., 190 Nepal, 45, 62, 64, 65, 66; Napal Labor Act, 63

INDEX

Netherlands, 86 Never Reward Someone for Engaging in Intentional Wrongful Actions (NRW) principle, 228, 229 Nicholas I, 190 Nieboer, H. J., 80, 109, 118, 119, 120 Nieboer-Domar hypothesis, 115, 117–22 Niger, 160n3 Nigeria, 45, 66 North, Douglas, 113, 114 North Africa, 38 Northrup, David, 113 The North Star, 205 Nuadom, 252–56 Nussbaum, Martha, 78 obligation, 232, 236 opportunism, 247 opportunity, absence of, 80 opportunity costs, 29 outsider, 80 Pacific, 38 Pakistan, 62, 63–64, 65, 66; Bonded Labor System (Abolition) Act of 1992, 63; Constitution of 1973, 63 parents, 40, 82, 96–97, 98, 99, 112, 114 paternalism, 254, 255 Patterson, Orlando, 101, 118 Pavcnik, Nina, 112 pawning, 165 payment, and Russian serfdom, 184, 186, 187, 193, 194, 195 Payne, Francis, 200 peasants: hiring of outsiders by, 126; in labor and tenancy model, 115, 116, 117, 118, 124, 125; and landlords’ collusion, 124; landlords’ encroachment on land of, 133; and land-to-labor ratio, 134; mobility of, 130; post-emancipation situation of, 123; and Russian serfdom, 183, 184, 185, 191, 193–94 Pease, Elizabeth, 208 Peel, Robert, 39 Pennsylvania Abolition Society, 9 Perbi, Akosua, 250–51 per capita income, 54, 70 Peru, 66 Phelps, Elizabeth, 244 Piaget, Jean, 247 Picquet, Louisa, 209–10

INDEX

plantations, 84, 87–88, 89, 119, 130, 249, 250, 254. See also agriculture Platonism, 229 Plymouth Church, Brooklyn, 210 Pokrovskii, M. N., 190, 191 police, 53, 60, 263, 264 politics, 90, 214, 216 politics of recognition, 257 pomeshchik serfs, 183–88 Popular Defense Forces (PDF) (Sudan), 149, 150, 260 pornography, 44, 51 Portugal, 250 Portuguese islands, 88 postcolonialism, 160 poverty, 18, 39–40, 43, 44, 60, 62–63, 71, 78–79, 96, 111, 134, 151, 217 preference bias, 238–39 price: in French West Africa, 161, 163, 164, 165, 169, 174, 175; on human beings, 200; information about, ix–x; and sex trade, 61; and slave owners, 13; of slaves, ix–x, 11, 12, 13, 14, 16, 17, 18, 20, 21, 22, 23, 24, 27–28, 29, 31, 85, 91, 103, 216 Price, Enoch, 205 profit, 118, 121, 165 property, 127, 214, 215. See also land property rights, 120, 130, 133, 134 prostitution: of children, 15, 37, 40, 41, 44, 50, 51, 72; laws against, 60, 61; in Mauritania, 167; trafficking for, 45; voluntary entry into, 32. See also sex industry, children in Protection Project, 46 Prussia, 39, 130, 133 psychology, 245, 246, 247, 257 public works programs, 71–72, 73 Puerto Rico, 86 Pufendorf, Samuel, 98, 99, 102 pull and push factors, in trafficking. See trafficking Quakers, 1, 3, 9, 201 Qu’ran, 176 race, 80, 90, 147, 148, 160, 256 rachat, 158–77; defined, 161. See also redemption racism, 257 Railton, Peter, 247

279 ransom, 165, 214, 221, 227–31 Rather, Dan, 146 Rawls, John, 226, 229, 256–57 redemption: alternatives to, 34–35; and alternative ways of doing good, 235–47; in Asante, 252; benefit of, 30; and child labor, 37–73; cost of, 30; data on, x–xi, 17, 31–32; and debt bondage, 70, 73; definitions of, 102; demand for, 21, 22; as diversion of resources, 17; and Douglass, 213–21, 227; as factor in market, 25–27; family-initiated in Sudan, 2, 151, 263; fraud in, 14, 146–47; for freeing of slaves, 9; in French West Africa, 159, 162; good vs. harm in, 225, 235, 236; Jok on, 27–28; by locals vs. outsiders, 27–28; by Madison, 145; matching model of, 20, 21–30; morality of, 225–34, 235–47; and partiality, 231; as promoting slavery, 1, 9, 10, 14, 15, 17, 28, 30, 31, 103, 143–44, 147; as promoting trafficking, 32; as raising opportunity costs, 29; as ransom, 227–31; as reward for crime, 103; for Russian serfs, 193–94; in Sudan, xi, 2, 31, 143–57, 181, 227, 229–30, 231, 232, 233–34, 260; as supererogatory, 236; and supply vs. demand in slaves, 9–18; time factor in, 2, 20, 21–22, 23, 25, 26; in United States, 200–211. See also emancipation; rachat refugees, 34, 163. See also internally displaced persons relief services, 151 religion, 160, 201 reparations, 94 responsibility, for self, 257–58 Richardson, Ellen, 214 Rieber, Alfred J., 190 Rift Valley Institute, x, 152–53 rights, 82, 233; of children, 83, 96; of freed ex-slaves, 85; in moral thinking, 232; respect for, 229; of slaves vs. slave owners, 86; trade-offs between, 78. See also human rights right vs. good, 225 Robinson, James A., 110, 135 Rogers, C., 42, 55 Rome, ancient, 96 Rozin, Paul, 243

280 rural population, 62, 63, 65, 71–72, 73 Russia: Emancipation Statutes of 1861, 191, 192, 193; government of, 184, 187, 188, 192, 193, 194; postemancipation economy of, 133; serfdom in, 135, 181–98; and Sudan, 194–98; voluntary slavery in, 98, 100 Rutherford, Thomas, 102 Satz, Debra, 40 Saudi Arabia, 45 Save the Children (SC-UK), 143, 147, 153, 155, 156 savings, 62, 63 Sears, David, 239 Seiberg, Katri, 61 self-made man, 209 self-respect, 256–57, 258 sellers: of child labor, 43; and Douglass, 216; of slaves, 22–24; and voluntary slavery, 80. See also slave dealers; slavery; trafficking Sen, Amartya, 77–78, 79, 83, 93, 108; Development as Freedom, 77, 80–81, 82, 94 Senegal, 167, 174 serfdom, 41, 86, 100, 103, 111; abolition of, 122, 123; chronology of, 182–83; as contractual adaptation, 113; Domar on, 118, 119; evolution of, 127; freedom from as human right, 43; and labor mobility, 126; mechanics of, 181–89; in Nieboer-Domar hypothesis, 115; in Russia, 118, 122, 135, 181–98; service for protection in, 113; and tied labor contracts, 122–28 serfowner, Russian, 181, 183–88, 189, 192 serfs, Russian, 183–89, 195, 197 sex industry, children in, 15, 33n, 34, 37, 40, 41, 44, 45, 50, 51, 61, 72. See also prostitution sexual assault, 30, 254, 260, 261, 263, 266. See also violence sexuality, 209–10, 257 sharecropping, 113, 193, 197 Sharpton, Al, 144 Sidgwick, Henry, 237, 240, 242 Sierra Leone, 45, 66 Silver, Morris, 96 Singer, Peter, 235, 236, 246

INDEX

skin color, 209–10, 255, 256 slave dealers: costs of, 10, 11, 12, 15, 16, 17; in matching model, 22–25, 26. See also sellers slave economy, 85, 120, 121, 123, 136–37 slave labor, 21, 62 slave masters: in French West Africa, 159, 161, 162, 163, 164, 165, 166, 169, 171, 173, 174, 175, 177; recovery of slaves by, 159, 161, 162, 164, 165, 166, 169, 171 slave owners: and chattel slavery, 227; coercion by, 79; compensation to, viii, 18, 86, 103–4, 198, 200–211, 216; cost to, 13, 16, 28; identity of, 216; and law, 196; legitimacy of claim of, 204; negotiation with slaves by, 85, 86; power over slaves of, 254, 255; property rights of, 195, 201; rights of, 86; as self-justifying, 30–31; as sellers, 22 slave raider, 9, 14, 22, 23, 24, 25, 26, 28, 29, 30, 31, 32, 33, 34 slavery: abolition of, 9, 93, 122, 123; abuse under, 260, 261; and Asante, 249–50, 251, 252–53; British abolition of, 86, 203, 219–20, 221, 249; and capability, 78; chattel, 10; and child labor, 41; and children, 254; as crime, 103, 211; cruelty in, 254–55; for debt satisfaction, 15; defenders of, 30–31, 95, 254; definition of, 4; Domar on, 118, 119; Douglass on, 221; and economic growth, 85; experience of, 14–15, 20, 29, 30–32, 259–67; and family, 84, 254; forms of, 79–80, 227; in French West Africa, 159–77; and General System of Preference, 62; and heteronomy, 257; and human rights, 43; and humans as property, 214, 215; and laws, 102, 103, 196, 202, 207, 210, 230, 255–56; morality of, 3, 202–3, 204, 208, 210, 211, 249–58; nature of, 147; in Nieboer-Domar hypothesis, 115, 120; and outsiders, 80; and poverty, 78–79; as preferred to death, 79; prevention of, 103; and race, 80; rate of flow in, 2, 18, 22, 28, 29, 32, 33; Rawls on, 226; redemption as promoting, 1, 9, 10, 14, 15, 17, 28, 30, 31, 103, 143–44, 147; refusal to leave, 84, 100; roots of, 18; and self-respect,

INDEX

257; and sex, 30, 209–10, 254, 259, 261; and social status, 256; and society, 80; in Sudan, ix, 9, 30–32, 143–57, 181, 227, 229–30, 231, 232, 233–34, 259–67; supply and demand model of, 2, 9–18, 20–21, 22, 26, 29–30; transition plan for abolishing, 18; and twoparent household, 93; in United States, ix, 78, 81, 83–84, 88, 94, 100–101, 119, 121–22, 127–28, 133, 200–211, 227, 230, 254; and United States Constitution, 213, 217; as unjust, 226; value of, 22; violence in, 30, 34, 261–62, 263; voluntary, 32, 34, 79–80, 94–104; and war, 98, 99, 147 slave(s): abuse of, 254; bargaining power of, 215; buyers of, 22, 24; cost of, 10, 11, 12, 15, 16, 17, 23, 24, 25, 27, 29; demand for, 2, 10–11, 12, 13–14, 16, 18, 21, 29; donko as, 251; family’s redemption of, 2, 151, 215, 263; freedom of, 9; in French West Africa, 159–60; fugitive, 83–84, 103, 204, 205, 207, 262, 263; inheritance of status as, 101; inherited, 166, 167; lower status of, 257; motivation of, 120; by nature, 99, 256; negotiation ability of, 85, 86; number of freed, 14; number of in captivity, 3–4, 9, 10, 12, 13, 14, 15, 16, 17, 18, 20, 21, 22, 24–25, 27, 28, 29, 30, 33; number of in production, 28; number of redeemed, 29; output of, 11; post-emancipation freedom of, 86; post-emancipation land for, 87, 89, 218; post-emancipation rights of, 85; price of, ix–x, 11, 12, 13, 14, 17, 18, 20, 21, 22, 23, 24, 27–28, 29, 31, 85, 91, 103, 216; as property, 120; property rights of, 195; purchase of, 225; recapture of, 12; recovery of in French West Africa, 159, 161, 162, 164, 165, 166, 169, 171; rights of, 86; self-purchase by, 103, 175–76, 200, 205, 207, 210, 215; sellers of, 22–24; slave owners’ power over, 254, 255; supply of, 10–11, 12–13, 17, 18, 21, 29; as term, 256; testimonies of, 259–67; value of, 25. See also emancipation; market slave society, 86 Smith, Adam, 110, 120, 121n6

281 Smith, Gerrit, 216, 217–19 Smith, James McCune, 220 Soares, Rodrigo, 60 society, 80, 83, 86 Society for Promoting the Manumission of Slaves, 9 Somalia, 45, 66 source countries, for trafficking. See under trafficking South Africa, 45 South America, 45 Southeast Asia, 15 Southerners, American, 201, 209 South Sudan, 30, 31, 143, 144, 146, 147, 149, 150, 154, 155, 157, 266 Soviet Union, former, 45 Spain, 86 Sri Lanka, 45, 66 Stampp, Kenneth M., 101 standard of living, 77, 80, 81, 82, 83, 85, 94, 95 Starr, Kenneth, 145 status: legal, 52, 60, 83, 85, 95; social, 3, 250–53, 254, 255, 256, 257. See also stigma Stauffer, John, 231 Ste. Croix, G. E. M. de, 96 stigma, 15, 255. See also status strangers, 237, 239, 244 subordination, 250–53 Sub-Saharan Africa, 37, 38 Sudan, 66, 164; attention on, 160; children in, 149, 150, 151, 153, 154, 155, 259, 260, 261–63, 266; and child trafficking, 45; cost of slave acquisition in, 15; Eradication of Abduction of Women and Children (CEAWAC), 143, 153, 154, 155, 156, 262, 263; ethnicity in, 34; family in, 2, 149, 151, 155, 156, 260, 266; government of, x, 143, 144, 145–46, 147, 148, 150, 151, 152, 153, 154, 196, 197, 259, 260, 266; labor in, 197, 198, 259, 260, 260–61, 265–66; law in, 196, 198, 230; morality in, 230; Popular Defense Act, 150; redemption in, xi, 2, 31, 143–57, 181, 227, 229–30, 231, 232, 233–34, 260; and Russia, 194–98; slavery in, ix, 9, 30–32, 143–57, 181, 227, 229–30, 231, 232, 233–34, 259–67; war in, 2, 148, 150, 153, 154, 259, 266, 267;

282 Sudan (continued) women in, 149, 150, 154, 155, 259, 260, 263–64, 266 Sudan Mission to the United Nations, 144, 145 Sudan People’s Liberation Army (SPLA), 146, 149, 150, 151, 152, 153, 156 Sudan People’s Liberation Movement (SPLM), 155, 156 supply: in economic modeling of slavery, 2, 10–11, 12–13, 17, 18, 21, 29; elasticity of, 2, 13, 15, 16, 17; in French West Africa, 165; in trafficking, 44, 46, 49, 54, 56–57, 58–59, 60, 61 survival, 79, 82, 100 sweatshops, 38, 40 Swinnerton, Kenneth, 42, 55 sympathy, 147, 241, 245 Tappan, Arthur, 216 taxes, and Russian serfs, 184, 187 Taylor, Charles, 257 technology, 80, 111, 114, 115, 119, 120, 128, 136 teleological ethics, 225 Temkin, Larry, 236 tenancy, 109, 113–22, 124 Teresa, Mother, 245 Thailand, 45, 50, 66 Thomas, Robert, 113, 114 Timbucto, 217, 218 Timbuktu, 159, 160, 161, 162, 167, 170, 175 time, and redemption, 2, 20, 21–22, 23, 25, 26 tirailleurs. See French African Army (tirailleurs) Togo, 66 Toumanoff, Peter G., 191 tourism, 45, 50–51 trade, 50, 51, 61. See also exports trafficking, 32, 34; and age, 44; and bargaining power, 43; of children, 40, 41, 44–61; on cross-national basis, 42–43; definition of, 44; fraud in, 32, 44, 49; host countries for, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 58, 61; hub countries for, 46, 47, 48, 49, 50, 52, 53; of illegal immigrants, 42; and laws, 44, 46, 51–53, 58–59, 60–61; for prostitution, 45; pull and push factors

INDEX

in, 44, 46, 49, 54, 56–57, 58–59, 60, 61; source countries for, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56–57, 58–59, 60, 61; of women, 32, 44–49, 51–53 Trinidad, 87 Tutu, Osei, 249 Uganda, 66, 152 Ukraine, 45 unfamiliarity, 244 United Arab Emirates (UAE), 45 United Nations: and child labor, 41; Convention on the Rights of the Child (1989), 39, 44; International Agreement for the Suppression of White Slave Traffic (1904), 41; and Mauritania, 166; Migrant Workers Convention, 51, 52; protocols on elimination of women and child trafficking, 51, 52, 60; Slavery Convention (1926), 64; and Sudan, 31, 145; Supplementary Convention on Slavery (1926), 64; Supplementary Convention on the Abolition of Slavery (1956), 41; Survey of Crime Trends, 53 United Nations Children’s Fund (UNICEF), 1, 9, 143, 144, 147, 153, 155, 156 United Nations Commission for Human Rights, 144 United States: abolition of slavery in, 86, 196, 197; agriculture in, 88, 90; child labor in, 39; children in, 84; and debt bondage, 62; emancipation in, 88, 89, 193; and legal vs. economic status, 95; living standard in, 81; postemancipation economy of, 130, 133; post-emancipation period in, 88, 89, 90–92, 93; segregation in, 197; slavery in, ix, 78, 81, 83–84, 88, 94, 100–101, 119, 121–22, 127–28, 133, 200–211, 227, 230, 254; and Sudan, 143, 144, 146, 152; women in, 84, 207, 209–10 United States Constitution, 127, 201, 213, 217, 255, 256 United States Department of State: Human Rights Report, 64; Trafficking in Persons Report, 46 urban population, 38 utilitarianism, 237

283

INDEX

utility, 226 Uzbekistan, 45 value: of American slaves, 204; and human life, 227; in matching model, 22, 25, 26, 27, 28; prior ranking of, 225. See also cost; price Van, Pham Hoang, 39 Vic, Karl, 146 vice, 233 Vietnam, 45 villages de liberte. See liberty villages (villages de liberte) violence: and Douglass, 213–14, 216, 221, 254; and Garrison, 213; by landlords, 119; in slavery, 30, 34, 260, 261–62, 263, 266; and Smith, 218. See also coercion virtue ethics, 231–34 voluntary labor, 109, 110 voluntary prostitution, 32, 33n voluntary servitude, 108, 230 voluntary slavery, 32, 34, 79–80, 94–104 wages: of African Americans after emancipation, 91; and bonded labor, 110–11; and child labor, 39, 42, 110–11; and child trafficking, 44; and debt bondage, 62, 63, 67, 70, 71; and labor and tenancy model, 117, 118, 119; and land-to-labor ratio, 132; and moral hazard, 125, 126; and output collapse following emancipation, 130; for slave labor, 120; and tied contracts, 132. See also income Wakefield, E. G., 109, 118, 119 Walker, David, Appeal, 202

Walker, James, 206 Walker, Paralee, 206 war: captive of, 251; and child trafficking, 45; and slavery, 98, 99, 147; in Sudan, 2, 148, 150, 153, 154, 259, 266, 267 Washington, Madison, 213 wealth, 90, 96 Webster, Daniel, 218 welfare, 29, 30, 34, 35, 77, 81, 95, 110, 112 West Africa, 250 Wheeler, Mary, 244–45 White Fathers, 163 whites, American, 88, 91, 121–22, 200, 202, 204, 205, 208, 255 Williams, Bernard, 232 Wol, Ariath Ariath, 266 women: and abusive marriages, 94; and American slavery, 84, 207, 209–10; in Asante, 252; domination of, 92; after emancipation in United States, 89, 90–91, 92; and freedom, 79; legal rights of, 83; in Mauritania, 166, 167; and men, 82, 83; and rational choice, 83; respect for, 257; and sex trade, 34; in Sudan, 149, 150, 154, 155, 259, 260, 263–64, 266; trafficking of, 32, 44, 45, 46–49, 51–53; violence against, 30. See also gender Wright, Henry C., 227 wrong, benefit from, 228, 229 Zajonc, Robert, 238, 244 zakat (Islamic tax), 174 Zeinab, 263 Zionism, 146, 147, 152 Zoido-Lobaton, Pablo, 53