Anthropology, Economics, and Choice

Anthropology, Economics, and Choice

THIS PAGE INTENTIONALLY LEFT BLANK

Anthropology, Economics, and Choice

michael chibnik

University of Texas Press   Austin

Copyright © 2011 by the University of Texas Press All rights reserved Printed in the United States of America First edition, 2011 Requests for permission to reproduce material from this work should be sent to: Permissions University of Texas Press P.O. Box 7819 Austin, TX 78713-7819 www.utexas.edu/utpress/about/bpermission.html ♾ The paper used in this book meets the minimum requirements of ANSI/NISO Z39.48–1992 (R1997) (Permanence of Paper).

Library of Congress Cataloging-in-Publication Data

Chibnik, Michael Anthropology, economics, and choice / Michael Chibnik.   p.  cm. Includes bibliographical references and index. iSbn 978-0-292-72676-5 (cloth : alk. paper) — iSbn 978-0-292-72902-5 (pbk. : alk. paper) — iSbn 978-0-292-73535-4 (e-book) 1. Economic anthropology—Case studies. I. Title. Gn448.c47 2011 306.3—dc23 2011019003

Contents

Preface vii Acknowledgments ix

Introduction 1

1.

How Important Is Decision Making? 20

2. Choices between Paid and Unpaid Work 38 3. Risk, Uncertainty, and Decision Making 60 4. Experimental Games and Choices about Cooperation 90 5. Who Makes Household Economic Decisions? 118 6. Is There a Tragedy of the Commons? 142 Conclusion 164 Notes 173 References 177 Index 197

THIS PAGE INTENTIONALLY LEFT BLANK

Preface

During 2008–2009 the world’s economy was mired in a terrible recession. As nations struggled with rising unemployment, failing businesses, massive deficits, and plummeting stock prices, their angry residents wondered what had caused these problems and how they might be resolved. Economists were often in the news during this time. Some commentators blamed economists for promoting policies that led up to the crisis; others looked to them for predictions about what might happen next. Prominent economists wrote books and columns analyzing the recession and proposing paths to recovery. When most economists offer explanations of events such as the global recession of 2008–2009, they are guided by their discipline’s ideas about how individuals and groups allocate scarce resources. Many are influenced by rational choice theory, which assumes that such decisions are made by well-informed, intelligent people who consciously or subconsciously weigh the risks, costs, and benefits of alternative actions. Yet we all know that decision makers are often muddled about what they want, confused about the possible costs and benefits of different choices, and uncertain about what the future might bring. Economic anthropology is a small subfield of a discipline that is much less influential than mainstream economics. Nonetheless, the heterodox views of economic anthropologists present an important, iconoclastic challenge to conventional ways of looking at choice. These scholars think that theories based on rational choice rely too much on elaborate mathematical models that make dubious assumptions about a limited set of variables and pay too little attention to the context and complexities of real-world decisions. Most economic anthropologists analyzing decision making emphasize how historical changes, cultural norms, and socioeco-

viii Anthropology, Economics, and Choice

nomic institutions constrain the choices possible for different groups of people at particular times and places. The heart of this book consists of a comparison of the ways that different social sciences—particularly economics and anthropology—have looked at five important issues in the analysis of economic decision making: choices between paid and unpaid work; ways people deal with risk and uncertainty; how individuals decide whether to cooperate with one another; the extent to which households can be regarded as decision making units; and the celebrated “tragedy of the commons.” These issues involve basic questions in the social sciences such as the distinction between “use value” and “exchange value,” the extent to which human beings are altruistic, the ways in which societies attempt to limit the untrammeled pursuit of self-interest, the nature of “rationality,” and the degree to which “economics” and “culture” can be analytically separated. In recent years, economists have become increasingly confident that their methods can aid both in the understanding of such general questions and in the solving of down-to-earth practical problems. Although I do not share the economists’ confidence in the usefulness of their theories, I make no claims to have provided definitive answers to any of the questions that I raise in this book. My goal is to provide accessible explanations, as jargon-free as possible, of the ways that different groups of scholars have thought about some fundamental, intriguing questions about decision making. I have included numerous ethnographic examples of decision making from my own research. I do so because I think that the pluses and minuses of abstract theories can most easily be understood and assessed by looking carefully at their applicability (or lack of applicability) to real-world examples. I have been an economic anthropologist for almost four decades. It would be natural to assume that my research methods incorporate techniques drawn from economics. This was certainly the case at the outset of my career when I carried out doctoral fieldwork in Belize. With a background in mathematics and a commitment to anthropology as a science, I attempted to apply models from economics to the agricultural decisions I studied. Although these models were useful, I found that they did not help my understanding of important aspects of what I observed. During subsequent research in Peru, Mexico, and United States, I became increasingly convinced of the value of ethnographic approaches to choice. I now think that anthropologists and economists see the world in very different ways. This book begins by exploring why this is so.

Acknowledgments

In 1976 I applied for a postdoctoral position in the Department of Anthropology at the University of California, Berkeley. I proposed to use my time as a postdoc to write a book about some knotty problems in economic and ecological anthropology. These included the monetary value of subsistence production, decision making under risk and uncertainty, the effects of household composition on agricultural decision making, and the ways that societies attempt to prevent environmental damage caused by individuals pursuing their “rational” self-interest. Although I spent a worthwhile, enjoyable year as a postdoc at Berkeley, I unsurprisingly did not come close to writing the ambitious book I had proposed. Nonetheless, in the ensuing years I have continued to conduct research on issues related to those discussed in my postdoc application. Much of what I have written here consists of my current thoughts about these issues. This book is quite different from the one I planned to write all those years ago. After conducting fieldwork in diverse field sites, I have become more convinced of the value of ethnography and history and less willing to embrace the parsimony of mathematical models of human behavior. Theories have come and gone as anthropologists and economists in new research settings have been forced to revise their ideas about decision making. I wrote this book between 2008 and 2010. For the most part I did this alone, consulting with only a few people. But in a larger sense, the book is the outcome of more than three decades of research and teaching. Here I can mention only a few of the many institutions and people that have helped me. The writing of this book and much of the research discussed here have been funded by Career Development Awards and other kinds of insti-

x Anthropology, Economics, and Choice

tutional support from the University of Iowa. I have also received important support over the past several decades from Columbia University, the University of California–Berkeley, the National Science Foundation, the National Institute of Mental Health, the Welte Institute for Oaxacan Studies, and the Federación de Campesinos de Maynas in Peru. My biggest debt is to all the people in the field who have listened to my never-ending questions about their lives. The following people have also helped me greatly in various ways: Shirley Ahlgren, Saúl Aragón, Peggy Barlett, Gerald Britan, Carole Browner, Holly Carver, Katharine Chibnik, Jeffrey Cohen, Beth Conklin, Rudi Colloredo-Mansfeld, Wil de Jong, Gudrun Dohrmann, Virginia Dominguez, Susanna Donaldson, Nora England, Arthur Goldhammer, Paul Greenough, Lourdes Gutiérrez Nájera, Charles A. Hale, Angelique Haugerud, Jorge Hernández, Douglas Hertzler, Brandi Janssen, Allen Johnson, Ann Kingsolver, Theresa May, Mark Moberg, Arthur Murphy, Tad Mutersbaugh, Tomomi Naka, Robert Netting, Christine Padoch, Charles Peters, Miguel Pinedo-Vásquez, Beverly Poduska, Silvia Purata, Martha Rees, Cerisa Reynolds, Tana Silva, Christine Szuter, Amy Todd, Stephen Tulley, Victor Vásquez, Lois Wasserspring, Ronald Waterbury, and Jim Weil. While writing this book, I did much of my background reading and editing in the coffee shop of Prairie Lights Bookstore in Iowa City. This comfortable space is a wonderful refuge from economic downturns, midwestern floods, and university politics. Chapter 5 is a substantially revised and expanded version of a previous publication: Experimental Economics in Anthropology: A Critical Assessment. American Ethnologist 32 (2) [2005]: 198–209.

Introduction

Three decades ago Gary Becker wrote a book in which he made extraordinary claims about the usefulness of economic approaches for the understanding of questions in social sciences. In a now-famous introductory essay to The Economic Approach to Human Behavior (1976), Becker succinctly states his views: I have come to the position that the economic approach is a comprehensive one that is applicable to all human behavior, be it behavior involving money prices or imputed shadow prices, repeated or infrequent decisions, large or minor decisions, emotional or mechanical ends, rich or poor persons, men or women, adults or children, brilliant or stupid persons, patients or therapists, businessmen or politicians, teachers or students. (8)

By “the economic approach” Becker means one in which “all human behavior can be viewed as involving participants who maximize their utility from a stable set of preferences and accumulate an optimal amount of information and other inputs in a variety of markets” (14). He asserts that his approach provides a “unified framework for human behavior” that was sought by but eluded such renowned thinkers as Jeremy Bentham, Auguste Comte, and Karl Marx. In modified form, the types of analysis advocated by Becker have become increasingly influential in the social sciences. In this book I examine the advantages and disadvantages of this “economic” approach to diverse practical problems involving decision making. I contrast this approach with “anthropological” analyses that focus on the historical changes, cultural norms, and socioeconomic institutions that constrain the choices

2 Anthropology, Economics, and Choice

possible for different groups of people at particular places and times. My principal argument is that the methods of economics alone are insufficient for understanding the complexities of choice.

Economic and Anthropological Approaches to Choice Becker’s book, filled with graphs and equations, was written primarily for economists. His definition of an “economic approach” would be likely to mystify noneconomists unsure about the technical meanings of “utility,” “preferences,” “inputs,” and “markets.” Even readers familiar with the language of economics might wonder how anyone could possibly know when utility is being maximized, what it means to say that preferences are “stable,” and what exactly is an “optimal” amount of information. In 1992 Becker won the Nobel Memorial Prize in Economic Sciences “for having extended the domain of microeconomic analysis to a wide range of human behaviour and interaction, including nonmarket behaviour.”1 As Becker hoped, economic approaches have become prominent in fields such as political science and are now also of some importance in sociology, geography, and history. Such approaches are often lumped together under the term “rational choice theory,” a reference to their assumptions about utility maximization. An enormous scholarly literature (among others, Elster 1989, Gigerenzer 2008, Parsons 2005, Schweers Cook and Levi 1990) describes rational choice theory and debates its merits. There are also numerous books aimed at general readers that popularize the basic ideas of the theory. Some of these books (for example, Becker and Becker 1997; Harford 2005, 2008; Levitt and Dubner 2005) describe insights gained from applications of rational choice theory; others (including Gladwell 2005, Groopman 2007, Lehrer 2009, Thaler and Sunstein 2008) pay more attention to the theory’s limitations. Many of the critics of economic approaches in social sciences emphasize findings from experiments in cognitive psychology that show that people often make suboptimal choices. Rational choice theories have certain features that make them attractive to scholars advocating “scientific” approaches to social phenomena. Their focus on a restricted number of key variables can allow hypotheses to be tested in controlled situations. The mathematical elaboration of these theories provides clear statements of suggested relationships among different variables. Many ethnographically oriented anthropologists, however, are repelled by the very features that make rational choice

Introduction 3

theories appealing to many economists and political scientists and intriguing—if debatable—for many psychologists. The reliance of rational choice theories on a limited number of assumptions is diametrically opposed to an entrenched holistic anthropological tradition that stresses multiple interrelated influences on cultural practices. The underlying idea that humans everywhere have similar motivations conflicts with anthropologists’ longtime emphasis on cultural diversity. The methodological individualism typical of most applications of rational choice theory is unacceptable to both anthropologists focusing on supra-individual characteristics of groups and institutions and those whose research examines how history and political economy influence cultural practices. The treatment of “preferences” in rational choice theory illustrates why so many anthropologists are uncomfortable with economic approaches. For economists, “preferences” refer to individual (or group) rankings of the desirability of different outcomes. For example, an individual might prefer apples to oranges or spending money on a swimming pool rather than making a charitable donation. Such preferences are ordinarily assumed to be “given,” meaning that the analyst is unconcerned about why individuals have them. Becker, as usual, takes an extreme position: Since economists have little to contribute . . . to the understanding of how preferences are formed, preferences are assumed not to change substantially over time, nor to be very different between wealthy or poor persons, or even between persons in different societies and cultures. (1976b:5)

Almost any anthropologist would be appalled by Becker’s casual assumption of cultural uniformity over space and time. I would not argue that all economists share Becker’s explicit disinterest in cultural variability and change. But it is clear that most have little professional interest in cultural differences. The description and explanation of such differences, in contrast, are at the heart of most ethnographic work. Approaches consistent with rational choice theory nonetheless have been influential within varied subfields of anthropology. Anthropologists adopting ideas from evolutionary psychology (such as Bliege Bird and Smith 2005) have argued that diverse aspects of cultural behavior are the result of humans attempting—not necessarily consciously—to maximize their genetic contributions to future generations. Ecological anthropologists studying hunter-gatherers (Winterhalder 1987 and others) use the economistic optimal foraging theory in their efforts to explain food-

4 Anthropology, Economics, and Choice

getting patterns. Demographic anthropologists (such as Nag, White, and Peet 1978) have attributed increases and decreases in fertility patterns to changes in the economic value of child labor. Rational choice theory might be expected to be especially appealing to anthropologists whose research focuses on the economy. Economic anthropologists, however, have had an uneasy relationship with mainstream economists. This discomfort can be traced in part to the areas where anthropologists have historically done fieldwork. In the first part of the twentieth century, most ethnographic research took place in settings quite different from those usually studied by economists. In many anthropologists’ research sites, goods and services were exchanged primarily via kin networks. Anthropologists working with members of tribal societies in the deserts of Africa, the islands of the Pacific, and the forests of Amazonia were understandably skeptical about the general applicability of theories designed to explain decision making in the economic institutions of states. During the 1950s and 1960s theoretical discussions in economic anthropology were therefore dominated by what came to be called the formalist-substantivist debate. These impassioned arguments focused on the relevance of economic theories and models in societies where markets were either absent or of limited importance. The formalists (Burling 1962, Cook 1966, LeClair 1962, to name a few) argued that economic theory was directly applicable to such places. Their empirical work stressed methodological individualism, made extensive use of quantitative methods, and treated culture primarily as an unanalyzed shaper of preferences. Inspired by the writings of Karl Polanyi (1944, 1957), the substantivists (including Dalton 1961, Sahlins 1972) dismissed the relevance of western economic theory, emphasized institutions rather than individuals, and provided ethnographic descriptions of exchange systems such as reciprocity, redistribution, and markets. By the 1980s most economic anthropologists regarded the formalistsubstantivist debate as a dead end.2 Furthermore, the vast majority of anthropologists were by this time conducting research in societies where markets are important. Nonetheless, there remained a clear divide within economic anthropology between those advocating the use of formal models based on utilitarian assumptions and those emphasizing the importance of history and culture. Model-oriented economic anthropologists examined topics such as demographic influences on household labor allocation (Durrenberger 1984), the effects of risk aversion on decision making (Cancian 1979), and competition over the use of commonly held

Introduction 5

resources (McCay and Acheson 1987). Such research differed greatly from the many historical studies in the 1970s and 1980s influenced by dependency and world systems theories (for example, Mintz 1985, Roseberry 1983) that examined international trade, inequalities between the rich and poor nations, and the socioeconomic relations associated with commodity chains. Such historically and politically oriented research is today ordinarily framed in the context of globalization. Studies of consumption also have become prominent in economic anthropology (among them Howes 1996, Miller 1998, Orlove 1997, Rutz and Orlove 1989); these often incorporate ideas from cultural studies and postmodernism. Despite this divide, even model-oriented contemporary economic anthropologists do not ordinarily uncritically accept mainstream economic theory. They often argue that such theories need to be modified to be applicable to many situations in both western and nonwestern settings. The great majority of economic anthropologists—whatever their theoretical position—distance themselves from conventional economic theory.

The Value of Ethnography My research in Belize, Peru, Mexico, and the United States over the past several decades has focused on the work lives and economic strategies of individuals and households. In my efforts to figure out how and why people in these diverse places make decisions about their livelihoods I have read many models of human behavior consistent with rational choice theory that have been proposed by economists, anthropologists, cognitive psychologists, and evolutionary biologists. These models employ to greater or lesser degrees the economic approach advocated by Gary Becker. Although these models have sometimes helped my understanding of practical issues, I found that they were more often of little use because of their intentional lack of attention to relevant ethnography, history, and political economies. In their efforts to explain everything, they explained very little. The case studies in this book are intended to show the practical relevance of what can seem to be arcane intellectual debates. They show that the issues examined are not abstruse theoretical conundrums of interest only to scholars; instead, they are essential to our understanding of how and why people around the world make important decisions about their livelihoods and the welfare of their families. The details in the ethnographic examples illustrate what rational choice models leave out.

6 Anthropology, Economics, and Choice

Making Choices Economic approaches to decision making have been criticized on both practical and theoretical grounds. Many scholars argue that the expected utility framework underlying such approaches is difficult to apply to realworld situations. Some go further and question the reasoning behind utilitarian explanations of cultural practices. Expected Utility and Rational Choice Most contemporary versions of rational choice theories are based on something called “expected utility.” This theoretical approach assumes that decision makers are able to make approximations of the expected utilities (payoffs) associated with alternative choices. The idea is that decision makers select the option that provides the greatest expected utility. Writers about expected utility (and more generally rational choice) differ as to whether they consider their theories to be prescriptive or predictive. Those taking a prescriptive approach say that while their methods tell us what decision makers should do, no claims are made that people actually make optimal choices. Those taking a predictive approach argue that much human behavior can be explained by looking at the extent to which various choices maximize decision makers’ expected utility (self-interest). There are several obvious problems associated with attempts to apply an expected utility framework—whether prescriptive or predictive—to on-the-ground decision making. The most important of these difficulties are the following: 1. Expected utility theory assumes that decision makers can assign values (utilities) to alternative outcomes of decisions. “Utility,” however, is a murky concept that is almost impossible to define and measure. Although economists often examine monetary returns to alternative choices, people’s decision making is often influenced by other goals such happiness, leisure, and risk avoidance. Some goals such as “happiness” can be difficult to measure. Even when outcomes can be quantified, there may be no obvious way to combine the different kinds of payoffs from a particular outcome into a single measure of “utility.” How can, for example, the happiness one gets from winning an athletic competition be compared to the monetary rewards from such a championship? 3 2. Expected utility models of decision making depend in part on deci-

Introduction 7

sion makers’ views about the probabilities of alternative outcomes when particular choices are made. In many decision-making situations (such as the chances that a new restaurant will be successful), these probabilities are difficult or impossible to estimate. Even when such probabilities can be guessed reasonably well from past experiences (weather records) or statistical theory (state lotteries), decision makers’ estimates of the chances of possible outcomes may not be realistic. 3. Estimates of expected outcomes (payoffs) from particular choices must specify what length of time is being considered. Decision makers, however, may have hazy ideas about time frames. This can cause analytic problems when the short-term and long-term consequences of a decision differ. The potential long-term consequences of a decision often differ from those in the short term. Smokers, for example, must weigh the immediate pleasures of a cigarette against their increased chances of getting lung cancer some day. Their decisions about whether to smoke depends on whether the time frame being considered is the next five minutes or the next twenty years. The “rational” decision might well be to smoke if only the next few minutes are considered and to abstain if the next two decades seem relevant. 4. Expected utility models assume that there are no problems in specifying which individual or group is making a particular decision. Although such models usually assume that choices are made by an individual, they sometimes examine decisions made by groups such as families, corporations, labor unions, and political parties. In many situations, however, decision-making units are hard to isolate. Decisions that may seem to be made by individuals are often influenced by the ideas and concerns of many people.

Broader Meanings of Rationality Many books and articles about “rationality” discuss utilitarian explanations of cultural practices that assume that people are consciously attempting to achieve measurable goals. A brief look at the polemical writings of an advocate and a critic of such explanations shows well how debates over the usefulness of the concept of “rationality” reflect fundamental disagreements about the causes of human behavior. The advocate is Tim Harford, an economist whose books (2005, 2008) aim at making the often-abstruse ideas of his discipline comprehensible for general

8 Anthropology, Economics, and Choice

readers. The critic is the famous anthropologist Marshall Sahlins, a longtime outspoken opponent of economic and ecological explanations of cultural patterns. In his book The Logic of Life: The Rational Economics of an Irrational World (2008:9), Harford describes his perspective on decision making: Rational people respond to incentives: When it becomes more costly to do something, they will tend to do it less; when it becomes easier, cheaper, or more beneficial, they will tend to do it more. In weighing their choices, they will bear in mind the overall constraints upon them; not just the costs and benefits of a particular choice, but their total budget. And they will also consider the future consequences of present choices.

Harford’s comments are deliberately consistent with the expected utility approach to decision making. Notice, however, the imprecision of words such as “cheap,” “easy,” “future,” and “beneficial.” This is in part because Harford is trying to make complicated ideas accessible for readers who may understandably lack the patience to decipher technical definitions of utility and the ability to calculate probability-based expected outcomes. But Harford’s use of ordinary language also makes it easy for him to loosely interpret diverse types of behavior as being “rational.” He is not attempting to rigorously test hypotheses. Some years ago Sahlins wrote a book in which he harshly criticized numerous noted anthropologists for their utilitarian explanations of cultural phenomena. In Culture and Practical Reason (1976), Sahlins argues—in language less accessible than Harford’s straightforward prose—that such explanations constitute “an epistemology for elimination of culture itself as the proper anthropological object” (83). He goes on to say: Without distinctive properties in its own right, culture has no title to analysis as a thing-in-itself. Its study degenerates into one or another commonplace naturalisms: the economism of the rationalizing individual (human nature), or the ecologism of selective advantage (external nature). (Ibid.)

Sahlins here is not explicitly criticizing expected utility approaches to rational choice. He is instead generally arguing against what he regards as reductionist explanations of cultural practices that emphasize their usefulness to either individuals or groups. Sahlins is hardly alone in making

Introduction 9

such critiques (Douglas and Wildavsky 1982, Lupton 1999, and others do as well); the influential postmodern movement in anthropology during the latter part of the twentieth century was based in part on a related rejection of attempts to make anthropology a social science seeking lawful regularities in human behavior. Although much of this book consists of my criticisms of what I see as the limitations of economic models, I want to be clear than I am not advocating culturalist analyses. My discussions of the shortcomings of rational choice approaches to particular issues in economic anthropology are not intended to be read as rejections of utilitarian explanations of human behavior. I actually think that anthropologists should whenever possible seek such explanations. I agree more with Harford than with Sahlins.

Migration from Mexico to the United States The problems associated with attempts to apply economic models to realworld situations can be seen by examining closely the decisions that two brothers made about migrating temporarily from Mexico to the United States. Rational choice theory provides an incomplete and in some ways inadequate framework for understanding why the brothers made different choices. In 2003 Jorge and Oscar Morales began to think seriously about leaving their community near the city of Oaxaca in southern Mexico and migrating temporarily to California.4 The two brothers, both married and in their late twenties, were the youngest of seven children in a hard-working artisan family. Their village of a thousand people had prospered in the 1990s by selling brightly painted wood carvings to tourists and dealers in folk art. Jorge and his wife, Sandra, had three young children; Oscar’s wife, Veronica, had just given birth to their second daughter. The two brothers and their families lived in a large compound with Alberto and Alicia, the parents of Jorge and Oscar. Alberto and Alicia sold masks and jewelry that they made at Monte Albán, a nearby archaeological site. Oscar, three years older than Jorge, was one of the most highly educated artisans in Oaxaca. He excelled in his studies, obtaining a degree in business administration from a local university. For several years Oscar earned a good living by local standards by working as an intermediary for Roger, a large-scale dealer in folk art from Arizona. Roger sent orders to Oscar for particular types of carvings. Oscar arranged with local artisans to make the pieces, which he later collected, packed, and sent to Arizona.

10 Anthropology, Economics, and Choice

Oscar was a decent carver and painter, but his pieces are unremarkable in a place where there are many talented artisans. He stood out instead for his business skills and reliability. Oscar was one of the most respected men in his community. By 2003 he had already served as the (unpaid) treasurer and secretary in the local government. Although Jorge was a better-than-average student, he enjoyed his studies less than Oscar did and left school after attending prepa (the equivalent of a U.S. high school) for a few years. Jorge had little interest in marketing crafts and did not join Oscar and another brother in their work for Roger. He was, however, a talented artisan who specialized in making meticulous small carvings of animals including goats, cats, and giraffes. Sandra, a skilled painter, helped with the pieces but was limited in what she could do because of child care responsibilities. The demand for Oaxaca wood carvings began to decline around 2000. Although sales of well-made, relatively expensive pieces remained good, fewer tourists, shop owners, and wholesalers were buying the cheaper carvings that comprised the bulk of the trade. In 2002 Roger decided to stop selling Oaxacan wood carvings and to concentrate instead on other Latin American crafts. Oscar lost his job with Roger and needed immediately to find other ways to support his family. Jorge’s income also decreased because of fewer sales of his pieces. Like many other Mexicans their age, Jorge and Oscar knew that temporary migration to the United States could provide income that could support their families. Several of their older brothers regularly went back and forth to California, where they seemed to have little trouble finding work. Although these brothers migrated without legal documents and worried about the difficult border crossing, they had been able to make their trips without serious difficulties. If Jorge and Oscar went to California, they could stay with family members or friends already there and use these connections to find steady if low-paid employment. The advantages and disadvantages of such migration are well known in both the United States and Mexico. The wages for unskilled work in the United States are much higher than those for even most professional jobs in Mexico. Because of this discrepancy in wages, remittances (money sent home by migrants) are one of the major sources of income in Mexico. Many migrants hope that they might be able to settle permanently in the United States and make better lives for themselves and their families. Some migrants—whether or not they see their border crossing as permanent—also regard their journeys as exciting opportunities to see another part of the world.

Introduction 11

Most migrants, however, are reluctant to leave Mexico. Crossing the border without papers can be dangerous; even those who safely enter the United States live in constant fear of being caught by immigration officials and being jailed or deported. Perhaps the most common reason that people give for disliking migration is the loneliness of being separated from loved ones. Even in an age when telephone and Internet communication is easy and inexpensive, being away from home can be painful (Cohen 2004:7). Oscar and Jorge reacted differently to the possibility of migrating. For Oscar this was an alternative only to be pursued as a last resort. He was devoted to his family and comfortable in his position in his community. Oscar did what was expected by authorities; living illegally in the United States would be especially uncomfortable for him. Furthermore, Oscar was friendly with many folk art dealers, academics, and tourist agents in both the United States and Mexico. These contacts, he thought, would help him find new sources of income. Although Jorge was not enthusiastic about leaving Oaxaca, he had fewer reservations about migration than Oscar. Lacking Oscar’s education and social networks, Jorge had fewer options if he remained in Mexico. Perhaps also he was not as content as Oscar with his life in Oaxaca. Jorge had never earned as much money as Oscar. While popular, Jorge did not have the deep respect that Oscar had earned through his community service. The decisions Oscar and Jorge made about migration were not entirely independent. Both brothers thought it best if one remained in their community to help with their elderly parents, their children, and the maintenance of the family compound. While such considerations would not prevent both brothers from migrating if economically necessary, they clearly influenced Oscar’s thinking. However, if one brother decided to migrate and send remittances back home, this would help everyone living in the compound. In 2003 Jorge left for Los Angeles. Although he phoned his family most days, as of 2010 Jorge had not returned to visit even once because of fears about increased vigilance at the border. Jorge, who now speaks good English, works days at an insurance company and nights in a grocery store. He shares living space with an ever-changing cast of relatives. Jorge has done well economically compared to most Mexican migrants; his brothers in Los Angeles, for example, work fewer hours at less well-paid and more physically taxing jobs. Through his connections with university professors (including me), Oscar was able to legally visit the United States several times to give

12 Anthropology, Economics, and Choice

wood-carving demonstrations and sell pieces. After making several such trips, Oscar obtained a ten-year tourist visa that allows him to enter the United States without problems. His trips to the United States, which ordinarily include visits with Jorge and other relatives, last only a few weeks. Oscar continues to serve in important political positions in his community. In recent years Oscar has not done as well economically as he had hoped after losing his job with Roger. He has supported his family by making and selling wood carvings, teaching computer skills to local children, and working occasionally for visiting scholars. Wood-carving sales in Oscar’s community fell sharply between 2006 and 2008 because political problems in Oaxaca resulted in fewer tourists visiting the state. In reaction to these economic problems, Oscar founded an artisans organization that promotes the cooperative marketing of wood carvings. Although the organization has had some success and tourism in Oaxaca is rebounding, Oscar and many of his neighbors can no longer rely on wood carving as a good source of income. There is no question that Oscar could have earned more money over the past several years if he had chosen to migrate. But he is content with his life in Mexico and almost certainly would be less happy in Los Angeles away from his family.

When Is Migration “Rational”? The applicability of rational choice models has been questioned (Gladwell 2005, for example) for the many situations in which people make decisions without consciously and systematically weighing the costs and benefits of alternative actions. The choices Jorge, Oscar, and other Mexicans face about migration, however, involve a careful consideration of the pros and cons of leaving home. Migration is a perennial topic of conversation in Mexican communities; families may discuss for years whether particular members should migrate. Jorge and Oscar’s decision-making processes nevertheless illustrate the limitations of rational choice approaches. Although these approaches can help us understand the reasons for variations in the choices made by members of different socioeconomic groups, their intentional downplaying of ethnographic detail leads to thin, unsatisfying analyses. The basic idea underlying rational choice models is that individuals will pursue actions that are in their self-interest. In the case of Mexican

Introduction 13

migration, this suggests that individuals who would benefit most from migration are more likely to go to the United States than those who would benefit less. While this appears to be a truism of no explanatory value, rational choice approaches have led researchers to specify the socioeconomic conditions in Mexico and the United States that lead certain groups to be more likely to migrate than others. Much research on this topic (such as Massey, Goldring, and Duran 1994) has therefore consisted of useful examinations of statistical correlations between migration rates and membership in particular demographic and economic groups. Such studies have helped us understand, for example, why most migrants from Oaxaca to the United States are young men. An examination of the migration decisions made by Jorge and Oscar shows what is left out of such analyses. Using an economic model to determine which choices were in their self-interest is difficult to do because of the slipperiness of the concept of “utility,” the unpredictable outcomes of different decisions, the differences between short-term and long-term consequences of actions, and the fuzzy nature of decision-making units. The Uselessness of “Utility” There are two main reasons Jorge and Oscar made different decisions about migration. First, Oscar had job skills and social connections that might be more useful in Oaxaca. Second, the brothers placed different emphases on the relative importance of the economic and personal consequences of migration for themselves and their families. A researcher focusing on rational choice might attempt to measure the differences in local job opportunities, but such an approach would be of no help in understanding the different weights that Jorge and Oscar placed on economic and psychological considerations. An advocate of rational choice could only make the meaningless observation that Oscar placed higher “utility” than Jorge on the psychological well-being of himself and his family. Although from this perspective both brothers acted “rationally,” the concept of “utility” does not aid our understanding of how their psychological make-up influenced their choices. Unpredictable Consequences When Jorge and Oscar discussed going to the United States, they had some idea of what their lives would be like in California. Still, they did not know what kinds of work they would find, whether they would have

14 Anthropology, Economics, and Choice

trouble with the border patrol, and how happy or unhappy they would be away from their community. They could make only vague guesses about what their economic and personal situations would be like over the next several years if they stayed in Mexico. Their migration decisions therefore only partly fit expected utility theory assumptions about the ability of decision makers to specify the probability of alternative outcomes to their choices. Short Term versus Long Term When Jorge and Oscar contemplated migration, they were unsure about a time period on which to base their decision. Migration might be shortterm if large amounts of money were earned and economic conditions in Oaxaca improved. If the economic situation in Oaxaca worsened (as turned out to be the case), migrants might spend a long time—perhaps even the rest of their lives—in the United States. Who Makes Choices? Studies of migration ordinarily implicitly or explicitly assume that decisions are made autonomously by individuals like Jorge and Oscar. This is not always so. Jorge and Oscar, like many other potential migrants, often talked with various members of their families about the possibility of leaving. Jorge might not have left without knowing that Oscar would probably stay; Oscar might not have stayed without knowing that Jorge was leaving. The views of the wives, parents, and siblings of Jorge and Oscar also influenced their decisions. Ethnography and Rational Choice Models Economists, political scientists, and sociologists taking rational choice approaches to migration use statistical analyses to compare the relative importance of a small number of key variables thought to influence decisions to stay or leave. While such analyses have provided valuable information about migration patterns, they can be of limited help in understanding particular choices made by individuals. I question the extent to which Jorge’s and Oscar’s decisions can be analyzed without examining the details of their lives, local family structures, and economic and political developments in the state of Oaxaca.

Introduction 15

Economic Theory, Rational Choice, and the Financial Crisis The recent worldwide financial crisis has led economists to pay more attention to critiques of strict versions of rational choice theory (Cassidy 2009 and Fox 2009 provide good popular overviews). Consumers, investors, bank officials, and policy makers made decisions during the past decade that in retrospect were “irrational” in their underestimation of the possibility of disastrous outcomes. Their choices seemed inconsistent with ideas such as the rational expectations hypothesis, part of an influential theoretical framework proposed by the economists Robert Lucas (1987) and John Muth (1961). John Cassidy nicely summarizes this framework: Lucas . . . [assumes] that everyone knows exactly how the economy works. People aren’t merely aware that unemployment is somehow linked to inflation, which is linked to interest rates; they all have the same (correct) mathematical model of the economy in their heads, which they use to form expectations of wages, prices, and other variables. . . . By invoking the rational expectation hypothesis, Lucas could simply write down some equations to describe how workers, firms, and the government behave, put a mathematical expectation operation in front of them, and derive a solution that was consistent with the decision rules of everybody in the economy. (2009:99–100)

It would be foolish to argue—as anthropologists sometimes do—that until recently almost all economists shared the worldviews of Lucas and Muth. Prominent economists like John Maynard Keynes (1937), Francis Bator (1958), and Richard Thaler (1988) have long questioned the assumption that economic decision makers with complete knowledge of the possible outcomes of alternative actions consciously calculate utilities when making choices. In subsequent chapters I discuss in detail some recent and not-so-recent critiques of this assumption. In order to give readers a feeling for the flavor of these critiques, I briefly present here three ideas offered by economists who question strict versions of rational choice theory. The “herd mentality” (Scharfstein and Stein 1990) refers to the tendency of economic decision makers such as investors to imitate the often risky choices of others even when their own inclinations and calculations suggest that they should do otherwise. Imitation occurs not only because individuals going along with the crowd doubt their own judgment or fear

16 Anthropology, Economics, and Choice

being mocked for nonconformist behavior. In certain respects it makes economic sense to conform even when doing so appears to be “irrational.” When individuals go along with the crowd and things turn out badly, the blame is shared by everybody. If individuals differ from the crowd, they bear sole responsibility when things go wrong. Conservative investors working for banks or Wall Street firms during the heights of the dot-com boom of the late 1990s and the housing bubble of the past decade would have lost substantial bonuses and perhaps their jobs if they declined to follow their peers in making risky allocations of money into Silicon Valley startups and subprime derivatives. Behavioral economics consists of attempts to apply findings from laboratory experiments to economic decision making in the outside world. Most behavioral economists (including Thaler 1988) have been strongly influenced by research carried out in the 1970s and 1980s by the psychologists Daniel Kahneman and Amos Tversky. Kahneman and Tversky demonstrated in a series of ingenious experiments (1982a) that most people in laboratory settings in western societies are unable to make the calculations required by strict versions of rational choice theory. Instead, they rely on rules of thumb and unsubstantiated beliefs that enable them to make complex decisions quickly. Researchers in behavioral finance have used the ideas of Kahneman and Tversky in their explanations of trend following, speculative bubbles, and poor corporate decision making (Cassidy 2009:198). Scholars focusing on asymmetric information directly challenge the assumption of rational choice theory that all participants in economic transactions are well informed about the possible consequences of alternative actions. They point out that in many transactions, not all participants have equal access to relevant information. In a classic paper, George Akerlof (1970) gave the example of the market for used cars. Sellers are ordinarily more familiar than buyers with the condition of cars for sale. A buyer, knowing this, often distrusts the claims of a seller about a car’s condition and fears being saddled with a lemon. One result is to depress the price of cars that are actually in good shape. Akerlof makes the less obvious argument that another consequence would be that some sellers of good cars would take their vehicles off the market, leading to an increased proportion of lemons for sale. In the past decade, economists have used ideas about asymmetric information in their analyses of problems in the financial sector and the health care industry (Cassidy 2009:156–163). These economists criticizing rational choice theory have cogently described the empirical difficulties associated with assumptions about om-

Introduction 17

niscient, calculating decision makers. Their often sensible policy recommendations (as in Thaler and Sunstein 2008) usually involve urging institutions to adopt incentives that will lead fallible decision makers to make choices that are more likely to help both themselves and society as a whole. Although I find the work of these critics provocative and insightful, their analytic approaches nonetheless differ from those of most economic anthropologists. Like other economists, these critics rely heavily on mathematical models using only a few variables, assume that decisions are made by either individuals or groups acting as a unit, deemphasize history and ethnography, and separate “economy” and “society” in their analyses. Their examples are almost entirely confined to the decisionmaking situations and institutions that conventional economic theory was designed to explain. They rarely attempt to apply their ideas—some of which seem relevant—to the kinds of choices that ethnographers typically examine in their fieldwork in western and nonwestern settings alike. Some economists are truly heterodox in the attention they pay to history and ethnography and their deemphasis of individual decision making. Many such economists are influenced by feminist or Marxist theory. These radical critics of conventional economics are a small and uninfluential minority in academic and policy circles in the United States and many other countries. The generalizations I sometimes make in this book about economists are not meant to apply to these scholars.5

Issues in the Analysis of Choice The fundamental assumption of mainstream economics is that careful analyses of decision making help us understand human behavior in diverse situations. Because this assumption is not self-evident, the first chapter looks carefully at critiques of studies focusing on choice. Although these critiques highlight certain weaknesses of economic theory, I conclude that decision making is an important aspect of cultural adaptation and change. The five chapters that follow examine knotty problems in the analysis of decision making. The first of these, chapter 2, considers attempts to calculate monetary values for unpaid labor and production for home consumption. Such calculations are aimed at improving our knowledge of how people make choices between such work and paid labor. Economic analyses of these decisions entail comparisons of the “utility” of paid and unpaid work. Although there are formidable theoretical and practical

18 Anthropology, Economics, and Choice

problems associated with such comparisons and calculations, I conclude that they are necessary for an understanding of many important choices. Chapter 3 compares approaches taken by economists, cognitive psychologists, and anthropologists in their analyses of decision making in situations of risk and uncertainty. These scholars differ greatly in the extent to which they accept the expected utility assumption that decision makers estimate the probabilities of different outcomes to alternative choices. I argue that this assumption—a fundamental part of strict versions of rational choice theory—is almost always unrealistic. Cognitive psychologists agree, showing in laboratory experiments that participants are often unfamiliar with basic concepts of probability. While I do not dispute these findings, my critique is different from that of the psychologists. I question the extent to which the results of their experiments can be extrapolated to real-world decision making. In my view, there is no substitute for ethnographically rich descriptions of the complexities of decision making in particular risky and uncertain situations. Chapter 4 takes a critical look at economic experiments in anthropology. In one of the stranger episodes in the history of economic anthropology, in the late 1990s and early 2000s widespread publicity and significant funding were given to experienced ethnographers running carefully designed experiments in their field sites in nonwestern settings. The purpose of these experiments was to test the cross-cultural validity of ideas from evolutionary biology and economics about decisions related to cooperation. The experimenters made the ambitious claim that their results provided important insights about the “nature of human nature.” Again, the relevance of experiments in quasi-laboratories to real-life decision making can be questioned. This chapter, more than any of the others, illustrates the dangers of making simplistic assumptions about the motivations for human behavior. The next two chapters examine situations in which it is difficult to determine whether individuals or groups are making decisions. Chapter 5 examines definitional problems associated with the concept of “households.” Economists and anthropologists sometimes treat households as decision-making units. National censuses often provide information at the household level, implicitly assuming that this is a more or less obvious socioeconomic unit. As ethnographers have observed, in many places it is impossible to neatly delineate a group of people as a “household” that shares resources and makes choices together. I nonetheless argue that the advantages of collecting and analyzing data at the household level outweigh the disadvantages brought about by this particular simplification.

Introduction 19

Chapter 6 discusses influential theories about what has been called “the tragedy of the commons.” According to these theories, communal resources are frequently destroyed by individuals pursuing their selfinterests. Ethnographers have shown that such theories do not always accurately depict decision-making situations. In many situations there are cultural rules and social institutions that prevent the untrammeled exploitation of publicly held resources. The conclusion summarizes the major differences in the ways anthropologists and economists study choice. I conclude that the mathematical models of economists must be complemented by more descriptive, empirical approaches that consider the context within which choices are made.

CHAPTER 1

How Important Is Decision Making?

[T]he difference between economics and sociology is very simple. Economics is all about how people make choices. Sociology is why they don’t have any choices to make. DueSenberry 1960:233

A theory of how people make their economic choices is without interest and probably impossible until we have tackled the prior questions of the factors determining what choices are available to them. White 1976:36

When I went to Belize in the fall of 1971 to conduct research for my doctoral dissertation, my overall goal was simple enough. I wanted to learn how rural residents of three communities made decisions about how much time to spend over the course of a year on alternative ways of producing food and earning income. This turned out to be a difficult task, involving careful analyses of the costs, benefits, and risks associated with different ways of earning a living. For the most part, I avoided thinking too much about why these costs, benefits, and risks were what they were. Explanations, for example, for the poor road system and low price of corn seemed beyond the scope of my research. I was having enough trouble collecting basic information on the yield of different crops, the composition of households, and the techniques involved in catching fish. While I was writing up my thesis, I sometimes questioned the value of what I was doing. Many of the so-called choices I was analyzing were greatly affected by historical, political, and macroeconomic forces that rural Belizeans could not control. Perhaps my research—while not useless—was relatively unimportant. Maybe I should have spent more time learning about the history and political economy of Belize.

How Important Is Decision Making? 21

I was not alone in worrying about such issues. The question of the explanatory power of studies of decision making is related to two classic debates in social science. The first concerns the relative importance of fine-grained local studies and wider-ranging examinations of regional, national, and international economics and politics. The second is about the usefulness of methodological individualism as opposed to the analysis of cultural norms and supra-individual social institutions. I cannot pretend here to be able to resolve these debates. My goal is the more modest one of contrasting the ways in which economists and anthropologists deal with the limitations of decision-making studies. Economists usually regard most constraints on decision making as “givens” beyond the scope of their analyses. They often, however, analyze how incentives offered by different macroeconomic policies affect the choices that people make. Economic anthropologists take several approaches to decision making. Some attempt careful formalist analyses of how people make choices in particular times and places. Others adopt the substantivist position that economic choices are limited because they are embedded in social institutions. Most emphasize how local decisions must be understood in the context of wider historical, economic, and political circumstances. Although many practicing economic anthropologists combine these approaches, they differ in which ones they emphasize. Before contrasting the approaches of economists and anthropologists to the limitations of decision-making studies, I look at the Belize case in more detail. Pragmatic examinations of particular cases, I think, provide insights about general issues that are obscured in programmatic theoretical proclamations.

How Much Choice Did Rural Belizeans Really Have? Much of my research for my doctoral dissertation consisted of attempts to understand decisions that rural Belizeans in the early 1970s made about how to allocate their work time between cash crops and wage labor. I began by asking people to tell me what they thought were the advantages and disadvantages of these two ways of earning income. Using what they told me as a starting point, my analyses included calculations of monetary returns per unit of time for different types of cash cropping and wage labor, estimates of the risks and uncertainties associated with various ways of earning a living, and creations of models of labor allocation that appeared to maximize cash income over the course of a year. I also described how individuals’ ability to participate in different types

22 Anthropology, Economics, and Choice

of income-producing activities was affected by their particular circumstances. Farmers living near a road, for example, could market cash crops more easily than those living in more remote places. Although I was confident that my descriptions and analyses (Chibnik 1975, 1980) advanced our understanding of how and why these decisions were made, there were times when I wondered about the value of such detailed studies of choice. Did it really matter that I understood why Belizean farmer X planted more rice than corn and rarely worked in wage labor, while Belizean farmer Y planted no rice at all and instead did wage labor, grew corn, and raised citrus crops? Was it not more significant that both farmers X and Y were in a situation that no matter what they did, the returns to labor were so low their families could hope for no more than a modest standard of living? Another way of looking at decision making by rural Belizeans in the early 1970s in the area where I did research would be to emphasize the constraints that limited their choices and affected the input/output ratios associated with different ways of making a living. Such constraints were ecological, technological, economic, and political. They included a poor road system, a low population density, a small internal market for cash crops, a complicated land tenure system, a newly instituted state Marketing Board that guaranteed prices for certain agricultural products, the increasing availability of international aid, the colony’s political relationship with Britain, local soil systems, military threats to Belize from Guatemala, the operations of foreign citrus companies, economic competition from Mennonites, the availability of agricultural advice from government officials, and the beginnings of what later became a flourishing tourist industry. Although I discussed these matters in my publications at some length, they were not the focus of my research. My concerns about the limitations of ethnographic studies were shared by many other sociocultural anthropologists in the 1970s. During this period, scholars in diverse disciplines (for example, Gunder Frank 1969, Wallerstein 1974) were advocating political economic studies that emphasized how economic conditions in “satellites” such as Belize were affected by the activities of governments and firms in “metropoles” such as Britain and the United States. The central idea of this “dependency theory” was simple: rich countries gained their wealth and power through unequal economic relations with poor countries. Furthermore, within poor countries there were parallel exploitive relations between the rich and poor and between urban centers and rural peripheries. Dependency theory has been criticized for its neglect of the particularities of local and regional

How Important Is Decision Making? 23

histories and its unwillingness to grant agency to the activities of those individuals and groups resisting the incursions of capitalist enterprises. Nonetheless, there can be no question that the general ideas of dependency theory help explain historical developments and economic conditions in many parts of the world. One long-ago day I was discussing these ideas with another graduate student who was conducting research among farmers in rural France, a place not usually thought of as prime territory for theories decrying the effects of imperialism. My colleague announced that she was planning to do research on political maneuvering on agricultural issues in the French parliament. The decisions made there were the ones that really counted in the lives of the French farmers with whom she worked. I was taken aback. Why not also study the history of the French parliament, the invention of the machines used in farming, the origins of the land tenure system, and countless other topics that in some way affected the lives of the people in the rural community where she was conducting research? Surely, I thought, my colleague had to decide that certain matters were outside the scope of her project. Despite my skepticism, my colleague had raised an issue of fundamental importance, one that has continued to preoccupy anthropologists to this day. Ethnographic fieldwork focusing on a single community has become less common than multisited research. Almost all studies that do focus on particular places now include detailed discussions of regional histories, national politics, and economic interactions between local communities and the outside world. The language in which these discussions have been framed, however, has changed over time as dependency theory has become less fashionable, giving way first to studies of resistance and later to analyses of globalization emphasizing multidirectional flows of goods and information. As I wrote my dissertation I also considered the implications of the ongoing formalist-substantivist debate for my study of decision making. Without consciously taking sides in this contentious discussion, I had been adopting a formalist approach emphasizing how “rational” decision makers went about making choices. The substantivists argued that such an approach was fundamentally mistaken because it ignored how economic actions were often embedded in local social organization and culture. Although I was not entirely sure what “embedded” meant, the general idea was clear enough. The substantivists thought that economic choices in many societies were constrained by social customs and institutions such as kinship obligations and religious taboos. While the dependency theorists

24 Anthropology, Economics, and Choice

emphasized how external institutions such as states and firms limited decision making, the substantivists focused on internal sociocultural factors that prevented people from having all that much choice. The substantivist argument seemed of limited relevance to my research in Belize. I was not working in a place with a long-standing cultural tradition and complex forms of indigenous social organization. The groups I worked with—including the Mayas—were all descendants of people who had either been brought to Belize as slaves or migrated there while the colony was under British control. The only way that the substantivist position seemed pertinent was in local patterns of marriage and residence. Arguably as a result of economic conditions under colonialism, men in two ethnic groups in the area—Creoles and Caribs (Garifuna)—often remained in their parents’ homes well into adulthood while fathering and supporting children with women living elsewhere, usually with their own relatives. Maya men and women, in contrast, married young and established their own households. These cultural differences clearly influenced economic activities. But such differences affected the local economy much less than, for example, low prices for cash crops, poor roads, and a small internal market.

What Economists Intentionally Ignore Three academics went off on a sailboat together, a physicist, a chemist, and an economist. Unfortunately, they ran into a storm and the boat was wrecked on an uninhabited island. The only food they were able to rescue from the wreckage was a case of baked beans. As they got hungry, they began to wrestle with the problem of how to open the bean cans. The physicist said, “I’ll climb a tree and throw a can onto a rock and it’ll split open.” The others didn’t much like this idea because they thought the beans would just splatter everywhere. The chemist said, “We can soak the cans in salt water and they’ll rust through.” The others didn’t much like this idea because it would take too long. Then the economist said, “Hey—no problem, we’ll just assume a can opener.” claSSic joke about economiStS

In 1995 Amartya Sen was invited in to give the Frisch Memorial Lecture at the World Econometric Conference in Tokyo. Sen, who was awarded the Nobel Prize in Economics a few years later, decided to talk about “maximization and the act of choice.” As might be expected in a lecture to this audience, much of the talk was phrased in the language of mathe-

How Important Is Decision Making? 25

matical models. The beginning of the lecture, however, included an accessible discussion of two examples demonstrating the limitations of rational choice analyses of decision making. I present these examples in some detail because they clearly illustrate the differences between economic approaches to choice and those in disciplines such as anthropology and sociology. The first of Sen’s examples is a trivial matter—the selection of a chair at a garden party. Sen presents a situation in which someone is invited to a party where one chair is obviously more comfortable than the others. The guest nonetheless sits in a different chair. How, Sen asks, can such a choice be reconciled with the rational behavior assumed in conventional economic theory? Using the jargon of his discipline, Sen says that this might happen because the guest’s “preference for choice behavior may well be defined over ‘comprehensive outcomes,’ including choice processes (in particular, who does the choosing) as well as the outcomes at culmination (the distribution of chairs)” (1997:747). Fortunately for noneconomists, Sen goes on to explain what he has in mind in somewhat more comprehensible language. He suggests four reasons the selection of a lesspreferred chair might make sense: (i) Reputation and indirect effects: The person may expect to profit in the future from having the reputation of being a generally considerate person, and not a vigilant “chair-grabber.” (ii) Social commitment and moral imperatives: She may not think it morally “right” to grab the most comfortable chair, cutting others out, and such “moral sentiments” could be explicitly followed or only implicitly obeyed. (iii) Direct welfare effects: The person’s well-being may be affected directly by the process of choice (for example, by what people think of her—she may not enjoy the look she gets as she makes a dash for the great chair) . . . (iv) Conventional rule following: She may be simply following an established rule of “proper behavior” (as the ongoing norm), rather than being influenced by direct welfare effects, or even by any selfconscious ethics. (747–748)

According to Sen, the first of these explanations is most in harmony with the conventions of standard neoclassical economics. The other three explanations are based on “forces” outside of conventional economic explanations.

26 Anthropology, Economics, and Choice

Sen’s other example, of more substantive weight, is the much-debated paradox among economists about why people in democracies take the time to vote (Boudon 1998, Overbye 1995). When the number of voters is large, the ballot of any one individual has little chance of affecting the outcome. From a narrow cost-benefit perspective, it appears that voters would do better by staying away from the polls. Sen points out that the act of voting may be important to a person because of the cultural significance of political participation. Perhaps because such significance is difficult to measure, it is rarely incorporated into economic models of voting. From an anthropological perspective, Sen’s need to make these elementary points in such a rarefied setting suggests that something basic is missing from most economic analyses. Stripped of the jargon and the mathematics, Sen’s main point seems to be that culture matters in decision making. He is not the first to have noticed this. In any case, Sen’s remarks only begin to scratch the surface of the complexity of his examples. His comments ignore the historical and sociopolitical context of the choices he describes. In order for decision makers to choose chairs and pick candidates, they must be invited to garden parties and be permitted to vote. On a broader lever, such choices would be nonexistent without the historical development of the institutions of garden parties and political elections. Economists, of course, know that much is missing from their analyses. But many economists do not regard these omissions as a defect of their methods. As will be seen, they argue instead that the creation of simplified models actually improves our understanding of choice. Microeconomics and Macroeconomics The field of economics is conventionally divided into two subfields— microeconomics and macroeconomics. Most of the discussions in this book about the ideas and assumptions of economists focus on microeconomics because that subfield is more directly concerned with decisionmaking theories. Nonetheless, the typical assumptions of macroeconomists about the existence and cross- cultural relevance of certain institutions of western market societies form the context within which microeconomists make their analyses of choice. Microeconomics is usually defined as the study of individual entities such as firms or households. Most microeconomic analyses explicitly or implicitly assume market-based societies where firms are producers and

How Important Is Decision Making? 27

households are consumers. Gregory Mankiw, for example, concisely defines microeconomics as “the study of how households and firms make decisions and how they interact in specific markets” (2005:27). Macroeconomics examines either a national economy as a whole or large subdivisions such as government, household, or business sectors. The examples given in textbooks show how macroeconomic analyses ordinarily assume the institutions of contemporary market societies: Macroeconomics speaks of such economic measures as total output, total income, aggregate expenditures, and the general level of prices. (McConnell and Brue 2005:9) Today, macroeconomists examine a wide variety of areas, such as how total investment and consumption are determined, how central banks manage money and interest rates, what causes international financial crises, and why some nations grow rapidly while others stagnate. (Samuelson and Nordhaus 2005:5) A macroeconomist might study the effects of borrowing by the federal government, the changes over time in the economy’s rate of unemployment, or alternative policies to raise growth in national living standards. (Mankiw 2005:27–28)

Economists recognize that microeconomics and macroeconomics are related. Much of the aggregate data studied by macroeconomists is the result of large numbers of decisions made by individuals, households, and firms. Nonetheless, the two subfields are regarded as distinct and are often taught in separate courses at universities. Milton Friedman and Economists’ Use of Assumptions Long ago the famous economist Milton Friedman wrote a book in which he presented his views about the methodology of both microeconomics and macroeconomics. Although these ideas have been derided as a “simpleton’s version of positivism” (McCloskey 1998:xi), they continue to represent the mainstream approach to theory and modeling in economics. According to Friedman (1953:7) and many philosophers of science—not all simpletons—theories and hypotheses aim at making valid and meaningful predictions about phenomena not yet observed. They are

28 Anthropology, Economics, and Choice

intended to abstract essential features of a complex reality. Such theories and hypotheses are intentionally unrealistic: Truly important and significant hypotheses will be found to have “assumptions” that are wildly inaccurate descriptive representations of reality, and, in general, the more significant the theory, the more unrealistic the assumption. . . . The reason is simple. A hypothesis is important if it “explains” much by little, that is, if it abstracts the common and crucial elements from the mass of complex and detailed circumstances surrounding the phenomena to be explained and permits valid predictions on the basis of them alone. To be important, therefore, a hypothesis must be descriptively false in its assumptions; it takes account of, and accounts for, none of the many other attendant circumstances, since its very success shows them to be irrelevant for the phenomenon to be explained. (Friedman 1953:14–15)

Friedman therefore argues against the position that theories can be assessed by the realism of their assumptions (41). For Friedman, the only test of a theory is whether it leads to hypotheses that yield accurate predictions. This view would seem to contrast in two striking ways with past and present practices in economics. Although some economic theories are notorious for their failure to predict real-world events, their adherents often continue to espouse them. Furthermore, much of economics consists of exploring the mathematical implications of different models without much concern about how such models might be tested using empirical data. What is of relevance here, however, is Friedman’s deliberate lack of concern about realistic descriptions of choice and his implicit disdain of holistic methods, such as those prized in anthropology and history, that emphasize complex interrelationships of culture, institutions, and decisions. The continuing dominance of Friedman’s way of thinking about economics can be seen through an examination of the introductory sections of three leading contemporary textbooks.1 Economics, Principles, Problems, and Policies (McConnell and Brue 2005) is currently the best-selling textbook in the field. Economics (Samuelson and Nordhaus 2005) is a recent incarnation of what was the standard textbook in economics for decades. Principles of Microeconomics (Mankiw 2005) is written by a Harvard professor who runs a popular blog and was chair between 2003 and 2005 of President Bush’s Council of Economic Advisers.

How Important Is Decision Making? 29

All three textbooks begin by emphasizing the centrality of decision making to their field: Economics is the study of how society manages its scarce resources. In most societies, resources are allocated not by an all-powerful dictator, but through the combined actions of millions of households and firms. Economists therefore study how people make decisions: how much they work, what they buy, how much they save, and how they invest their savings . . . [M]aking decisions requires comparing the costs and benefits of alternative courses of actions. (Mankiw 2005:4–5) [A]t its core, economics is the science of choice. (Samuelson and Nordhaus 2005:xviiii) Economics assumes that human behavior reflects rational self-interest . . . Self-interested behavior is simply behavior that enables a person to achieve personal satisfaction, however it may be derived. (McConnell and Brue 2005:4)

Mankiw as well as McConnell and Brue early on make statements about assumptions mirroring the views of Friedman: Assumptions can simplify the complex world and make it easier to understand. To study the effects of international trade, for example, we may assume that the world consists of only two countries and that each country produces only two goods. Of course the real world consists of dozens of countries, each of which produces thousands of different types of goods. But by assuming two countries and two goods, we can focus our thinking on the essence of the problem. . . . The art in scientific thinking—whether in physics, biology, or economics—is deciding which assumptions to make. . . . [E]conomic models omit many details to allow us to see what’s truly important. (Mankiw 2005:21–22) Economic principles, or theories, are abstractions—simplifications that omit irrelevant facts and circumstances. Economic models do not mirror the full complexity of the real world. The very process of sorting out and analyzing facts involves simplification and removal of clutter. Unfortunately, this “abstraction” leads some people to consider economic theory impractical and unrealistic. This is nonsense! Economic theories

30 Anthropology, Economics, and Choice

are practical precisely because they are abstractions. The full scope of economic reality is too complex and bewildering to be understood as a whole. Economists simplify—that is, develop theories and models—to give meaning to an otherwise overwhelming and confusing maze of facts. Theorizing for this purpose is highly practical. (McConnell and Brue 2005:8)

From a microeconomic perspective, a key—if often unstated— assumption of many analyses concerns who exactly makes decisions of interest. The textbooks limit decision makers to a few entities that are said to interact in particular ways in market economies: Firms decide who to hire and what to make. Households decide which firms to work for and what to buy with their incomes. These firms and households interact in the marketplace, where prices and self-interest guide their decisions. (Mankiw 2005:9) [Microeconomics is] the branch of economics, which today is concerned with the behavior of individual entities such as markets, firms, and households. (Samuelson and Nordhaus 2005:5) [T]he market economy has two groups of decision makers: households and businesses, . . . (McConnell and Brue 2005:34)

From a macroeconomic perspective, the key assumptions revolve around what are regarded as the major institutions of a society that affect decision making. The textbooks assume state societies, which they divide into those with market systems and those with centrally planned economies. The texts are almost entirely devoted to market systems, which seem to be assumed to be superior to “command” economies: Every society needs to develop an economic system—a particular set of institutional arrangements and a coordinating mechanism—to respond to economizing problems. Economic systems differ as to (1) who owns the factors of production and (2) the method used to coordinate and direct economic activity. There are two general types of economic systems; the market system and the command system. . . . The alternative to the market system is the command system, also known as socialism or communism. (McConnell and Brue 2005:33)

How Important Is Decision Making? 31

This textbook focuses primarily on the market economy of industrialized nations. Before the rise of the market economy, going back to medieval times, aristocracies and town guilds directed much of the economic activity in Europe and Asia. However, about two centuries ago, governments began to exercise less power over prices and production methods. Feudalism gradually gave way to markets, or what we call the “market mechanism” or “competitive capitalism.” (Samuelson and Nordhaus 2005:25) Principle 6: Markets are usually a good way to organize economic activity. (Mankiw 2005:9)

What Anthropologists and Sociologists Refuse to Ignore If there is anything that exemplifies a certain common style in ethnographicallyoriented approaches to culture and society today, and sets them apart from other kinds of social science, it is the habit, irritating to colleagues in some other disciplines, frustrating to students, deemed perverse by potential funders, and bewildering to the public, of responding to explanations with the remark, “We need to complicate the story.” keane 2003:222

Anthropologists and sociologists have long been interested in the relationships among social institutions, cultural norms, and individual decision making. Some scholars emphasize how cultural norms and social institutions constrain decision making; others focus on how norms and institutions change over time as a result of individual choices in new circumstances. Anthropologists and sociologists have also written extensively about the related question of the extent to which the choices of individuals matter in the context of large-scale regional and global economic and technological historical changes. The general issue of the relative power of individual actions and systemic forces is often called the problem of structure and agency. In the first half of the twentieth century, sociocultural anthropologists placed little emphasis on individual decision making. Anthropologists (including Benedict 1934, Malinowski 1922) generally agreed that Western economic theory was mistaken in its fundamental assumptions about the universality of self-interested decision making. Most thought

32 Anthropology, Economics, and Choice

instead that economic behavior in tribal and peasant societies was determined by culturally specific institutions and norms. Sol Tax’s ethnographies of Mayas in Panajachel, Guatemala, in the 1930s and 1940s were a rare exception. According to Tax (1953), the Mayas were “penny capitalists” who attempted to economize and maximize in ways similar to those of entrepreneurs in industrial countries. Starting around 1950, increasing numbers of anthropologists began to think more carefully about the relationships among social institutions, cultural norms, and economic decision making. Raymond Firth’s distinction between “social organization” and “social structure” (1951) is perhaps the best-known attempt at this time to reconcile studies of individual choices and analyses of larger systems. Firth used the term “social organization” idiosyncratically to refer to the continuing choices that individuals make when faced with new opportunities. “Social structure” consists of long-standing institutions such as kinship systems, land tenure rules, and political organizations. As new opportunities arise, individuals make choices (social organization) that eventually lead to long-term institutional changes (social structure). Firth’s ideas were developed further by Fredrik Barth (1967) and Gerald Britan and Bette Denich (1976), who argued that individuals with similar goals and choices tend to make similar decisions. The result is changing cultural patterns that ultimately affect social relationships and institutional forms. Many anthropologists studying economics in tribal and peasant societies agreed with Firth and Barth about the importance of studying individual decision making. Others, however, continued to emphasize the ways in which cultural norms and social institutions constrained choices. This disagreement led to the famous, bitter, formalist-substantivist debate of the 1960s and 1970s. The most fundamental issue in this debate involved the definition of “economics.” For the formalists, economics is about the methods used to analyze decision making in situations when resources are scarce. The core of the field consisted of theories such as rational choice, which could be applied to people everywhere. The substantivists, as their name suggests, thought that economics is about the description of the operations of particular institutions associated with production, consumption, and especially exchange at different places and times. The substantivists’ most important claim was that in places where markets are either nonexistent or unimportant, economic choices are embedded within social institutions that prescribe culturally appropriate ways of making decisions about production, exchange, and consump-

How Important Is Decision Making? 33

tion. Most choices do not involve comparisons of costs and benefits as western economic theory would predict; they are instead dictated by cultural norms. The substantivists therefore paid little attention to the process of decision making; their focus was on social institutions associated with economic exchanges. In the years since the heyday of the formalistsubstantivist debate, anthropologists and sociologists have extended this line of argumentation to include western societies. These scholars say that even in societies with markets, many important resource allocation decisions are influenced greatly by institutional structures and cultural norms. The concept of “embeddedness” has recently been called the “central organizing concept” in economic sociology (Krippner and Alvarez 2007:220), a field that focuses on western, market-based societies. In a programmatic article that introduced the idea of embeddedness to many sociologists, Mark Granovetter restates the formalist-substantivist debate: How behavior and institutions are affected by social relations is one of the classic questions of social theory. . . . Much of the utilitarian tradition, including classical and neoclassical economics, assumes rational, self-interested behavior affected minimally by social relations. . . . At the other extreme lies what I call the argument of “embeddedness”: the argument that behavior and institutions are so constrained by ongoing social relations that to construe them as independent is a grievous misunderstanding. (1985:481–482)

Granovetter briefly considers the relationship between embeddedness and modernization. He says the level of embeddedness of economic behavior in nonmarket societies is less than is claimed by substantivists (1985:482– 483). Furthermore, modernization changes this level less than the substantivists believe. However, the level of embeddedness in all societies past and present is greater than is allowed for by formalist anthropologists and—more importantly—most economists. Although Granovetter felt compelled to offer this opinion, he really is not much interested in nonmarket societies. His main concern is the ways in which transactions in market societies are subsumed within hierarchically organized firms through networks of interpersonal relations. Granovetter distinguishes his network-oriented approach from both the “undersocialized” views of formalists and the “oversocialized” views of substantivists:

34 Anthropology, Economics, and Choice

[D]espite the apparent contrast between under- and oversocialized views, we should note an irony of great theoretical importance: both have in common a conception of action and decision carried out by atomized actors. In the undersocialized account, atomization results from narrow utilitarian pursuit of self-interest: in the oversocialized one, from the fact that behavioral patterns have been internalized and ongoing social relations thus have only peripheral effects on behavior. That the internalized rules of behavior are social in origin does not differentiate this argument from a utilitarian one. (1985:485)

In recent years, Granovetter has decided that the concept of embeddedness is so vague as to be meaningless (in Krippner et al. 2004). Nonetheless, his ideas remain influential in economic sociology. Most contemporary economic sociologists criticize classical and neoclassical economics for reasons similar to those of the substantivists. However, the interests of these two groups of scholars differ. Substantivists and their successors in anthropology examine social institutions and (to a lesser extent) cultural norms; economic sociologists focus on social networks. Although anthropologists have written extensively about social networks (as in Barnes 1972, Mitchell 1974, White and Schweizer 1998), they have been more concerned about the extent to which individuals can affect and resist global technological and economic changes (for example, Comaroff 1985, Moberg 1992). This latter question arises in part because of the growing influence of political economy analyses that emphasize the actions and goals of entities such as states, multinational corporations, political parties, insurgent groups, and development agencies. In their textbook on economic anthropology, Richard Wilk and Lisa Cliggett explain well the challenges such analyses pose for scholars interested in local decision making: On . . . [a] global scale, history appears as conflicting forces, as modes of production colliding with one another across the landscape, and there often seems little room for human action . . . [H]uman actors seem like tiny cogs on giant gears in some satanic factory. How can we put people back into the picture? . . . How can we account for the reality, power, and force of social structure and the same time grant some “agency,” some autonomous decision-making power to individuals? (2007:112)

Scott (1990) and others have examined how local people in certain places and times have attempted to improve their lives through group

How Important Is Decision Making? 35

action and organized resistance. There are lively debates among those who study such social movements about the degree to which common religious and cultural beliefs, especially those grounded in ethnicity, underlie the formation of effective local resistance to outside pressures. In some cases, common interests lead people from diverse groups to band together; in other situations differences in ethnicity, class, and politics prevent effective actions. The question of human agency is also part of a branch of anthropology called “practice theory” (Bourdieu 1977; Ortner 1984, 2006). This theory aims at exploring relationships between human actions (practice) and global entities (“the system”), with equal attention to the impact of the system on practice and the impact of practice on the system (Ortner 1984:148). Practice theory developed in reaction to what were regarded as defects in the prevailing anthropological theories of the 1960s and 1970s: [T]hey all had one thing in common: they were essentially theories of “constraint.” Human behavior was shaped, molded, ordered, and defined by external social and cultural forces and formation: by culture, by mental structures, by capitalism. . . . [A] purely constraint-based theory, without attention to either human agency or to processes that produced and reproduced these constraints—social practices—was coming to seem increasingly problematic. (Ortner 2006:1–2)

Practice theory has only rarely been directly concerned with individual economic decision making. Nonetheless, the issues addressed by practice theorists are clearly relevant to studies of choice.

Conclusions Some decisions obviously matter a lot. No one would deny the significance of the vote of the Supreme Court of the United States in the 1950s outlawing segregation in public schools or the decisions that many financial institutions made in the past decade to invest in risky hedge funds. Other decisions do not matter as much. Even I find it hard to care much about my choice of which route to take when bicycling the mile between my house and the university where I work. The extent to which a particular choice matters depends on both its consequences and the number of people affected. The choices of powerful decision makers such as the president of an oil company or the members of a national legisla-

36 Anthropology, Economics, and Choice

ture can have dramatic consequences for the lives of millions of people. Less powerful individuals and groups, in contrast, often find themselves in situations in which many choices they make have little effect on even their own lives. Although economists focus more directly than anthropologists on the study of choice, most scholars in both fields agree on the importance of studying significant decisions. The fields differ starkly, however, in how best to study such decisions. Economists attempt to create parsimonious models of choice that involve a limited number of clearly defined, measurable variables that can be tested with empirical evidence. They regard “economy” and “society” as separate spheres, each requiring its own distinctive type of analysis. In their models of choice, cultural norms and social institutions are treated as external factors that are relevant only in as much as they affect the values (“utilities”) that decision makers place on alternative outcomes. Economists are therefore rarely concerned with the complex interactions among cultural norms, social institutions, and decision making. They pay little attention to the ways in which choices are embedded in social institutions and the effects of cumulative decision making on the evolution of cultural rules and customs. Anthropologists place much more emphasis on the ways in which choices are constrained by cultural norms and social institutions. In the past, a few anthropologists interested in decision making (formalists) attempted to mimic the assumptions and methods of economists. However, the great majority of contemporary economic anthropologists— including many using mathematical models of choice—reject some or all of the methods and assumptions of economics. There is general agreement among anthropologists that most economists’ models distort reality by ignoring many variables that affect choices. Unlike Milton Friedman, they do not regard such distortions as a scientific necessity. Almost all anthropologists think that treating “economy” and “society” as separate categories is a basic analytic error. Economists are also sometimes scorned by anthropologists for their disregard of empirical evidence in their casual assumption that theories developed to describe behavior in market societies can be easily applied everywhere. Perhaps most importantly, many anthropologists and other social scientists think that economists fail to sufficiently emphasize the extent to which individuals’ choices are constrained by historical circumstances and power differentials between the rich and the poor. Although I agree with most other anthropologists about these shortcomings of economic theory and practice, I nonetheless would argue that

How Important Is Decision Making? 37

the study of choice is of fundamental importance in understanding both everyday life and cultural evolution.2 People in all societies are continually making conscious choices in which they weigh the costs and benefits of different options. As Firth (1951) pointed out long ago, cultural transformations are in part the results of economic decisions that individuals and groups make in changing circumstances. Most of the rest of this book compares how economists, anthropologists, and other social scientists look at different types of decisions. I begin with the complicated choices that individuals in many places must make between paid and unpaid work.

CHAPTER 2

Choices between Paid and Unpaid Work

Social scientists analyzing diverse economic decisions have often attempted to estimate the monetary value of unpaid labor and production for home consumption. In this chapter I discuss the reasons for these attempts and examine some of the theoretical and methodological problems associated with such efforts. My focus is on estimates of the value of three types of unremunerated work—subsistence agriculture, non-timber forest products (NTFPs), and domestic labor (housework). In order to understand why anyone would want to make such estimates, it is helpful to consider certain measurement problems that arise when economists attempt to make cost-benefit analyses. People making difficult decisions must weigh the advantages and disadvantages of alternative choices they might make. The method of costbenefit analysis (Boardman et al. 2006, Nas 1996) was developed to aid in such decisions. The basic ideas of cost-benefit analysis, which come directly from expected utility theory, are not hard to understand. The first step is to assign numerical values to the advantages and disadvantages associated with different possible courses of action. Next, estimates are made of the probability of different outcomes if particular options are selected. Finally, calculations based on these values and estimates indicate which possible choice has the most favorable cost-benefit ratio. In practice, cost-benefit analyses can be extraordinarily difficult. Most decisions have multiple costs and benefits, which may not all be easily measured by one metric such as money or time. A good example of this is a restaurant customer choosing between an expensive steak and a cheaper, less fattening salad. Costs and benefits influencing this decision might include price, taste preferences, and calories. The weight any individual customer assigns to each of these considerations varies; a rich patron of

Choices between Paid and Unpaid Work 39

a restaurant is less likely to be influenced by cost than someone who can barely afford eating out. But even if one somehow knows how important each factor is in influencing the decision of a particular customer, their advantages and disadvantages may seem incommensurable. The monetary cost and caloric content of steaks and salads, for instance, are obviously measured in different units. Moreover, even figuring out the unit with which to measure some variables such as taste preferences can be difficult. Economists conducting cost-benefit analyses usually emphasize inputs and outputs that can be measured monetarily. When confronted with a variable measured in other units (time, noise, odor, caloric content, taste preference), economists often attempt to devise a way to “impute” a monetary value. When imputation seems impossible, such costs and benefits may be omitted from an analysis. There are good reasons for such omissions. Economists’ models are simpler if the assumption is made that a decision-maker is attempting to maximize one single thing such as income from sales.1 In the societies where most economists work, monetary costs and benefits are a major influence on many important decisions. Scholars critical of conventional cost-benefit analyses such as Waring (1996) argue that the focus on easily measured monetary inputs and outputs results in economic analyses that neglect significant factors affecting certain choices. In particular, subsistence production and unpaid labor are often ignored in both input/output analyses and statistical measures such as gross domestic product. This neglect may influence public policies that affect people such as small-scale farmers and domestic workers. Some of these scholars therefore think it is useful to attempt to impute a monetary value to subsistence production and unpaid labor.

Ethnicity, Subsistence Production, and Wage Work in Belize I did not become interested in estimating the value of subsistence production and unpaid labor as a result of worrying about the shortcomings of the analytic methods of professional economists. Instead I was trying to understand a practical problem I encountered in my research—the reasons for differences in the economic activities of diverse ethnic groups in British Honduras (now the independent country of Belize) in the early 1970s. Belize is today a laid-back tourist destination renowned for its snorkeling, nature reserves, and archaeological sites. But at the time of my fieldwork British Honduras was a colonial possession scorned by travelers

40 Anthropology, Economics, and Choice

as a bizarrely isolated backwater. Writers about the colony were fond of Aldous Huxley’s sardonic observation that “if the world had any ends, British Honduras would certainly be one of them. It is not on the way from anywhere to anywhere else. It has no strategic value. It is all but uninhabited” (1934:32). The views of world travelers about British Honduras had not changed much in the almost four decades that elapsed between Huxley’s remarks and my fieldwork. Even one of my graduate school advisers reacted with puzzlement when I announced my research plans. “Why would you want to go there?” he asked. “It’s a pit,” he went on to say, doubtless comparing Belize City and its open sewers to the beaches of the well-known Caribbean islands where he conducted research. What attracted me to British Honduras were the historical circumstances that led to the isolation and desolation that bemused Huxley, my adviser, and other visitors. Although there were once sizable numbers of Mayas in the area that is now Belize, the region was sparsely populated when Europeans came to the Americas. The Spanish did not settle the region in significant numbers; instead most European residents in the seventeenth and eighteenth centuries were British foresters who extracted first logwood and later mahogany. The woodcutters needed a labor force, and starting around 1700, slaves were brought into British Honduras, mostly from Jamaica. After emancipation in 1833 the descendants of slaves continued to work in forests, often raising small plots of subsistence crops to feed their families. The largest population group in British Honduras in the nineteenth and early twentieth centuries was “Creoles” of mixed European and African ancestry. By 1970 the colony’s peoples were a complex ethnic mixture. Although about half the 100,000 or so residents were Creoles, there was a rapidly growing Spanish-speaking mestizo population in the north and west that was culturally similar to the Ladinos of neighboring Guatemala. About 10,000 “Black Caribs” (now called Garifuna) lived in the south of British Honduras. They were descended from escaped slaves of African origin and the indigenous peoples of the Caribbean islands of St. Vincent and Dominica. At the end of the nineteenth century, the Caribs were evicted from the Caribbean after being on the wrong side in conflicts between the British and the French. They settled along the Atlantic coast of Central America; most are bilingual in Carib (an indigenous language of South America) and either English or Spanish. Mayas who had migrated from either Guatemala or Mexico comprised another 10 percent of the population; they were usually bilingual in one of several Maya languages

Choices between Paid and Unpaid Work 41

and either English or Spanish. There were also numerous smaller ethnic groups in the colony, including German-speaking Mennonites and expatriates from Britain, the United States, and elsewhere. By 1971 the British were eager to abandon their colony in Central America. Britain had been pumping money into British Honduras ever since the demand for mahogany had lessened in the first part of the twentieth century. The place remained a colony only because of a longstanding territorial dispute between Britain and Guatemala. During the colonial era the British claim of part of the Central American mainland was disputed by Spain. The territory remained contested after the independence of most Latin American nations. In 1859 an agreement was reached in which British sovereignty would be recognized if Britain constructed a road connecting Guatemala City and the city of Belize on the Caribbean Sea. Because the road was never built, Guatemala continued to press its claims. Only the presence of the British army had deterred an invasion on several occasions. Although British Honduras was internally self-governed in the early 1970s, the fear of an attack by Guatemala prevented Belizeans from pursuing complete independence. British colonial officials and the internal Belizean government thought that agricultural development was necessary to improve the local economy and reverse the unfavorable balance of trade. Domestic shortages had forced the government in many years to import rice and red kidney beans, the staples of much of the population. Both attaining self-sufficiency in staple foods and increasing agricultural exports seemed feasible. British Honduras had a low population density and large tracts of unused, cultivable land. Diverse projects aimed at stimulating agriculture. The government built roads into remote hamlets, provided price supports, sold land to farmers planting permanent crops, formed cooperatives, and encouraged the use of fertilizers, herbicides, pesticides, and new types of seeds. These programs had limited success. Norman Ashcraft, an anthropologist who carefully examined agricultural development programs during this period, reported: In contrast to expectations, small-scale producers have not responded to guaranteed price incentives, for their contributions to government purchases [by the relevant state institutions] has been declining rather than increasing. It seems evident that most cereal farmers continue to sell only the excess over household needs and that excess is not increasing. (1973:90–91)

42 Anthropology, Economics, and Choice

I was struck by a “cultural” reason that was sometimes offered for the failure of these programs. Many projects were aimed at Creoles living in the center and south of British Honduras. Development officers claimed that Creole males regarded agriculture as an inferior, transitory occupation that occupied time away from more important forestry work. There is evidence . . . that the organization of forest work encouraged antagonism and even bred a contempt for agriculture; the relative freedom of the forester returning to the town for his considerable rest periods contrasted with the continuous attention which cultivation involved. (Settlement Report 1948:253) [Creoles] are, perhaps, not natural farmers and do not work the land continually and in harmony with it. (Romney et al. 1959:151)

Two themes were implicit in these assertions. When Creoles were described as “not natural farmers,” they were being compared to the Mayas, who planted larger fields and worked less as wage laborers. While the Creoles’ alleged preference for wage labor was thought to have been reasonable at earlier times when opportunities for forest work were abundant and incentives for cash cropping minimal, the continued aversion to farming was regarded as an irrational obstacle to economic development. Ashcraft, however, thought that the Creoles were sensible in not responding much to agricultural incentives (1973:91). He noted the poor road system, the high cost of transportation, and a cumbersome marketing board system that made it difficult for farmers to sell crops to the state at fixed prices. Ashcraft also observed that there were increasing employment opportunities on sugar and citrus plantations. Creoles, he suggested, might reasonably prefer the ready cash from wage work on commercial farms to the uncertain returns from their own farming. Although Ashcraft’s book on agricultural development in British Honduras came out after I finished my fieldwork, I knew about his views prior to conducting research from his doctoral dissertation (Ashcraft 1968) and personal correspondence. My initial reaction on reading about this controversy was to side with Ashcraft. I was at the time (and for the most part remain) committed to materialist approaches to economics. I assumed that human groups faced with changing material circumstances will ordinarily alter their economic strategies in attempts to solve whatever new problems arise. From this perspective, I thought that the cultural “traditions” of particular ethnic groups would not usually pose a significant

Choices between Paid and Unpaid Work 43

obstacle to economic development programs. I was, however, willing to consider that I might be wrong in particular cases and thought this might be an empirical question that I could investigate as part of my fieldwork. I eventually decided to focus my research on how Creole, Carib, and Maya residents of three villages in the Stann Creek (now Dangriga) district of southern British Honduras made decisions over the course of a year about how to allocate their household labor. A key part of my analysis consisted of calculations of input/output ratios for diverse economic activities. I was especially interested in comparing the monetary returns per unit of labor for farming and wage labor. Were Creoles, for example, preferring wage labor to farming even in circumstances where agriculture seemed to offer better returns? Or were their choices between these activities just the results of straightforward cost-benefit comparisons unrelated to Creole culture and history? Although I never thought that answering this question would be easy, I learned during fieldwork how naive I had been in framing my research problem. Members of village households rarely were faced with direct choices between wage work and farming. Some participants in farming— the elderly, children, most women—rarely had the option to do wage work. Moreover, agriculture had a seasonal cycle in which periods of heavy labor demands only sometimes coincided with opportunities for wage work. Even when villagers decided to do wage work during a peak part of the agricultural cycle, they often were able to persuade family members to help out with tasks such as clearing fields, planting, or harvesting. Comparing the returns per unit of labor for wage work and farming in any case was not simply a matter of attempting to calculate the amount of money earned per hour. Villagers talking about the advantages and disadvantages of wage labor and farming did not restrict their discussion to the potential cash income from these two types of work. Agriculture in British Honduras, as everywhere, was a risky undertaking; yields and prices for crops varied from field to field and from year to year. Income from farming came many months (or in the case of tree crops, years) after fields were cleared. Wage labor, in contrast, provided a reliable, quick source of income. Farming required a labor commitment of at least five or six months; wage labor could be taken on for short periods when money was needed. The theoretical and methodological problem that puzzled me the most, however, was how to place a “value” on subsistence agriculture—crops that were not sold because they were grown for home consumption. Suppose, for example, that someone (either a local farmer or an outsider such

44 Anthropology, Economics, and Choice

as an agricultural economist) wanted to calculate the returns per labor unit of an acre of corn.2 For corn sold at market, one could make a rough calculation by estimating the yields, work input, and selling price. But what if the corn was eaten at home or fed to animals? And what should an analyst do if some corn was sold at market and some was consumed at home? This analytic problem consisted of two distinct questions. Could an estimate of monetary value be made for subsistence agriculture? If so, how should such an estimate be made? These were not idle academic inquiries. Most rural households in British Honduras obtained some of what they ate by farming, fishing, and hunting. By so doing, they reduced the amount of money that had to be spent on food. Clearly, members of such households choosing between wage work and other activities were influenced by their perceptions of the value of subsistence production. In my thesis I eventually reached the following conclusion about the value of subsistence agricultural production among rural residents of British Honduras: [I]f one wishes to calculate the total value of a farmer’s harvest [nowadays I would say “the harvest of a farm household”], . . . the value of a given unit of a crop used for home consumption is not equal to [the value of ] the same unit sold at market. If it is assumed that agricultural activity and diet are independent of one another [in retrospect a shaky assumption], food that is needed for subsistence purposes, but is not grown, must be purchased. For this reason crops consumed at home should be valued at the price they sell for at stores and not at the price the farmer gets when he [Of course, today I would not use the male pronoun.] sells them. (Chibnik 1975:81)

Although I realized that other researchers must have confronted similar problems, I made no attempt to search the scholarly literature to see how this issue was ordinarily handled. A few years later, I decided that this issue was of sufficient general interest that it would be worthwhile to examine the value of subsistence agricultural production in more depth (Chibnik 1978).

Imputing Value to Subsistence Production Although it is obviously impossible to place a monetary value on labor and subsistence production in societies without cash, numerous attempts

Choices between Paid and Unpaid Work 45

have been made to measure the “efficiency” of different types of foodgetting methods in such places. These efforts were perhaps most visible in the 1960s and 1970s when scholars from diverse disciplines were comparing the caloric returns per unit of labor input and per unit of land of different subsistence techniques in their efforts to develop theories about cultural evolution (for example, Boserup 1965, Harris 1977) and “optimal foraging” (reviewed in Smith 1983). For example, the anthropologists Richard Lee and Allen Johnson calculated labor inputs and caloric returns for various ways to get food among the !Kung in the Kalihari desert in Botswana (Lee 1979) and the Machiguenga in the Peruvian Amazon (A. Johnson 1978). Their sophisticated, meticulous comparisons of input/ output ratios suggested that labor devoted to certain food-getting methods was more efficient in certain respects than that devoted to alternative subsistence techniques. The analysts of energetic efficiency rarely equated efficiency with “value.” An examination of one of Johnson’s findings suggests why they were reluctant to make this seemingly obvious inference. Johnson noted (1978:81) that the Machiguenga obtained about 45 calories growing maize for every calorie of work effort; the comparable return for hunting and gathering was 0.8 calorie. This would suggest that labor on maize was much more valuable than that spent hunting and gathering. One wonders why the Machiguenga ever foraged at all. As Johnson explains, the reasons are not mysterious. Because foraging and corn growing can be done at different times, the Machiguenga rarely directly chose between these two options. Furthermore, the Machiguenga received various noncaloric “returns” (protein, nonfood animal and vegetable products) from hunting and gathering that were not available from farming. I think that attempts to estimate any sort of value of labor and subsistence production are fruitless in societies without money. In such places the famous distinction between “use value” and “exchange value” is relevant.3 Production for use value aims at the acquisition of specific goods; production for exchange value, in contrast, aims at the accumulation of general “wealth.” I would not want to argue that there is no production for exchange value in societies without cash. Nonetheless, the preponderance of production for use value in such societies makes attempts to compare the “value” of different types of labor or subsistence production impossible for all practical purposes. The remarkable economist A. V. Chayanov lucidly explained many years ago why the absence of money hindered the comparison of the value of different economic efforts. Chayanov was one of a large group of Rus-

46 Anthropology, Economics, and Choice

sian scholars carrying out detailed empirical studies of the agricultural activities of “peasant” households in the first part of the twentieth century.4 His work on these projects convinced him that certain aspects of classical economics could not be applied to the analysis of certain kinds of farming.5 One of his arguments (originally published in the 1920s) was that conventional calculations of profit were impossible in nonmonetarized societies: [In such circumstances] the question of the profitability of various expenditures cannot arise—for example whether growing hemp or grass would be more profitable or advantageous. For these plant products are not interchangeable and cannot be substituted for one another, therefore, no common standard can be applied to them. (Chayanov 1966:4)

Attempts to assign numerical values to unpaid labor and subsistence production seem more plausible in societies with money. One might, for example, estimate the economic value of unpaid labor on a task such as house cleaning at what it would cost to pay someone to do this work. Similarly, the economic value of tomatoes grown in a garden and consumed at home could be estimated at what it would cost to buy them in a market or store. Even in societies with money, estimates of these values are not at all straightforward. Certain types of unpaid work (repairing toasters) and subsistence production (freshly hunted meat) may not be readily available for purchase at some times and places. More importantly, Chayanov argued that household agricultural labor—and presumably other forms of unpaid work—cannot be valued at the prevailing wage rates for farmworkers (1966:88–89). Chayanov pointed out that on “capitalistic” farms, where all work is paid for, increases in labor inputs without corresponding income gains can be disastrous because profit equals gross income minus outlay on materials minus wages. On “family farms” using unpaid labor, however, increases in labor inputs without corresponding income gains do not necessarily lead to monetary losses. Because labor inputs that are disastrous in capitalistic agriculture can be acceptable on farms using unpaid workers, valuing family labor as equal to the wages of hired workers is, according to Chayanov, absurd. This leads to family farms being regarded as operating at a loss in situations when they are actually making money. Chayanov agreed (somewhat ironically, given his opposition to “capitalist” economics) with certain neoclassical economists (such as Krishna 1969) that no monetary value at all can be assigned to family agricultural

Choices between Paid and Unpaid Work 47

labor and subsistence production even in a highly monetarized context. Their argument is that not all units of subsistence labor and production are equally valuable; initial work inputs and outputs are worth more than later ones. The first part of a harvest may be necessary to avoid starvation; this is less so of later parts. Furthermore, there may be diminishing returns for additional labor inputs after a certain point for many types of work, such as weeding and clearing fields. In addition, the value of any particular input of labor may vary among households according to their particular situations. If all other things are equal, adults in a household with several young children may be willing to work harder at subsistence production (place higher values on later inputs of labor) than adults in a household with no young children. Such arguments can easily be extended to non-agricultural unpaid labor. Although such arguments are thought-provoking, they suffer from a fundamental flaw. People often do choose between subsistence agriculture and cash crops (Moerman 1968, Scott 1976:23–24) and between unpaid work at home and wage labor. In the United States, for instance, many farm households have decided to grow very little for home consumption and instead buy most of the food they eat. Analogously, couples in the United States with young children often engage in difficult decisions about whether both husband and wife should seek paid work. If they decide to do so, paying for child care may be necessary. The alternative usually involves one member of the couple engaging in paid work (still usually the husband) and the other staying at home and caring for the children. Although such choices are clearly influenced by nonmonetary considerations, it would be foolish to deny that costs are also usually relevant. If people sometimes choose unpaid labor and subsistence agriculture in preference to wage work and cash cropping, they must be at least in part subjectively comparing the monetary implications of their options. The question remains as to how an outside analyst should place a monetary value on unpaid work in such cases.

The Value of Crops Consumed at Home In my article on the value of subsistence production (Chibnik 1978), I discussed the then-common practice of valuing crops consumed at home at the prices for which they could be sold. I wrote that the economist John Mellor had made convincing arguments against this practice. According to Mellor (1966),

48 Anthropology, Economics, and Choice

The farmer correctly attaches a higher price to production for home consumption than to production for sale since he [sic] pays the retail price for what he buys and receives the wholesale price for what he sells. (200)

Mellor illustrated his point by discussing a hypothetical farmer growing both wheat and cotton. This farmer sells only a small portion of the wheat, consuming the rest at home. If the price of wheat dropped relative to that of cotton, it might be reasonable for the farmer to reduce wheat production and increase cotton production until no more wheat is produced for sale. Mellor then argued: In considering further shifts of resources out of wheat, the farmer should compare not the wholesale [selling] price of wheat, but the [retail] price he would have to pay if he purchased wheat . . . No further cut in wheat production below subsistence needs will be made until the relative price of wheat has dropped by at least the difference between the buying and selling price of wheat. (Ibid.)

Mellor thus suggests that the subsistence production of wheat (in this case) should be valued at its retail price. I concluded that such a valuation was “the most plausible procedure that had been suggested” (Chibnik 1978:569). In the three decades since I published this piece, social scientists have written extensively about estimating the economic value of unpaid domestic work, non-timber forest products, and hunting and fishing for home consumption. Very little new has been said, however, about placing a monetary value on crops and animals raised for subsistence purposes by households participating in market economies. I therefore restrict my discussion about agriculture in what follows to as yet unresolved methodological questions about problems associated with valuing subsistence production at retail prices. Farming is inherently a risky enterprise. Yields vary depending on weather conditions, the extent to which weeds and pests can be controlled, and the effectiveness of capital expenditures such as irrigation and fertilizer. When crops are sold, there is inevitably uncertainty about prices, even when governments provide a guaranteed market. Farmers in market economies thinking about the advantages and disadvantages of different crops therefore must consider variability in yields and prices as well as average returns to inputs of land, labor, and capital.

Choices between Paid and Unpaid Work 49

Long-established (“traditional”) subsistence crops are often less risky than agricultural products sold at markets. Time-tested varieties of staple crops usually provide adequate yields in varying environmental conditions. Cash crops, in contrast, may be recent introductions to an area that is subject to blights and provide good yields only in optimal growing conditions. Considerable uncertainties regarding prices are usually associated with cash crops; variation in selling prices is obviously not a risk factor for staple crops consumed at home. Most importantly, staple crops have high use value because they are eaten; many cash crops (coffee, tea, cotton, tobacco) have relatively little use value if not sold. All other things being equal (yields per unit of land and labor, selling prices at market), farmers are therefore likely to prefer staple food crops to cash crops. If farmers’ decisions are influenced by both risk factors and differences between buying and selling prices of staple crops, they therefore should value subsistence crops even higher than retail price. Valuing subsistence crops at market prices makes the debatable assumption that if they are not grown, they must be purchased. Even in places where households grow much of what they eat, the relationship between food consumption patterns and the availability of different crops is not self-evident. In order to ensure that most of what they consume is homegrown, some households may alter their diets and eat less-preferred subsistence foods. Others may prefer to purchase food if necessary to maintain their customary diets. If a household fails to produce enough of a staple crop to meet its preferences, the most likely response of its members might well be both to purchase some food and to shift diets. Such a reaction challenges the assumption of unchanging consumption patterns that underlies the valuation of subsistence crops at retail price. Retail prices of agricultural products are usually lowest at harvest time, when supplies are high. Even if one assumes that subsistence crops should be valued at their retail prices, one should probably not use the price at harvest. Instead, some other method (perhaps the average price over the course of a year) seems preferable. Some crops (corn, soybeans) that are grown in part for home consumption can be fed to domestic animals. Certain products from these animals (meat, eggs, milk, cloth) are consumed at home; others are sold at market. Estimating the monetary value of crops fed to animals is a daunting methodological task. Nonetheless, this value (however calculated) differs from that of both crops consumed at home and those sold at market. Researchers assigning a value to subsistence production must consider what proportion of a crop is consumed by domestic animals.

50 Anthropology, Economics, and Choice

Non-Timber Forest Products The destruction of tropical rainforests is widely recognized as one of the major environmental problems of our time. Much of this damage is caused by entrepreneurs attempting to either earn money or gain land rights through the sale of timber and the use of formerly forested land for crops and cattle raising. Environmentally oriented scholars from diverse disciplines (among them Godoy and Lubowski 1992:423 and Peters, Gentry, and Mendelsohn 1989:655) have observed that policy makers often assume that tropical forests have no value unless they are logged or farmed. These scholars have emphasized the ecological importance of non-timber forest products such as wild food plants, honey, resin, spices, game, firewood, charcoal, and raw materials for handicrafts. While NTFPs are not always environmentally sustainable, they ordinarily cause much less ecological damage than either timbering or agriculture. Ecologists have argued that NTFPs are economically valuable. In an influential article about NTFPs that appeared in the prestigious British science journal Nature, the ecologists Charles Peters and Alwyn Gentry and the economist Robert Mendelsohn attempted to calculate the market value of all the tree species in one hectare in the Peruvian Amazon. They concluded that tropical forests are worth considerably more than has been previously assumed and . . . the actual market benefits of timber are very small relative to those of non-wood resources. Moreover, the total net revenue generated by the sustainable exploitation of “minor” forest products are two or three times higher than those resulting from forest conversion [logging, cattle raising, and farming]. (1989:655)

In the late 1980s, enormous publicity was being given to the worldwide disappearance of tropical rainforests. The findings of Peters, Gentry, and Mendelsohn were described in a lengthy article in the science section of the New York Times (July 4, 1989) and were cited in a sharply worded editorial in that newspaper about the dangers of excessive logging. When Peters, Gentry, and Mendelsohn assigned a monetary value to the NTFPs in a small plot of Amazonian rainforest, they made several simplifying assumptions. They considered only tree products and ignored other NTFPs. Their calculations were restricted to products that could be sold at markets; they did not consider items that were used exclusively for home consumption. In cases where NTFPs or wood could be

Choices between Paid and Unpaid Work 51

both sold and consumed at home, no attempt was made to measure local people’s actual allocation of these products between market sales and subsistence use. Perhaps most importantly, the authors did not discuss the extent to which an increased production of NTFPs might lower market price. The limited demand for many NTFPs suggests that valuing them at current market price overestimated their potential value in comparison with timber. My focus here is on the neglect by many authors discussing NTFPs— exemplified by Peters, Gentry, and Mendelsohn—to examine carefully the value of forest products used exclusively or primarily for subsistence. In a recent perceptive article examining the economic value of unmarketed NTFPs Charles Delang argues, In many communities . . . NTFPs that are directly consumed play a more important role in the livelihood of the population than the cash earned with the sale of NTFPs or other commodities. Indeed, the NTFPs that are consumed rather than sold on market can be considered income in kind rather than in cash. Thus, ignoring the role of NTFP consumption in the livelihoods of rural populations gives a very distorted view of the importance of NTFPs—and of their economic values. It also gives the wrong message that everything that does not go through the market, and does not have a market price, does not have an economic value, and therefore is not worth protecting. (2006:65)

Delang goes on to review several methods that have been tried to estimate the economic value of unmarketed food plants.6 The three methods he discusses that are most commonly used—opportunity cost, contingent valuation, and substitute product value—all are associated with significant theoretical and methodological difficulties. Furthermore, they result in strikingly different conclusions about the economic value of particular NTFPs. Opportunity cost. Scholars using this method (for example, Adamowicz et al. 1998) measure the amount of time local people spend collecting NTFPs for subsistence purposes. Researchers then place a monetary value on this time, ordinarily the local wage rate. In so doing, they are using the same method of valuing unpaid labor that Chayanov criticized so strongly. This method makes the questionable (“opportunity cost”) assumption that time collecting NTFPs could otherwise be allocated to earning money via wage labor. Moreover, it seems counterintuitive to as-

52 Anthropology, Economics, and Choice

sign different monetary values to the same NTFPs in a particular place as wage rates change. An additional methodological problem arises because people often combine trips to collect NTFPs with other activities. Determining what percentage of their time on such trips should be counted in calculations is not straightforward. Contingent valuation. This method, which does not seem to me to be particularly useful in the case of NTFPs, has been used to estimate the value of diverse environmental “goods,” including protected areas (Dharmaratne, Sang, and Walling 2000) and whales (Bulte and Van Kooten 1999). In the language of ecological economics, “consumers” are asked how much compensation they would be willing to accept for the loss of environmental benefits such as clean water or access to a beach. For unmarketed NTFPs, this would mean asking people what they considered to be a fair market price for these products. Most scholars who have seriously considered the application of this method to NTFPs (for example, Delang 2006:65, Godoy and Lubowski 1992:427) have a hard time imagining that sensible answers could be given to such inquiries. Substitute product value. This method (Godoy, Lubowski, and Markandya 1993:272) involves finding marketed NTFPs that are “similar” to the unmarketed NTFP whose value is being estimated. The unmarketed NTFP is assigned either the selling or retail monetary value of the similar marketed product. When the retail price is used, this is analogous to Mellor’s method for placing a monetary value on agricultural crops produced for home consumption. The obvious methodological problem here is identifying marketed NTFPs that can be considered close substitutes for unmarketed NTFPs.

In a monthlong study in a Thai village, Delang compared the results of opportunity cost and substitute product methods in placing monetary values on wild plants collected by the Karen, a local indigenous group. For the opportunity cost method, Delang used the local wage rate for casual agricultural labor to impute a monetary value for the time spent collecting wild plants. For the substitute products method, Delang (with the help of the Karen) grouped 134 wild food plants that were gathered (but not marketed) into six categories—leaves, stems, roots, fruits, flowers, and bamboo shoots. The prices of all food plants sold in a nearby market town were recorded. With the help of the Karen, Delang grouped the commercial food plants into the six categories that had been devised for the NTFPs. The average price of each category of commercial food

Choices between Paid and Unpaid Work 53

plants was then used to calculated the value of the wild food plants consumed at home. The two methods resulted in surprisingly different estimates of the value of collected wild food plants. Using the opportunity cost method, Delang calculated that the value of NTFPs collected by the “average” household (5.33 members) was about fifteen days of wage labor at eighty Baht (the Thai monetary unit) per day. In contrast, the substitute production method resulted in a much higher valuation: [T]o make up for the loss of wild food plants, the Karen would have to buy food plants in the market. The (surrogate) market value of the wild food plants that the average household . . . consumes is 11,505 Baht per year. This much higher value corresponds to approximately 114 days of work at 80B/day. Thus, by working for 15 days gathering wild food plants, the Karen are able to save the same amount of money they earn in 114 days. (2006:70)

The assumption that the Karen would buy comparable food plants at market if they did not collect them is questionable. They might well instead be able to purchase cheaper foods. Nonetheless, Delang’s study clearly shows the difficulties associated with attempting to value subsistence production in particular cases.

Unpaid Labor in Industrial Societies Although economists analyzing industrial societies ordinarily focus on monetary transactions, many important tasks are carried out by unpaid workers. The best-known such tasks are chores associated with maintaining a household such as cleaning, cooking, and child rearing. But many other activities essential to the well-being of industrial societies are done by volunteers in hospitals, food banks, and other community institutions. Such unpaid labor is often culturally regarded as not being “work” (Fink 1987). The problems involved in attempts to assess the “worth” of unpaid labor exemplify the general difficulties that economists face when confronting “values” that cannot be easily measured numerically. An examination of several relevant cases in the United States shows again why such analytic problems cannot be dismissed as abstract complexities of interest only to academics.

54 Anthropology, Economics, and Choice

In the mid-1970s I eked out a living for couple of years working as a freelancer on diverse social science projects. One of my jobs was on a project funded by the National Academy of Sciences that was evaluating the federal government’s Experimental Technology Incentives Program (ETIP). As its name suggests, the ostensible goal of ETIP was to examine how small-scale changes (experiments) in government policies could stimulate technological innovation (Britan and Chibnik 1980). Many of the experiments were, to my discomfort, based on free-market assumptions that government regulations were discouraging entrepreneurialism. A typical project tried to stimulate innovation in the pharmaceutical industry by shortening the time the government took to approve new drugs. The shortened approval period in my view led to less information about the long-term effects of these drugs. I looked at one experiment, however, that did not seem to be affected by ideological considerations. Government agents in charge of buying diverse items thought that the procurement methods were not costeffective. Contracts for products such as washing machines were given to the lowest bidder, ignoring considerations such as durability, repair costs, and energy use. The idea in the experiment was to devise procurement systems for different products that considered not only initial prices but also the expected costs over their lifetimes. I was struck in particular by economists’ analyses of the lifetime costs of lawn mowers. The policy makers compiling a list of factors that they wished to include in their procurement models recognized that noise was an important aspect of lawn mower performance. Although the economists wanted to reward bidders with quieter lawn mowers, they were unable to come up with a way to put a monetary value on noise. The methods the economists created eventually omitted noise as a factor to consider in government lawn mower purchases. By so doing, they implicitly rewarded companies that did not allocate research funds and time to developing innovative ways to make lawn mowers quieter. Two more recent examples further illustrate how social policy can be influenced by difficulties in imputing monetary values to environmental variables. Over the past two decades, large-scale industrial hog lots have become a common feature of the landscape in rural parts of states including Iowa, Illinois, and North Carolina. Local social movements have opposed these facilities because of their economic effects. Because the hog lots are usually run by large corporations, they are part of an increasingly capital-intensive agriculture that discourages the continued existence of small and mid-size farms. What I want to focus on here, however, are en-

Choices between Paid and Unpaid Work 55

vironmentalists’ objections to the confinement facilities’ effects on air and water pollution. While monetary values can be placed on certain aspects of air and water pollution, the smell of hog lots has so far been omitted from economic discussions of their effects. No one also been able to show, for example, that the odor of hog lots influences nearby property values. Nonetheless, the smell certainly affects the quality of rural life. Wind power is often advocated as a sustainable energy source that is environmentally less destructive than fossil fuels and nuclear power. One might think that environmentalists would therefore applaud a proposed offshore New England project to harness wind energy. However, the Cape Wind project became famous (Williams and Whitcomb 2007) for its opposition from wealthy, politically powerful environmentalists such as the Kennedy family. The proposed wind turbines were regarded as unsightly by the usually environmentally conscious residents of nearby Nantucket Island. Although much about the environmental consequence of wind turbines can be measured monetarily (costs of construction and maintenance, value of energy produced), the financial costs of their aesthetic downside can only be calculated by making speculative estimates about their effects on property values. But the opponents of wind turbines on aesthetic grounds are surely thinking about more than property values. Given the myriad theoretical and methodological problems associated with measuring seemingly intangible costs and benefits, why would anyone think it worthwhile to try to impute a monetary value to unpaid labor in industrial societies? In a provocative book written some time ago, Marilyn Waring argues (1996) that such imputation matters because this labor is mostly carried out by women. By making such work visible, imputation has the potential to influence public policy and cultural values. Waring shows how the activities of typical women in two quite different societies are ignored by policy makers compiling measures of national productivity: Consider Tendai, a young girl in the Lowveld, in Zimbabwe. Her day starts as 4 a.m. when to fetch water, she carries a thirty litre tin to a borehole about eleven kilometres from her home. She walks barefoot and is home by 9 a.m. She eats a little and proceeds to fetch firewood until midday. She cleans the utensils from the family’s morning meal and sits preparing a lunch of sadza for the family. After lunch and the cleaning of the dishes, she wanders in the hot sun until early evening, fetching wild vegetables for supper before making the evening trip for water. Her day ends at 9 p.m., after she has prepared supper and put her younger

56 Anthropology, Economics, and Choice

brothers and sisters to sleep. Tendai is considered unproductive and economically inactive. According to the international economic system, Tendai does not work and is not part of the labour force. Cathy, a young, middle-class North American housewife, spends her days preparing food, setting the table, washing dishes, dressing her children, disciplining children, taking the children to day-care or to school, disposing of the garbage, dusting, gathering clothes for washing, doing the laundry, going to the gas station and the supermarket, repairing household items, keeping an eye on or playing with the children, making beds, paying bills, caring for pets and plants, putting away toys, books, and clothes, sewing or mending or knitting, talking with door-to-door salespeople, answering the telephone, vacuuming, sweeping and washing floors, cutting the grass, weeding, and shovelling snow, cleaning the bathroom and the kitchen, and putting her children to bed. Cathy has to face the fact that she fills her time in a totally unproductive manner. She, too, is economically inactive and economists record her as unoccupied. (13)

The reason these domestic activities do not show up in national accounts, Waring says, has to do with economists’ culturally based notion that “work” should be part of measures of productivity only if it is paid. By making unpaid labor invisible, economists are “devaluing” (in the sense of insufficiently esteeming) such work. Perhaps more importantly, the lack of imputation results in policies that fail to recognize the significance of unpaid labor. Although Waring discusses such policies in “developing countries,” she says less about what they might be in industrial societies. I would imagine that they would include—to give only a few possible examples—divorce settlements, social security (given only to those with paid work), and insurance compensation. Waring lists several methods for imputing the value of unpaid work. These methods, mostly derived from market wages, resemble those previously discussed: 1. the wages of substitute workers who could be hired for performing productive activities in the household (substitute workers could be either polyvalent: domestic servants, housekeepers, family aids [sic]; or specialised: cooks, launderers, babysitters); 2. the wages of workers performing tasks similar to those required by household tasks;

Choices between Paid and Unpaid Work 57

3. the wages of workers performing tasks requiring qualifications similar to those required by household tasks; 4. wages foregone in the market by those engaged in unpaid household work (that is, “the opportunity cost” of their time); 5. the average wages of market workers (sometimes differentiated by sex, age, education, residential area, legal minimum wages, and so on). (227)

She notes that the computation of the percentage of the contribution of women to measures of national productivity differ depending on which method is used. According to Waring, opportunity cost calculations (“average hourly wage, by sex, of full-time civilian workers, less income tax”) gives a higher percentage than the replacement value alternative (“paying another person to carry out all the housekeeping or homemaking or mothering tasks”). The reason for the difference is that the tasks of the replacement workers are poorly paid. I do not understand how Waring can assume, using the opportunity cost method, that women forgoing housework will (as a group) be able to get work that is paid an average wage. In any case, her results contrast with Delang’s study of the value of wild plant collection in a Thai village, where the replacement cost method yields a much higher estimate than the opportunity cost method. By imputing a value to unpaid labor, Waring is doing exactly what Chayanov warned against. The reason many households in industrial societies are economically viable is precisely that much labor is unpaid; imputing a monetary value to such work therefore seems misleading. Nonetheless, Waring would agree with the use of Chayanov’s term “self-exploitation” to describe unpaid household labor.

Conclusions Imputing a monetary value to subsistence production and unpaid labor is reasonable in societies with cash economies. Many important choices in such economies require decision makers to weigh the costs and benefits of paid and unremunerated work. When these decision makers opt for unpaid work, they are implicitly valuing the net benefits more highly than alternative paid labor. Because a monetary value can be assigned to paid work in these circumstances, it is at least plausible that one might be able to impute a higher value to unpaid options that are deemed preferable.

58 Anthropology, Economics, and Choice

Furthermore, ignoring unpaid work and subsistence production in national economic statistics can lead to misleading measures of both the total production of an economy and the relative contributions of different segments of the population. What does it mean, for example, when the World Bank reports that in 2007 the gross domestic production per capita in Malawi was US$266? If this figure omits subsistence production and unpaid labor (as it almost certainly does), an erroneous impression is given of both the scale of the country’s economy and the importance of the activities of members of many rural households. Only two methods are commonly used to impute monetary values to subsistence production and unpaid labor. The first is to calculate an opportunity cost—the monetary returns available from alternative paid work. The second is to calculate a replacement value—the cost of hiring someone to carry out a particular subsistence activity. As we have seen, there can be daunting technical difficulties associated with such calculations. Moreover, the results of the two methods in particular situations often lead to strikingly different imputations of value. A more fundamental problem arises when the imputation of monetary values obscures the economic significance of whether an activity involves the exchange of cash. Consider, for example, two farms using different types of labor to raise a cash crop. Farm A uses only unpaid household labor; farm B employs hired workers. As we have seen, Chayanov noted many years ago that in some circumstances farm A can operate at a profit while farm B is losing money. I am not at all sure what it means in such cases to impute a monetary value to the household labor on farm A. Some years ago I read a report for the Agency for International Development in which an economist used the opportunity cost method in cost-benefit analyses of corn farming in Belize. Using reasoning similar to that of the economists criticized by Chayanov, he concluded that many households raising corn using unpaid labor were operating at a loss even though their income from this crop was greater than what they spent on inputs such as weed killer, rent, and fertilizer. Although this conclusion struck me as absurd, the alternative of ignoring the economic value of unpaid work also seemed unacceptable. Perhaps the best reason to impute a monetary value to subsistence production and unpaid labor is to draw attention to the economic importance of these activities. Certainly, this can be done rhetorically by stressing how essential domestic work, volunteer service, and subsistence crops are to the well-being of all societies. But we live in a world where statistics—

Choices between Paid and Unpaid Work 59

however imperfect they may be—greatly influence states’ allocations of diverse resources. Furthermore, economic models of household economic behavior are often used in proposals by government entities and NGOs to improve the conditions of poor people in both rural and urban areas. Such models, statistics, and analyses can only be improved by a formal consideration of the worth of unremunerated work.

CHAPTER 3

Risk, Uncertainty, and Decision Making

How am I supposed to think about consequences before they happen? New Yorker cartoon caption of a chilD arGuinG With upSet parentS, november 9, 2009, p. 60

When I bought a new home in early 2009, I was faced with several complicated financial decisions. How much of my savings should I use for a down payment and how much should I borrow via a mortgage? Should I take out a bridge loan during the period between the closing date on the new house and the sale of my current residence? Which assets should I sell to cover the down payment? The world economy at the time was in the midst of a severe recession. The value of my stocks and bonds had dropped precipitously over the previous year; I had no idea if this plunge would continue. In order to learn how the tax implications of the market slump might influence my decisions about paying for the house, I made an appointment with my accountant. He recommended that we also consult one of his associates who handled investors’ portfolios. Although the meeting began with a discussion of taxes, we spent most of our time considering my investment strategies and plans for retirement. The financial adviser confidently showed me a complicated chart he had devised that showed the risks and returns associated with different types of investments over the past thirty years. His idea was that this chart could be used in conjunction with my financial goals to determine how best to pay for my house and make future investments. My immediate response was to ask the adviser about his apparent assumption that the future would resemble the past. I pointed out that prominent economists seemed to have no idea about what would be

Risk, Uncertainty, and Decision Making 61

the outcome of the current financial crisis. Why, then, would I want to base my investments on what had happened in previous decades? The adviser could not give a reassuring response to my query and could only say that economic conditions at this time seemed unusually uncertain. The type of question I raised was not at all unusual. Almost every decision that human beings make is affected by the unpredictability of future events. In the early twentieth century, the economist Frank Knight argued (1921) that such decision-making situations could be characterized as either “risky” or “uncertain.” In risky situations, decision makers can estimate the probabilities of different outcomes resulting from particular decisions. In uncertain situations, decision makers have no idea what the probabilities are of different outcomes. In the real world, however, most decisions-making situations combine elements of both risk and uncertainty. Decision makers often have vague ideas about what the odds are of different outcomes but are unable to make even ballpark estimates of probabilities. My house financing choices, for example, seemed to fall in the middle of a risk-uncertainty continuum. In this chapter I explore approaches that economists, cognitive psychologists, and anthropologists have taken toward analyzing decision making in situations of risk and uncertainty. Perhaps because of conceptual problems, economists and psychologists usually either ignore or downplay Knightean uncertainty when examining decision making. Anthropologists have taken diverse approaches to risk and uncertainty. Some ethnographers have attempted to employ the theoretical models of economists and the findings of laboratory experiments of psychologists in their analyses of decision making in real-world settings. Most anthropologists, however, reject the approaches of economists and psychologists as being unrealistic. Even in situations of low uncertainty when risks can be quantified to some extent, the models of economists and psychologists seem to these anthropologists to be largely irrelevant to the activities of most people making complicated decisions. Before discussing more thoroughly different theoretical approaches to risk and uncertainty, I present in some detail three examples of decision making drawn from my fieldwork. Although these case studies omit or gloss over many of the complex factors influencing choices, they provide sufficient ethnographic material with which to assess the advantages and shortcomings of varied theoretical approaches to risk and uncertainty.

62 Anthropology, Economics, and Choice

No-Till Farming in Iowa, 1982 During the early 1980s I conducted research on decisions about the adoption of no-till farming techniques in Van Buren County in southeastern Iowa (Chibnik 1987). No-till is an important soil conservation measure that had been recently introduced to the Midwest. Farmers using this technique in Iowa plant corn or soybeans in previously unprepared soil by “opening a narrow slot, trench or band of sufficient width and depth to obtain proper seed coverage” (Choi and Coughenour 1979:1). The use of herbicides enables farmers to control unwanted weeds and grasses without turning the soil. At the time of my research, no-till was receiving extensive publicity in Iowa. The Des Moines Register, the state’s leading newspaper, and magazines including Wallaces Farmer and Successful Farming ran articles about the technique. No-till yield contests were regular features of state fairs. This publicity led farmers to be increasingly willing to experiment with no-till. Van Buren County was one of the most active no-till areas in the state. Although there were no more than a dozen no-tillers in Van Buren County in 1978, by 1982 the local Soil Conservation Service estimated that 20 to 30 percent of the county’s 1,100 farmers had some acreage in no-till. Iowa farmers were experimenting with no-till because of its potential as an effective soil-saving technique that might also reduce labor time and fuel costs. In the 1980s, severe soil erosion problems were threatening the prominence of midwestern corn and soybean farming. Many farms had lost half their original topsoil. Despite deteriorating soils, farmers had been able to increase corn and soybean yields by applying massive amounts of fertilizers, herbicides, and pesticides. These energy-intensive practices, however, had become less cost- effective as the price of fossil fuels soared and gains in productivity leveled off. Furthermore, the extensive use of chemical fertilizers had caused soil damage. Besides these on-farm costs, soil erosion was causing off-farm environmental problems such as the silting of reservoirs and streambeds and the deterioration of water quality. Technological and macroeconomic changes had caused some farmers to become dissatisfied with older soil conservation practices. Large tractors could not be operated easily on fields with intricate systems of contours, terraces, and shelter belts. Expanding markets and soaring land values (with consequent higher rents and taxes) led farmers to shorten crop rotations and plow under pastures and woodlands. The search for new conservation methods led many farmers to attempt

Risk, Uncertainty, and Decision Making 63

to control erosion with reduced tillage systems. These systems minimally disturb the soil during seedbed preparation and subsequent weed control. Although some writers several decades earlier (such as Faulkner 1943) had recommended that farmers reduce or abandon plowing to preserve soil, such a change was not practical at the time because of the absence of effective nonmechanical methods of weed control. Chemical research after the end of World War II, however, enabled the development of better herbicides and pesticides. These new chemical compounds better distinguished between crops and weeds, and they provided greater effects per gram of material applied. No-till, the most radical of the various reduced tillage systems, essentially consists of the substitution of chemicals for plows. Because no-till eliminates plowing and cultivation, farmers switching partly or entirely to this technique can save considerably on labor. Such time saving was particularly useful for part-time farmers and dairy and feedlot operators whose fieldwork was a sideline for their main enterprise of raising animals. The equipment required for no-till differs somewhat from that used in conventional midwestern agriculture. Special planters or attachments to conventional planters are needed to penetrate barriers of dead organic material and place seed in the soil. The absence of plows, disks, harrows, cultivators, and other heavy equipment used in turning the soil allows no-till farmers to use considerably smaller tractors than those employed in conventional tillage. Farmers adopting no-till are able to reduce fuel expenses considerably because they spend less time with machinery in the field. If no-till is adopted entirely, there is a fuel saving through the use of smaller tractors. However, overall, no-till may not save money. Most farmers combine notill with plow agriculture; as a result their expenditures on machinery are higher than if they used only one of these techniques. Energy savings from fuel are usually counterbalanced by increases in the cost of purchasing and applying chemical herbicides and pesticides. Because no-till soybeans ordinarily require a greater application of weed killers than no-till corn, the cost per acre of no-till soybeans is usually somewhat higher than when that crop is conventionally grown. No-till farming has several disadvantages. The extensive use of herbicides may endanger the health of farmers and result in chemical pollution of the water table, rivers, and streams. Chemicals may change the physiology and development of crops so they become more susceptible to pathogens present in low levels in the field. Weeds resistant to the chemicals in herbicides have evolved. Besides these environmental risks, no-till can be difficult logistically. In comparison to conventional farm-

64 Anthropology, Economics, and Choice

ing, no-till requires a greater knowledge of soil characteristics and the proper application of chemicals. Furthermore, the reduction in the number of passes over a field gives no-till farmers few opportunities to correct mistakes. Early adopters of no-till confronted numerous risks and uncertainties. When farmers plant a crop using a time-tested technique, their situation with respect to yields usually falls near the risk end of a risk-uncertainty continuum. Although such farmers may not know how well their crops will do, they have some notion of the probabilities of good, fair, and poor yields in any particular time and place. When Van Buren County farmers first tried no-till, however, their situation was closer to the uncertainty end of the continuum. They soon discovered that no-till yields were more variable than those of conventional crops. Farmers often found that incomplete weed killing resulted in low yields of no-till corn and, especially, soybeans. Weeds were particularly problematic when farmers or hired sprayers had an inadequate knowledge of proper methods of chemical application. Weeds also turned out to be troublesome under certain climatic conditions. During cold springs, notill crops emerged later than those planted conventionally and were not as able to compete successfully with fast-growing weeds. Because activation of most herbicides requires water, in dry years no-till farmers ran risks of incomplete weed killing and of herbicide residues. Farmers learned from soil conservation officials, agricultural magazines, and their neighbors’ and their own experiences that no-till was especially risky for certain crops, soil types, and topographies. No-till works better for corn than soybeans because of the greater amount of weed control required for the latter crop. Many tillable acres in Iowa (43 percent, according to Cosper 1983) turned out to have soils unsuited for no-till. Drainage problems led many Van Buren County farmers to avoid using no-till in flatlands susceptible to flooding. Although many Van Buren County farmers were concerned about the long-term dangers of soil erosion and the health risks of excessive use of herbicides and pesticides, their agricultural decisions were primarily motivated by monetary considerations. Yield variation was the most important risk affecting the relative profitability of no-till and conventionally grown crops. Not entirely predictable differences in input costs were also relevant. No-till farmers were particularly affected by changing prices of herbicides and pesticides; variations in fuel prices were more important for conventional cultivators.

Risk, Uncertainty, and Decision Making 65

Agricultural Credit in the Peruvian Amazon, 1986 In the mid-1980s I conducted more than a year of research among people living in floodplain communities along the Amazon River near the Peruvian city of Iquitos (Chibnik 1994). Slow boats are the only means of transportation between these communities and Iquitos. Most villagers are descendants of local indigenous peoples and immigrants who came into the area during the rubber boom between 1870 and 1910. These ribereños are mostly monolingual in Spanish and do not self-identify as Indians. The overall standard of living was low in the communities where I did fieldwork. More than one in every ten children died before the age of two. The mean annual household income in three villages where I conducted surveys was about 7,000 intis (US$500). Ribereño livelihoods depended primarily on farming and fishing. Their principal foods were manioc ( yuca) and plantains. Rice was the most important cash crop. There was a sizable demand for rice among the residents of Iquitos, where a government buying center provided a guaranteed market and price. One of my projects involved looking at how ribereños made decisions about whether to get state-sponsored agricultural credit for rice production (Chibnik 1990, 1994:173–191). Most rice was grown on small plots in mud deposits that appear annually when the Amazon River falls. Because these deposits follow the ever-changing course of the river, they are impermanent. Every year new ones form and land that was once mud flats remains underwater. Although there were varied expenses associated with rice cultivation, the most important cost (about half of total monetary outlays) was for harvesting labor. Ribereños grew yuca and plantains using household labor and cooperative work groups (mingas). These crops are grown yearround on land types that flood either occasionally or not at all. With little monthly variation in rainfall and temperature in this part of Amazonia, labor demands for harvesting for yuca and plantains were evenly distributed over the course of a year and never high. Rice harvesting, in contrast, had to be done in a short window of time before the river covered mud flats. On plots of more than half a hectare, household labor was insufficient for harvesting. Cooperative work groups were difficult to mobilize because numerous households were harvesting simultaneously. Therefore, harvesting large crops of rice required hired labor. For many years, rice farmers seeking loans had to rely on rural money lenders (habilitadores) who charged high rates of interest. In the 1970s the

66 Anthropology, Economics, and Choice

state began making rice loans available through the Agrarian Bank.1 The process of getting a loan was time-consuming and costly. Farmers had to obtain separate loan approvals for specific activities such as land clearing, weeding, and harvesting. Borrowers had to travel to branches of the Agrarian Bank in Iquitos or elsewhere to get an entire season’s loan initially approved and to receive approvals for each loan portion. Because most villages were from five to fifteen hours by boat from the nearest branch of the Agrarian Bank, trips to get loan approvals involved a stay of at least one night away from home. Limited hours (nine in the morning to noon, five days a week), crowded conditions, and slow service at the bank often resulted in delays of several days. Farmers ordinarily did not know what the bank’s credit policies would be in the coming year. The bank made little effort to disseminate information about these policies, which often changed. Ribereños received most of their information about bank policies from a regional rice producers organization. The fundamental risk that rice producers encountered when they took agricultural credit was that their net incomes from sales would be less than the cost of their loans (principal plus interest plus travel costs). Because prices were guaranteed, profits depended mostly on rice yields. There were two principal reasons rice yields were sometimes unacceptably low. In some years farmers were unable to complete harvesting before the river rose. In other years farmers were unable to pay off loans taken out before the agricultural cycle began when they were unable to gain access to sufficient good land. This occurred most often when many mud deposits were not inundated after the river rose, leaving much of the land weed-covered. Farmers unable to repay loans found themselves in a precarious economic position. Although amnesties were sometimes given after regional disasters, indebted farmers were ordinarily forbidden to take out new loans from the Agrarian Bank. They were forced to borrow from habilitadores. Furthermore, when heavily indebted farmers sold rice to state buying centers, they did not receive cash. Instead, their debts were reduced by the value of the rice sold. Habilitadores sometimes took advantage of this government policy by buying rice at low prices from indebted farmers needing money immediately and reselling it to buying centers at higher prices. A research assistant and I interviewed twelve farmers about their experiences with rice loans from the Agrarian Bank. These farmers had

Risk, Uncertainty, and Decision Making 67

taken out a total of ninety-one such loans between 1970 and 1985. Of these loans, sixteen (18 percent) were not paid back when due, and in all but three of these cases insufficient rice production was the reason for the loan default. While these data do not allow more than a tentative estimate of the probability of a loan default because of insufficient production, they suggest that the chance of this occurring was considerable. Even if in most years the probability of a loan default was low, simple mathematical calculations show that a regular borrower was likely to default eventually.2 All but one of the twelve rice growers interviewed had defaulted at least once. Furthermore, even those borrowers able to repay loans often lost money or made minuscule profits. Because the chances of poor rice yields in mud flats can be roughly estimated, obtaining agricultural credit might have seemed to fall on the risk side of a risk-uncertainty continuum. Taking out rice loans appeared to be a calculable risky strategy that most farmers would be likely to avoid. Such an analysis, however, ignores a major motivation for seeking rice loans. Farmers often borrowed in order to insure against economic or medical setbacks or to even out seasonal fluctuations in cash. Ribereños sometimes deliberately took out loans to pay for routine household maintenance; at other times they were forced by emergencies to divert money away from producing and marketing rice. If loans were repaid on time, the Agrarian Bank might not monitor their use. Even when farmers used so much credit for household expenses that loan repayment was impossible, indebtedness to a bank was sometimes regarded as preferable to borrowing from an habilitador. The conventional wisdom about “peasant risk-aversion” (such as Ortiz 1973, Scott 1976) was therefore of only limited use in understanding ribereños’ willingness to seek rice loans from the Agrarian Bank. Farmers in the floodplain communities wanted to make money, avoid long-term indebtedness, and ensure a steady food supply and medical care for their families. Rice loans increased the probability of both significant monetary profits and long-term indebtedness. Credit provided insurance against starvation and medical disasters. Obtaining rice credit thus increased certain types of risk and decreased other types of risk. Farmers differed in the extent to which they weighed these varying types of risk (and uncertainty) when making decisions about whether to take out loans from the Agrarian Bank.

68 Anthropology, Economics, and Choice

Oaxacan Wood Carvers, 1998 Since 1994 I have been conducting research on the sale of wood carvings (alebrijes) from the Mexican state of Oaxaca in the global folk art market (Chibnik 2003). Even the most casual tourist in the city of Oaxaca, the state capital, can see that craft production is an integral part of the local economy. Visitors to the historic center pass shop after shop featuring the regions’ artisans. Tours visit villages where pottery, weavings, and textiles are made. Vendors in squares and markets hawk handmade rugs, earthenware, tin ornaments, wall hangings, and cloth and leather belts. Few tourists realize that their purchases are only a small part of the local trade in arts and crafts. The livelihoods of most rural potters, weavers, and wood carvers depend primarily on sales to intermediaries from Mexico, Europe, Japan, Canada, and, most importantly, the United States. My research has focused on the activities of carvers in two communities, Arrazola and San Martín Tilcajete. In the 1970s and early 1980s these artisans sold their pieces mostly to store owners in the city of Oaxaca. Wood carving during this period was ordinarily a part-time occupation for a few adult males; women and children occasionally helped with painting and sanding. In the mid-1980s, wholesalers and store owners from the United States began to visit Arrazola and San Martín to buy carvings. Wholesalers could earn significant amounts of money by selling carvings in the United States at three to six times their cost in Mexico. Carvings became more complicated and paint jobs more ornate as artisans competed to show their skills. The craft since then has been a family activity in which the typical, although not universal, division of labor is women painting and men carving and also doing some painting. Children often help out with various tasks, most importantly sanding. Adult men do less than half the work on many carvings. The wood-carving boom of the 1980s and 1990s had dramatic effects on the economies of Arrazola and San Martín. In Arrazola some families abandoned agriculture altogether and worked as full-time carvers. Although most carving households in San Martín continued to farm, agriculture became a secondary, subsistence- oriented activity. Despite artisans’ success selling wood carvings during these years, in both communities significant numbers of people migrated temporarily or permanently to the United States. At the beginning of this century, sales of cheap wood carvings dropped dramatically, and many low- end artisans stopped making pieces. There remained a good market for more expensive alebrijes.

Risk, Uncertainty, and Decision Making 69

My most intensive fieldwork took place in the late 1990s, when the market for all Oaxacan wood carvings was still strong. An important part of my research was determining how and why people in households involved in the trade in alebrijes allocated their time among wood carving, wage labor, schooling, and migration to the United States and elsewhere. To a certain extent, these decisions were made by individuals weighing the costs and benefits of different options. But these choices were also in part family decisions influenced strongly by discussions among adult members of households. The economic goals of Oaxacan wood-carving households were more ambitious than those of the Peruvian ribereños with whom I had worked in the 1980s. Ribereños were primarily motivated by their desires to ensure a steady food supply for their families and to have money in case of medical emergencies. Oaxacan wood carvers, in contrast, aspired— usually in vain—to a middle-class standard of living including automobiles, diverse electronic devices, and good educations for their children. They compared their economic situation not only with other Oaxacans but also with their relatives and friends who had migrated (usually without legal documentation) temporarily or permanently to the United States. With the exception of migration to the United States, wood carving was clearly the most promising short-term economic option in the late 1990s for families in Arrazola and San Martín. The returns per labor day were markedly higher than those from wage labor and crops such as corn, beans, garlic, and onions. The demand for alebrijes was sufficiently high that wood carving was not particularly risky if families had artisans of average talent. Even the cheapest and simplest pieces could be sold to visiting tourists, store owners in Oaxaca, and folk art dealers from the United States. Children, the elderly, and mothers of young children who lacked either the time or the ability to work at full-time wage labor could help out with wood carving. Why, then, were there many people in Arrazola and San Martín who did not participate in the wood- carving boom? Why did many woodcarving families continue to devote considerable time to growing corn and beans? Some people either did not enjoy making alebrijes or lacked talent for carving and painting. Long-term considerations of risk and uncertainty, however, were the primary reasons families often encouraged certain members to migrate, stay in school, or pursue economic activities other than wood carving. Oaxacan artisans in the late 1990s thought (correctly) that the market for alebrijes would worsen fairly soon. If everyone in a family specialized in painting and carving alebrijes, their tal-

70 Anthropology, Economics, and Choice

ents would all be economically useless in a few years. Furthermore, only the most successful wood-carving families had a standard of living that approached middle-class. Families with members of diverse skills might therefore do better in the near future than those where everyone worked on wood carvings. Although farmers suffered from frequent droughts and were able to earn only small amounts of money even in years when the weather was good, many rural Oaxacan families were even in the short term reluctant to abandon the limited food security made possible though growing corn and beans. Farmers could ordinarily retain access to their fields (available via complex rules of communal land tenure) only if they continued to cultivate them. Such access was a form of subsistence insurance if woodcarving sales fell. The two principal alternatives to wood carving that young men and women considered were obtaining secondary (in some cases also university) education and migration. The potential upside to both these choices was greater than that for wood carving. Education could lead to comfortable jobs in government or business. Migration offered possibilities of earning salaries that were orders of magnitude greater than those in Oaxaca. Many families in Arrazola and San Martín received some money from relatives in the United States. Rural Oaxacans knew, however, that investments in education and migration did not always pay off for either participants or other family members. Time spent away from a community could result in a loss of labor on wood carving. This was ordinarily not a serious problem; there were usually family members who could paint or carve in place of those unavailable because of studies or migration. More importantly, neither education nor migration was a sure path to the middle class. Mexico has many educated people working at menial jobs. This is especially noticeable in Oaxaca, one of the poorest states in the country. Migrants to the United States who were able to cross the border (usually not a problem) sometimes could get only intermittent, poorly paid employment. Even when migrants obtained steady work (most I met were able to do this), there was great variation in how much money they sent home. The most risk-averse household economic strategy in both the short term and the long term therefore involved a combination of wood carving, agriculture, wage work, education, and migration. Wood carving, wage work, and agriculture provided food and income in the short term. The other activities had the potential for greater incomes in either the near future (migration) or the long run (education).

Risk, Uncertainty, and Decision Making 71

Common Themes Iowa farmers experimenting with no-till, Peruvian ribereños taking out loans from the Agrarian Bank, and rural Oaxacans allocating household labor all had to cope with the unpredictable consequences of their choices. Although the ethnographic details of these examples vary, they share certain features that illustrate why simple models often fail to capture the complexity of decision making in situations of risk and uncertainty. In each of these cases decision makers must consider diverse risks, make choices that differ in their positions on a risk-uncertainty continuum, and take into account both short-term and long-term implications of their actions. Furthermore, the potential costs and benefits of alternative choices cannot always be easily measured. There are two important similarities in the ways in which the decision makers described here (and many others in different places and times) respond to situations of risk and uncertainty. Iowa farmers, ribereños, and Oaxacan artisans diversified their activities in order to reduce the chances of disastrous consequences resulting from the economic choices. They all also attempted to convert uncertainty into risk through information gathering and small-scale experimentation. Multiple Risks and Uncertainties Most choices involve multiple risks and uncertainties. Iowa farmers choosing between no-till and conventional methods had to consider yield variations, potential input costs, and the chances of health problems resulting from the use of chemicals. Ribereños weighing the possibility of taking out a loan were influenced by flood patterns, inflation, and the chances of family members incurring medical expenses. Oaxacan artisan households considering different types of work had to contend with an unpredictable market for crafts, the perils of being undocumented laborers in the United States, and a volatile Mexican economy that limits opportunities for educated people. Individuals differ in the weights they place on the importance of the various risks and uncertainties associated with particular choices. Moreover, analyzing the risks and uncertainties of variables such as yields, costs, and prices is enormously complicated. The risk and uncertainty of corn yields in Iowa, for example, arises from yearly variations in temperature, rainfall, weeds, pests, and a host of other factors. Farmers’ ideas about yield risks cannot be captured merely by examining “objective” his-

72 Anthropology, Economics, and Choice

torical records of fluctuations in corn production. Instead, farmers’ subjective perceptions of risk and uncertainty are influenced by their ideas about the myriad underlying causes of such variations. Variability in the Extent to Which Risks Can Be Estimated The decisions presented in the case studies took place in contexts that were quite different from Knight’s situations of “pure risk” in which the probabilities of different outcomes are known. Instead, they all fell somewhere on a risk-uncertainty continuum. Some variables (corn yields with conventional farming in Iowa, flood patterns in the Peruvian Amazon) were closer to the risk end of the continuum because of historical records and the reasonableness of assumptions that the future would resemble the past. But even in these cases decision makers could only provide rough estimates of the chances of different outcomes. Many other factors affecting the decisions described here are closer to the uncertainty end of the continuum. These included variability of no-till soybean yields on flatlands in Iowa, the chances of a member of a ribereño family being bitten by a snake, and the odds that a large-scale buyer of folk art would like a new creation by a Oaxacan wood carver. Short- term and Long- term Considerations The potential risks, uncertainties, costs, and benefits of alternative choices in the short term may differ from those in the long term. This can be seen most clearly in the Oaxaca example, where short-term payoffs from wage labor, farming, and artisanry must be weighed against potential longer-term payoffs from migration and schooling. Although farming, wage work, and artisanry are clearly less risky in the short term than migration and education, this may well not be the case in the long run. Ribereño borrowers and Iowan farmers are also influenced by both short-term and long-term considerations of risk and uncertainty. Perhaps counterintuitively, taking out credit is probably safer in the short term for ribereños than concentrating only on crops that can be grown without loans. Borrowing ensures the ability to hire labor when needed, enables a greater diversity of economic activities, and provides money in case of medical emergencies. In the long run, borrowing may be the riskier strategy because of the likelihood of floods leading to defaults that place ribereños in untenable economic situations. Iowa farmers experimenting with no-till may initially have poor yields because of their unfamiliarity

Risk, Uncertainty, and Decision Making 73

with the technique. The soil conservation made possible by no-till, however, may make this farming method preferable to conventional agriculture in the long run for particular crops on certain types of land. Incommensurability of Costs and Benefits Estimating the risks and uncertainties of alternative choices entails calculating diverse potential costs and benefits. These costs and benefits can be difficult to measure. Even when measurement is relatively straightforward, the costs and benefits may not be easily comparable. Iowa farmers adopting no-till must think about the potential income from corn and soybeans grown using the new technique, the health risks resulting from greater exposure to chemicals, and the satisfaction gained from preserving soil for future generations. Ribereños taking out loans weigh the risk of defaulting, the chances of improving their standard of living, and the peace of mind provided by greater ability to cope with medical emergencies. Young Oaxacan men considering migration to the United States must consider the chances of finding jobs, the dangers of crossing the border, and the loneliness of being away from their families. Attempts to convert these diverse costs and benefits into a single measurable unit such as money, while tempting for their analytic convenience, may well be unrealistic. Strategies for Coping with Risk and Uncertainty Many social scientists (among them Netting 1993:33, Scott 1976:24) have observed that diversification of economic activities is a common, useful strategy for reducing risk and uncertainty. The idea is that people who engage in several activities will have something to fall back on if problems arise with one of their sources of food and income. This is codified in folk wisdom in the adage “Don’t put all your eggs in one basket.” The three cases presented here provide further evidence of the ubiquity of diversification to avoid risk. Many rural households in Iowa in 1982 split their land between no-till and conventional crops and also took offfarm work. Ribereño households borrowing money from the Agrarian Bank in 1986 to hire laborers to help with the rice harvest also fished, collected fruit, and grew corn and manioc using family labor. Oaxacan artisan households in 1998 raised subsistence crops, made wood carvings, worked as wage laborers, and migrated temporarily to the United States or other parts of Mexico.

74 Anthropology, Economics, and Choice

In an article I wrote some years ago (Chibnik 1981), I argued that decision makers often take actions to make costs more risk-like and less uncertain-like. These actions, I suggested, commonly take the form of low-cost (in time or money), small-scale experimentation and information gathering. Although I thought at the time that this was an original observation, I now know that a few economists had already reached this conclusion. K. E. Warner (1974), for example, had made essentially the same argument in the context of the adoption of innovations. He pointed out that when possible, potential adopters seek information about the cost and value of an innovation via their own and others’ experiments. (Marra, Pannell, and Ghadim 2003 and Rogers 2003 review more recent work on this topic.) Iowa farmers talked to their neighbors about no-till, ribereños attended meetings about credit organized by the leaders of rural unions, and Oaxacan artisans asked their relatives about the sales of new types of carvings. The examples presented here suggest that detailed ethnography is necessary to understand the complex considerations that influence decisions in situations of risk and uncertainty. In what follows, I show that the great majority of economists and psychologists ignore such complexity, instead creating radically simplified models of choice. Although most anthropologists take a nuanced ethnographic approach, some prefer the more mathematically tractable treatments of risk and uncertainty by economists, psychologists, and ecologists.

Disciplinary Approaches to Risk and Uncertainty Economists Economists have written extensively about the definition and measurement of “risk.” Some see risk primarily in terms of the variability of outcomes possible when a particular choice is made. A farmer planting corn, for example, knows that yields and prices vary from year to year. These economists define risk mathematically by empirical measures of variability using statistics such as the standard deviation. Other economists, such as Nassim Nicholas Taleb, place more emphasis in certain situations on the chances of a disastrous outcome.3 Poor corn farmers might be especially concerned about a harvest being so bad that they will be unable to feed their families. In such cases, a definition of risk might rest on the probability of a harvest being unacceptably low.

Risk, Uncertainty, and Decision Making 75

Some economists have used sophisticated models (Meyer and Meyer 2006) to describe and analyze situations in which decision makers are concerned with both variability and the chances of a disastrous outcome. They deploy complicated mathematical equations to describe riskaversion preferences of individual decision makers such as farmers or investors. Eliciting risk-aversion attitudes in a form amenable to mathematical modeling is hardly straightforward. Many models combine information from interviews and untested assumptions about “rational behavior.” Most economic models of decision making in risky and uncertain situations are based on expected utility theory. Such models assume that probabilities can be assigned to the likelihood of different outcomes resulting from a particular choice. Some analyses calculate these probabilities from “objective” data about what has happened in the past (such as crop yields over several years). Other analysts adopting a “Bayesian approach” base their models on “subjective” probabilities of different outcomes elicited by interviewing decision makers. Economists have often noticed (for example, Smith and Mandac 1995 on Philippine farmers’ estimates of future rice yields) that subjective and objective probabilities can differ considerably. They conclude that decision makers acting on subjective probabilities often choose suboptimal options, given their riskaversion preferences. Diverse methods have been used in attempts to determine risk-aversion preferences and subjective probabilities (Meyer and Meyer 2006; Marra, Pannell, and Ghadim 2003). A typical question about risk-aversion preferences might ask stock investors if they prefer either a certain return of 6 percent annually or a 90 percent chance of a 20 percent return and a 10 percent chance of a 50 percent loss. A typical question about subjective probabilities might ask Iowa farmers what they think the chances are of their corn yields being between 150 and 200 bushels per acre. The assumption is that sufficiently careful questioning along these lines will allow the construction of realistic models of decision making. A recent summary of the state of the art on risk aversion (Meyer and Meyer 2006) shows well how the work of economists—almost impenetrable to outsiders—has resulted in only the most meager conclusions. The authors, Donald Meyer and Jack Meyer, conclude after many pages of definitions, assumptions, and differential equations that any measure of risk aversion for a decision maker is specific to the particular outcome variable over which the measure is defined or estimated.

76 Anthropology, Economics, and Choice

Even small changes when defining or measuring an outcome variable can significantly alter the risk aversion function representing the decision maker’s risk preferences. All statements concerning the magnitude, slope, or any other property of a risk aversion measure must implicitly or explicitly identify the outcome variable to which the statement applies. (91)

Putting this in plain English is not simple. Readers must understand that the models under discussion all assume that decision makers are concerned with only one aspect (“the outcome variable”) of a decision. That is, they regard the multiple aspects of risk considered by decision makers such as Iowa no-till farmers, Peruvian borrowers from the Agrarian Bank, and Oaxacan wood carvers as too complex to be mathematically modeled. Furthermore, these models all assume that the decision-making situation falls at the absolute risk end of a risk-uncertainty continuum with known probabilities of different outcomes. Even with such unrealistic assumptions, Meyer and Meyer find that the precise definition of the “outcome variable” (examples given include wealth, income, and “rate of return”) greatly influences models’ predictions of the risk-taking propensities of decision makers such as investors or farmers. Meyer and Meyer reach only one substantive conclusion from their painstaking survey of the relevant literature on risk aversion. They find— given a host of assumptions and definitions—that wealthy people are, on the whole, more willing to take risks than less wealthy people. This unsurprising finding seems a puny payoff for economists’ many mathematical models and empirical studies. Economists analyzing decision making when outcomes are variable ordinarily ignore Knightean uncertainty (Taylor 2003:254). This neglect reflects the inadequacy of mathematical tools to model behavior in the face of uncertainty. Taylor observes that as we move from risk to uncertainty, it becomes “more and more difficult to conceptualize and define random variables,” adding that “even the traditional notion of a sharply defined random variable as a way of modeling an uncertain system is problematic” (253). Even the best decision makers apparently do not know how they should act in uncertain situations. During the worldwide economic crisis of 2008–2009, economists often noted the problems uncertainty caused for investors. The chief economist of the International Monetary Fund commented in The Economist in early 2009 that

Risk, Uncertainty, and Decision Making 77

[f ]rom the . . . index of stock market variability . . . to the dispersion of growth forecasts, even to the frequency of the word “uncertain” in the press, all the indicators of uncertainty are at or near all-time highs. What is at work is not only objective, but also subjective uncertainty, or what economists, following Chicago economist Frank Knight’s early twentieth century work, call “Knightean uncertainty.” Objective uncertainty [Here Blanchard is writing about what I am calling “risk.”] is about what Donald Rumsfeld (in a different context) referred to as the “known unknowns.” Subjective uncertainty is about the “unknown unknowns.” When, as today, the unknown unknowns dominate, and the economic environment is so complex as to appear nearly incomprehensible, the result is extreme prudence, if not outright paralysis. (Blanchard 2009)

Because their mathematical toolkits are unable to model, describe, or prescribe the behavior of decision makers in uncertain situations, economists unsurprisingly have written little about this topic. As Jens Beckert observes (1996:813), most economists reinterpret situations of uncertainty as situations of risk in the sense that the individual decision maker has the information to make probability calculations. A few economists such as C. Robert Taylor (2003) understand that many decision-making situations should be placed along a risk-uncertainty continuum. This recognition rarely leads to their employing realistic models in their analyses. The problems again are both mathematical and conceptual. Taylor aptly comments that “the topology between these two precisely defined extremes [pure risk and pure uncertainty] is blurred at best and filled with epistemological gaps at worst . . . [T]here is a spectrum of increasing fuzziness as we move toward pure uncertainty” (253). John Maynard Keynes is the most famous economist who has considered how decision makers act in situations when some uncertainty is present. The recent economic crisis led the New York Times Magazine to run an article in December 2008 on Keynes’s ideas about uncertainty in which Robert Skidelsky notes: Then, as now, economists believed that all uncertainty could be reduced to a measurable risk . . . By contrast, Keynes created an economics whose starting point was that not all future events could be reduced to a measurable risk. There was a residue of genuine uncertainty, and this made disaster an ever-present possibility, not a once-in-a-lifetime “shock.” In-

78 Anthropology, Economics, and Choice

vestment was more an act of faith than a scientific calculation of probabilities. (21)

In such situations, Keynes thought that investors relied on what he called “conventions.” These conventions include assumptions that the future will be like the past and that current prices correctly indicate “future prospects.” Another convention, perhaps not altogether consistent with the previous two, was “running with the crowd.” According to Keynes, in situations of uncertainty investors do not process information efficiently because they are unaware of which information is relevant. This leads to what economists later called the “herd mentality.” Cognitive Psychologists Starting in the late 1960s, cognitive psychologists have shown that (at least in laboratory experiments) human beings are not very good at making the probabilistic judgments required by expected utility theory. Daniel Kahneman shared the 2002 Nobel Prize in Economics for studies along these lines, many of which he carried out with his longtime collaborator Amos Tversky, who died in 1996. Kahneman and Tversky (1982a:3) showed that experimental subjects systematically err in making probability judgments because they engage in certain shortcuts (heuristics) to simplify decision making. Such heuristics bear some resemblance to Keynes’s conventions. Two of the most important heuristics are representativeness and availability. Decision makers using the representativeness heuristic assess the probability that an object is a member of a particular category by judging how similar the object is to what they perceive to be an ordinary member of that category. In so doing, they often pay too little attention to base rates, the relative sizes of different categories. In a typical experiment, Kahneman and Tversky (1982b:49) presented subjects with a list of nine fields of graduate study and information about the percentage of students in each field (the base rates). They then gave the subjects a personality sketch of “Tom” that fit stereotypical ideas at the time about computer scientists. Tom was said, for example, to have “a need for order and clarity” and “little feel and little sympathy” for other people. Subjects were then asked to estimate the probability that Tom was a student of computer science. Their estimates were based mostly on the personality sketch; the highly relevant base rate information was largely ignored. When the availability heuristic is used, decision makers judge probabilities by how readily they can think of examples fitting various possi-

Risk, Uncertainty, and Decision Making 79

bilities. The availability of such examples is affected by numerous factors, including personal experiences and media publicity. For instance, studies (such as Combs and Slovic 1979) have shown that people in the United States commonly make mistakes in estimating the relative proportions of accidental deaths by fire and drowning. Because fires are more spectacular (“vivid” in the terminology of cognitive psychology) and more likely to be given prominence in the media, people overestimate the proportion of deaths from fire and underestimate the proportion of deaths from drowning. Cognitive psychologists have found in laboratory experiments that the way in which probabilities in a situation are described (framed) influences decision making. In perhaps the most famous example, Kahneman and Tversky asked subjects the following question: Imagine that the U.S. is preparing for the outbreak of an Asian disease which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientific estimates [How can an estimate be “exact”?] of the consequences of the programs are as follows: If Program A is adopted, 200 people will be saved. If Program B is adopted, there is a one-third probability that 600 people will be saved and a two-thirds probability that no people will be saved. Which of the two programs would you favor? (1984:343)

Most subjects (72 percent) chose Program A. Kahneman and Tversky then presented identical options to the subjects but framed the probabilities differently: If Program C is adopted, 400 people will die. If Program D is adopted, there is a one-third probability that nobody will die and a two-thirds probability that 600 people will die.

When the options were framed in this way, Program C (identical to Program A) was chosen by only 22 percent of the subjects. Kahneman and Tversky developed “prospect theory” in an attempt to describe how human beings employ subjective probabilities in their decision making. Prospect theory is a modification of expected utility theory (Baron 2008:262 gives details). It maintains the basic idea of expected utility theory that choices are based on decision makers’ comparisons of the product (multiplication) of the subjective probability of particular outcomes and the utility of those outcomes. However, prospect theory

80 Anthropology, Economics, and Choice

adds that the subjective probabilities of different outcomes are distorted (from the “real” probabilities) because of framing issues and heuristics such as representativeness and availability. The underlying assumption of prospect theory is that with sufficient effort a researcher can elicit subjective probabilities. Perhaps for this reason, cognitive psychologists in recent years have mostly ignored the risk/ uncertainty distinction. Some years ago, however, Daniel Ellsberg (of Pentagon Papers fame) wrote about the effects on decision making of Knightean uncertainty, which he called “ambiguity.” According to Ellsberg (1961), decision makers attempt to avoid ambiguity and in so doing violate the axioms of expected utility theory. Because expected utility theory assumes situations of pure risk (where probabilities are known), I do not fully understand how expected utility axioms could ever apply to choices made under any degree of Knightean uncertainty. If such axioms are inapplicable, how can they be violated? The predominant view of contemporary cognitive psychologists on risk and uncertainty is summarized in Jonathan Baron’s influential textbook Thinking and Deciding (2008). Baron notes that the majority of psychological experiments assume situations of pure risk. Taking a subjective probability perspective, Baron argues: The personal theory of probability . . . implies that the idea of “unknown” probabilities makes little sense. Because probabilities are properties of the person, not the world, the only way in which a probability could be “unknown” is for the person not to have reflected enough about the situation. (282–283)

This seems like an unrealistic, peculiar view of how the world works. As I write, the world’s economy is in flux. Investors are having tremendous problems because of their uncertainties (in a Knightean sense) about all sorts of “outcomes.” These uncertainties cannot be resolved simply by further “reflection.” Nonetheless, Baron comes to a conclusion similar to that I and others have reached about how people react in situations of ambiguity/uncertainty. He agrees with Ellsberg that decision makers will often delay taking actions in such cases. Instead, they will seek information to make situations less ambiguous (284). Perhaps such information seeking is what Baron means by “reflection.”

Risk, Uncertainty, and Decision Making 81

Anthropologists Some quantitatively oriented anthropologists have attempted to determine the extent to which risk-related models of behavior from other disciplines help explain data collected in ethnographic and archaeological fieldwork. Economic anthropologists have emphasized shortcomings of the expected utility model in describing how farmers make choices when faced with Knightean uncertainty. Ecological anthropologists and archaeologists have argued that models from biology are useful in understanding the adaptive benefits of varied human actions with respect to risk. Anthropologists emphasizing “scientific” approaches to the analysis of human behavior have used the methods of experimental economics to test risk models in ethnographic settings. In contrast with those researchers who begin their analyses by looking at ideas from other disciplines, other anthropologists looking at decision making in risky and uncertain situations reject altogether models from economics and biology. Some agree with cognitive psychologists that such models do not provide realistic guides to how people actually make decisions. Their fieldwork emphasizes eliciting the cultural rules (heuristics) that decision makers employ in their efforts to cope with risk and uncertainty. Other anthropologists taking a “cultural” approach toward risk and uncertainty disdain the ideas of both mainstream economists and cognitive psychologists. These social constructivists say that cultures differ in the amount of attention they pay to each of the multiple risks and uncertainties associated with particular decision-making situations. Furthermore, within societies there is no consensus about the potential downsides of different choices. Perhaps the majority of anthropologists analyzing decision making take an ethnographic approach that describes the activities that individuals and groups take to avoid unacceptable risks. Although their descriptions may be consistent with the theoretical ideas of economics, behavioral ecologists, and cognitive psychologists, these empirically oriented anthropologists do not attempt to create and test models of decision making.

anthropoloGiStS examininG riSk moDelS from other DiSciplineS In the 1970s and 1980s Sutti Ortiz and Frank Cancian were the most prominent anthropologists examining economic models of the influence of risk on agricultural decision making. In Uncertainties in Peasant Farming

82 Anthropology, Economics, and Choice

(1973), Ortiz clearly states why indigenous (Paez) farmers in Colombia making varied choices could not take risk into consideration in the ways that conventional economic theory predicts: Not only do weather conditions vary to an extent that makes it impossible to determine the chances of crop failure, but farmers lack the technical knowledge to enable them to evaluate the chances with sufficient accuracy. . . . [F]or decisions to be made with certainty [Ortiz means “risk” here], four conditions must be met: 1, that frequency ratios must be obtained from numerous and uniform sets of performance; 2, that performances can be repeated; 3, that experiments from which the frequency ratios are derived do not destroy the circumstances in which they were performed, hence that it does not become a unique act, and that the system remains stable; 4, that we consider only the total result of a large number of trials and not each trial separately. A Paez farmer cannot determine frequency ratios. His economic environment is changing; his assets are so limited that he can invest them in a venture once and never again if that venture fails; his technological knowledge of production limits the number of possible decisions—for example, coffee is planted once in a lifetime when he is a young and an inexperienced farmer. (13) 4

Ortiz later emphasized that Paez farmers did not remember the information, even when it was available, needed to make the calculations required by expected utility theory. She notes, for instance, that “as farmers do not memorize price sequences, they are only able to talk about the most familiar outcomes, note the general trend in prices, and give a few examples of past prices” (1980:187). While Ortiz concludes that Paez farmers are unable to arrive at probabilistic estimates or forecasts, she concedes that they see some outcomes of choices as more likely than others (188). Even though Ortiz’s goal is to show that farmers attempt to maximize “returns” (however measured) as economic theories about rationality predict, she says that they cannot do so in the ways stipulated by expected utility theory. I am unsure what Ortiz means by “rationality” in this context. Cancian most thoroughly presents his views about risk and uncertainty in The Innovator’s Situation: Upper-Middle-Class Conservatism in Agricultural Communities (1979). In this book Cancian compares the effects of wealth and “rank” (social status) on farmers’ willingness to innovate in diverse societies. According to Cancian, there are several reasons wealthier farmers are especially willing to innovate. They are more able than their

Risk, Uncertainty, and Decision Making 83

poorer neighbors to economically withstand the adverse consequences of an unsuccessful experiment. Furthermore, innovations ordinarily require substantial investments of money that wealthier farmers are more able to afford. Finally, because wealthy farmers are more likely to be educated, they are better able to gain access to and understand relevant information. Because the Knightean uncertainty associated with a lack of information inhibits innovations, poorer farmers lacking knowledge about the possible consequences of choices are less likely to be the first to experiment with new farming methods. Cancian also argues, however, that the association of wealth with high status inhibits innovation. People with high rank have less to gain and more to lose from a random change in their status. Cancian points out that the combined effects of these facilitating and inhibiting effects on innovations are not straightforward: [T]he overall theory makes no specific prediction about the dominance of the inhibiting effect or the facilitating effect. The only clear prediction is that the strength of the inhibiting effect should decrease as uncertainty decreases. That is, early in the innovation process, when uncertainty about the outcome of using a new practice is high, the inhibiting effect should be at its strongest. As more and more people adopt an innovation, information spreads, uncertainty is reduced, and the inhibiting effect should gradually disappear. . . . Since the inhibiting effect and the facilitating effect “contradict” each other, the disappearance of the inhibiting effect means the emergence of the facilitating effect in the empirical curve. (19)

He concludes through an examination of empirical evidence from diverse places that the relationship between wealth and willingness to innovate is nonlinear. As would be predicted by purely economic models, the wealthiest farmers are overall the most likely to innovate and the poorest farmers are the least likely. However, inhibiting sociological considerations of rank lead to upper-middle-class farmers being less willing to innovate than lower-middle-class farmers. Unlike Ortiz and Cancian, a number of anthropologists writing about risk and uncertainty in recent years have largely accepted the assumptions of formal models from other disciplines. Some ecological anthropologists interested in subsistence decisions in the past and present have placed particular emphasis on models drawn from nonhuman behavioral ecology. These evolutionary models assume that organisms adapt to risky environ-

84 Anthropology, Economics, and Choice

ments in ways that will maximize their chances of passing their genes on to future generations. In an exhaustive review, Bruce Winterhalder, Flora Liu, and Bram Tucker (1999) examine numerous mathematical models related to risk in behavioral ecology. Many of these models are concerned with optimal foraging strategies. They conclude that only a few anthropological studies of subsistence have made rigorous use of such models. The authors describe four such studies that “exemplify a full risk-sensitive argument” (332). These archaeological and ethnographic studies examine why certain mixes of sheep and goats are used in an environment characterized by unpredictable droughts (Mace and Houston 1989), explore the rationale behind the use of multiple, dispersed plots by farmers in the Peruvian Andes (Goland 1993), simulate prehistoric Hopi exchange under three scenarios for interhousehold sharing (Hegmon 1989), and look at the conditions in which cooperative behavior was useful among the Northern Anasazi of Mesa Verde between 900 and 1300 (Kohler and Van West 1996). In the first part of this century some anthropologists (including Henrich and McElreath 2002, Kuznar 2001) have used methods taken from experimental economics in efforts to determine the subjective probabilities of different outcomes among farmers and pastoralists making decisions in nonwestern settings. These experiments involve attempts to determine propensities for risk taking by asking people to make choices— sometimes hypothetical, sometimes involving money—between sure returns and various lottery options (for example, 50 percent chance of return X and 50 percent chance of return Y). They compare the results of these experiments to expected utility theory and Cancian’s ideas about upper-middle-class conservatism. The obvious question is the extent to which choices made in such circumstances mirror real-life behavior.

anthropoloGiStS rejectinG riSk moDelS from other DiSciplineS Several decades ago the anthropologist Naomi Quinn sharply criticized the expected utility approach to the analysis of decision making. In an article with the straightforward title “Do Mfantse Fish Sellers Estimate Probabilities in their Heads?” (1978), Quinn explicitly rejects “the widespread assumption by economists that individual decision makers can and do construct probability distributions against which they assess the riskiness of uncertain decisions” (207–208). From interviews with women in Ghana who made decisions about when and where to market fish, Quinn concludes that their decisions were based on simple rules of thumb based

Risk, Uncertainty, and Decision Making 85

on information they gathered. They asked truck drivers and traders about market conditions the previous day, looked at supplies of fish at various beaches, and guessed (erroneously) that high demand in the market one day would be followed by low demand the next. Although Quinn’s methods were inspired by the work of cognitive psychologists, she does not accept all of their conclusions: [T]he fish sellers interviewed in the present study did not seem to rely upon representativeness, availability, or any other heuristic that psychologists have observed their subjects applying to probability judgment tasks. Perhaps in recognition that their judgments of probability “foretell the future” poorly, or perhaps because other, less complex or more concrete indicators of future market conditions are available to them, fish sellers avoid making probability estimates altogether. Instead, the sellers report using information about a number of currently available test indicators of future market conditions. Sellers assess each of these indicators independently and go to market only if none of their tests indicate unfavorable market conditions. (209)

Quinn made little attempt to examine the extent to which fish sellers in practice used the heuristics they reported. Her work inspired considerable research on varied topics in cognitive anthropology and may have had some influence on the work of Sutti Ortiz and other economic anthropologists in the 1980s. However, Quinn quickly turned to other interests within cognitive anthropology such as “schema” concerning American marriages (Quinn 1982, 1996). Few anthropologists nowadays examine the use of heuristics in economic decision making.5 Mary Douglas’s treatment of risk is markedly different from those of Ortiz, Cancian, and Quinn. Douglas’s “cultural” approach has had considerable influence outside of anthropology (for example, Lupton 1999). Douglas says that differences between “expert” and “lay” judgments of risk have nothing to do with ordinary people’s inability to think in terms of probabilities (1985:3). She argues instead (1992:8) that when people estimate probability and credibility, their judgments are influenced by culturally learned conventions, expectations, and categories. Douglas emphasizes intercultural and intracultural variations in judgments about what kinds of risks are considered to be dangerous. As Lupton points out (1999:45), Douglas asserts that the risks that receive most attention in particular cultures are those connected with legitimating moral principles.

86 Anthropology, Economics, and Choice

Although Douglas’s social constructivist views on risk would seem to be attractive to the many anthropologists disliking the scientistic models of economists, biologists, and psychologists, her ideas have been controversial even among those scholars sympathetic to her theoretical approach. In several publications (Douglas 1990, 1992; Douglas and Wildavsky 1982), Douglas and her colleagues argue that certain cultural values have led many people in western societies to overestimate the dangers of particular environmental and health risks. This has led Douglas and colleagues to take political positions at odds with those of many scholars (among them Kaprow 1985, Lupton 1999:57) who regard her as an apologist for conservative business interests. Douglas’s general point nonetheless is clearly right. Humans have often overestimated or underestimated dangers from threats such as epidemics, earthquakes, unhealthy foods, and terrorist attacks. Briscoe and Alersey-Williams offer an opinionated, popularly written review (2009) of such misestimates nowadays. A recent example was the “swine flu” scare in 2009, when the vice president of the United States warned unreasonably of the perils of traveling in confined spaces such as airplanes and subways and the Egyptian government tried to kill every pig in the country. Although cognitive psychologists might analyze such fears as the results of inappropriate use of heuristics, it would be foolish to deny that they are related to cultural concerns. Most economic anthropologists writing about decision making in situations of risk and uncertainty avoid theoretical discussions altogether. They instead describe activities that reduce risk such as planting in multiple fields, growing subsistence crops, engaging in multiple occupations, helping kin and neighbors in times of need, and patronizing stores that provide credit. The unconventional economist Sanjay Reddy derides such ethnographic description: In much of the more naive and derivative anthropological literature, risk is largely perceived as being simply “objectively” given (the risk, for example, of crop loss due to pestilence, of collapse in the price of a cash crop, or of attack by one’s neighbors) and the relevant questions are perceived to concern how it is that communities adapt to and live with these risks . . . Much of this literature has unreflectively and uninterestingly adopted economists’ conceptions to arrive at a picture of peasants and tribals undertaking probabilistic “risk” calculations as a means of arriving at key decisions. (1996:230–231)

Risk, Uncertainty, and Decision Making 87

Reddy’s comments betray a misunderstanding of the ideas of most economic anthropologists, who are ordinarily skeptical about the use of the models of mainstream economists. More importantly, there is nothing trivial about descriptions of how individuals and communities “adapt to” and “live with” risks. In an edited collection dominated by model- driven anthropological approaches to risk and uncertainty (Cashdan 1990), Michael Baksh and Allen Johnson cogently argue that one of the contributions anthropology is best equipped to make to risk studies is to provide an ethnographic context crucial for a full understanding of risk avoidance behavior . . . [M]ost students of risk in economic, social, and other behavior would argue that the real situations in which decisions concerning risk are made are far more complex than the formal models used to analyze them, and some would follow us in the further claim that realistic analyses of risk may need to be quite a bit more complex, and less elegant, than formal models will allow. To be sure, textbook analyses of risk, such has how to invest in a lottery ticket where probabilities of winning or losing wagers of various amounts are perfectly known, depict decisions far cleaner and simpler than the everyday decisions of most men and women, whether in the modern business world or in a distant subsistence economy. (194)

Baksh and Johnson proceed to illustrate their point (with which I completely agree) by describing the myriad activities of the Machiguenga of tropical Peru to deal with risks related to the environment, subsistence, social conflict, and cultural loss.

Conclusions Economists making expected utility assumptions have developed seemingly straightforward methods for analyzing how decisions should be made in situations of risk and uncertainty. They first estimate the probability of different outcomes when particular choices are selected. Their next step is to estimate the payoffs (utilities) associated with different outcomes. Using their estimates of probabilities and payoffs, they then calculate the expected (average) payoffs of the various options. “Rational” decision makers, they argue, should pick the option that provides the highest expected payoff.

88 Anthropology, Economics, and Choice

Many economists recognize that the expected utility approach to decision making has certain shortcomings. Individuals differ in their willingness to take on risk. Some such differences are idiosyncratic. Others, however, are the predictable consequences of poor people being especially unwilling to make risky decisions that may lead to the loss of their livelihoods. In their efforts to deal with differential “risk preferences,” economists have created complex mathematical models of decision making. Although many economists limit their work on risk and uncertainty to model building, others have attempted to compare the implications of their prescriptive models with actual decision making. The results have not been all that impressive. Cognitive psychologists have questioned the fundamental assumptions behind economists’ models of decision making. Their critiques have focused on the extent to which individuals can estimate the odds of different outcomes. In a famous series of ingenious, well-designed experiments, psychologists have shown that in laboratory settings “subjects” do not estimate probabilities of outcomes to decisions. They instead ordinarily use various heuristics to simplify decision making. The social science literature on risk and uncertainty is dominated by economists’ models and psychologists’ experiments. These scholars largely ignore Knightean uncertainty, usually preferring to analyze situations of pure risk (although the relevant probabilities may be “subjective” rather than “objective”). By neglecting uncertainty, economists and psychologists are failing to consider the vast majority of decision-making situations. Economists and psychologists writing about risk and uncertainty usually devote only a paragraph or two to the relevance of their models and experiments to on-the-ground decision making. They rarely discuss multiple risks and uncertainties, choices that fall along a fuzzy riskuncertainty continuum, hard-to-measure and often incommensurable costs and benefits, and conflicting short-term and long-term payoffs from decisions. The discussion in this chapter of choices encountered by Iowa farmers, Peruvian ribereños, and Oaxacan wood carvers shows how such complexities can affect decision making under risk and uncertainty in real-world contexts. There are obvious reasons social scientists interested in risk create simplified models of reality and conduct experiments in carefully controlled laboratory settings. Such models and experiments are designed to eliminate many of the confusing and extraneous variables that can obscure the effects of risk and uncertainty on decision making. The crucial

Risk, Uncertainty, and Decision Making 89

question is whether the gains from such simplifications outweigh the loss of nuance. In my view, they rarely do. Economists and psychologists do not seem much impressed by imprecisely phrased generalizations about how diversification reduces risk and information gathering reduces uncertainty. Vague as they may be, I think that such generalizations have been more helpful in understanding human behavior than the great majority of economists’ models and psychologists’ experiments. There is, however, no substitute for ethnographically rich descriptions of the complexities of decision making in particular risky and uncertain situations. This is the work of anthropologists. Through their detailed ethnography, anthropologists have made an invaluable, though often neglected, contribution to our knowledge of such choices.

CHAPTER 4

Experimental Games and Choices about Cooperation

Sociocultural anthropologists, unlike their colleagues in psychology, sociology, and economics, rarely experiment. The field situations of anthropologists ordinarily do not allow the intentional, careful manipulation of variables required in good experiments. Furthermore, most sociocultural anthropologists explicitly avoid such manipulation; their goal is to minimize the effects of their presence on what they observe. Many have theoretical objections to experimentation. The isolation of a few variables for analytic purposes in experimental research differs markedly from the many holistic studies in sociocultural anthropology that emphasize the complex interaction of multiple variables. In recent years, however, a close-knit group of sociocultural anthropologists has been conducting controversial cross-cultural economic experiments aimed at testing ideas from evolutionary biology and economics about cooperation and competition. In this chapter I critically examine the assumptions and logical underpinnings of these ambitious, intriguing experiments. I contrast the parsimonious theoretical models underlying the experiments with the messy ethnographic reality of daily life.

An Overview of Economic Experiments in Anthropology In the summer of 1996 an indigenous group in the Peruvian Amazon took part in an economic experiment. Under the supervision of Joseph Henrich, the Machiguenga played the Ultimatum Game, in which two participants have a fixed amount of money (the stake) that must be divided between them. One participant, the “proposer,” suggests how the stake should be divided. The other participant, the “responder,” may accept or

Experimental Games and Choices about Cooperation 91

reject this offer. If the offer is accepted, both participants receive the share suggested by the proposer. If the offer is rejected, neither participant receives any money. This brief piece of fieldwork by a graduate student at the University of California, Los Angeles, was the first of many anthropological experiments aimed at comparing the extent to which individuals in particular societies are willing to be generous with others. These cross-cultural experiments, well-funded by the standards of sociocultural anthropology, have received extensive publicity. In the late 1990s the MacArthur Foundation gave grants to anthropologists to study the Ultimatum and other games in fifteen small-scale societies. The results of the MacArthurfunded research were discussed in a cover story in Scientific American (Sigmund, Fehr, and Nowak 2002) and reported in other widely read publications such as Nature (Fehr and Gachter 2002), Science News (Bower 2002), and the Wall Street Journal (Wessel 2002). In 2002 the Cultural Anthropology Program of the National Science Foundation (NSF) gave its largest grant ($463,425 over a three-year span) for further research on this topic to many of the same anthropologists who had been funded by the MacArthur Foundation. Stuart Plattner, then head of the NSF Cultural Anthropology Program, commented that “this new research is cutting edge stuff . . . [that] will advance economic theory” (in Bower 2002:106). The most publicized finding of these experiments is a significant positive correlation between the degree of market integration of a society and the proportion of the stake offered to the responder in the Ultimatum Game. This correlation has been widely interpreted (for example, Ensminger 2002:60, Surowiecki 2004:125) as an indication that the development of markets is associated with greater fair-mindedness and trust of strangers. Proponents of experimental economics in anthropology confidently assert that this and other findings “illuminate the nature of human nature, the potential importance of culture, and the appropriateness of the assumption of self-interest that underpins much of social science” (Henrich et al., “Overview and Synthesis,” 2004:10). These experiments have received so much funding and publicity because they aim at providing novel insights about long-standing questions in both the natural and social sciences concerning cooperation and competition. Evolutionary biologists since Darwin have been interested in the origins of altruistic behavior, when individuals act in ways that are good for their group but appear to reduce the chances of passing on their genes to future generations. The now generally accepted way to resolve

92 Anthropology, Economics, and Choice

this seeming paradox is to argue that particular cases of altruistic behavior that threaten the survival of individuals actually help them pass on their genes by increasing the probability of survival and reproduction of closely related kin. While such kin selection explanations seem to work well for most species, they do not account for the many situations in which humans act in ways that really do seem to reduce the chances of their genes being passed on. In such cases, altruistic behavior involves self-sacrifice for either distant relatives or nonrelatives. Economists influenced by rational choice theory are often puzzled when individuals act in ways that do not appear to be in their self-interest. Such activities, often involving cooperation with others, are a challenge to the underpinnings of their discipline. The ways in which economists usually think about generosity have nothing to do with the genetic advantages of helping closely related kin. They instead sometimes argue that generous behavior in the short run actually helps individuals in the long run. If such arguments seem implausible, economists may resort to the unprovable claim that the culturally valued, hard-to-measure “utility” that an individual gains from a good deed outweighs more easily measured losses such as monetary cost and the risk of physical injury or illness. One of oldest issues in the social sciences concerns the ways in which different groups deal with conflicts between the individual pursuit of self-interest and culturally variable ideas about the public good. Societies clearly differ in their relative emphasis on individual “rights” and group well-being. Within democratic countries, there are often heated debates between advocates of “individual liberty” and those motivated primarily by their vision of the good of the community. The lines of these arguments do not always break along conventional conservative/liberal divides. Examples of such controversies in the contemporary United States include debates over marijuana use, gun ownership, abortion, same-sex marriage, and legalized gambling. The creation of experiments such as the Ultimatum Game in the 1980s took place because economists and psychologists wanted to use controlled laboratory settings to examine the extent to which individuals would sacrifice their self-interests in order to help others. Prior to the work of Henrich among the Machiguenga, the great majority of these experiments took place in industrial countries, often with university students as subjects. Some economists and psychologists worried that only limited conclusions could be drawn from the experiments because the participants shared similar culturally influenced ideas about generosity and selfishness. The appeal of the anthropological experiments in small-

Experimental Games and Choices about Cooperation 93

scale societies came from their broadening the pool of subjects by including participants from many cultures. This would, at least in theory, allow a better assessment of the effects of culture on generosity and selfishness. Information on cultural variability was also of great interest to evolutionary biologists theorizing about supposedly genetic reasons for altruism. If there was considerable variation in the extent of altruism in different societies, evolutionary biological explanations for such behavior might be questioned. I find much to admire about the goals and activities of the anthropologists carrying out economic experiments. Their methodological rigor, willingness to engage in dialogue with scholars in both the natural and social sciences, and commitment to cross- cultural comparisons are refreshing in a time when many anthropologists question any attempts to measure variables and test hypotheses. Nonetheless, I think this research exemplifies the dangers of ignoring ethnographic complexity when attempts are made to create models of human behavior that can be mathematically formulated and experimentally tested. Furthermore, the proponents of these experiments are in my view often insufficiently critical of theoretical assumptions in biology and economics and insufficiently attentive to relevant anthropological theory. The findings of these economic experiments are, I will argue, consistent with long-accepted ideas in sociocultural anthropology. To understand why this research is nonetheless of such great interest to scholars in other disciplines, I describe here the history of experimental economics and theories about “selfishness” in evolutionary biology and the social sciences. I critically examine in some detail the studies of those anthropologists conducting economic experiments. Emphasis is placed on the disjuncture between what people do in these experiments and how they behave in their daily lives. Before presenting these theories and studies, however, I think it is useful to present some concrete examples of the tension between economic selfishness and generosity among Oaxacan wood carvers. The complexity of such real-life examples suggests why a certain degree of skepticism may be warranted toward wide-ranging theories about competition and cooperation.

Openness and Secrecy among Oaxacan Wood Carvers Wood carvings from the Mexican state of Oaxaca entered the international folk art market on a large scale starting around 1985. The trade

94 Anthropology, Economics, and Choice

originated somewhat earlier as a result of the activities of shop owners based in the city of Oaxaca and two particular carvers from communities in the nearby Central Valleys, Manuel Jiménez (1919–2005) of Arrazola and Isidoro Cruz (1934– ) of San Martín Tilcajete. Jiménez began to carve wooden figures as a boy to pass time while tending animals. In the late 1950s and early 1960s, owners of crafts shops in Oaxaca bought a number of his carvings and sold them to prominent folk art collectors. By the late 1960s, Jiménez was giving exhibitions in museums in Mexico and the United States. His carvings were featured in books and films about Mexican art. As Jiménez’s reputation grew and rural roads improved during the 1970s, tourists and collectors began to visit his workshop in Arrazola. Isidoro Cruz learned to carve during a yearlong illness in the late 1940s. While he was working as an oxcart maker in the city of Oaxaca in 1968, Cruz’s carvings were noticed by Tonatiúh Gutiérrez, the director of expositions for the National Tourism Council. Gutiérrez, who knew about Manuel Jiménez’s work, urged Cruz to make masks for sale. Gutiérrez later became director of a national agency that attempted to increase sales of Mexican crafts. He appointed Cruz as head of the agency’s buying center in the city of Oaxaca. Cruz kept this job for four years. During this time he was able to get jobs for some men from San Martín Tilcajete in government offices around Mexico. These men were therefore able to become knowledgeable about the folk art business. Jiménez and Cruz differed greatly in their willingness to share information about their artistic methods. Manuel Jiménez was secretive, showing his techniques only to his children and a son-in-law. The experiences of Miguel Santiago, a relative of Jiménez who is now a skilled, successful artisan, illustrate the extent to which Jiménez hid his methods. In the early 1980s Miguel Santiago was unsure about what paint to use on his pieces. He tried a paint used on posters but gave up when a potential customer’s hands became smeared while handling a carving. Because Santiago knew that Jiménez had abandoned water-based aniline for another type of paint, he searched the ravines around Arrazola for paint cans that Jiménez had discarded. These cans rarely had labels. When Santiago finally found a can with a label, he still was unsure what the paint was. He suspected that the Jiménez family was using a kind of house paint, but he was not certain about this until a local shop owner confirmed his hunch. Because of Manuel Jiménez’s secrecy, the craft spread slowly in Arrazola, and there were only about six carving families in the community in 1985. In the next five years Jiménez’s methods became well known, and by

Experimental Games and Choices about Cooperation 95

1990 most families in Arrazola had at least one artisan member. Although many of the new carvers and painters in Arrazola were talented, few could match Manuel Jiménez’s skill and none had his reputation. Manuel remained until his death the most famous carver and wealthiest resident of Arrazola. When interviewed, he often disdained the carvings of his neighbors, calling them “copiers.” In a typical statement, the maestro proclaimed: Mine is a sacred history . . . I am not just anybody. I am a real tiger. I was born intelligent. Everyone here is living off my initiative. If I hadn’t started carving, no one would be doing anything. I invented the whole tradition. They should make a statue for me in the plaza, with an arrow pointing to the house and rename this street Jiménez Street. (In Barbash 1993:19)

Isidoro Cruz, in contrast, was open about his methods and taught many of his neighbors how to carve and paint. A number of carvers in San Martín quickly became well known for their pieces and earned more money from the craft than Cruz did. Their pieces rarely resembled those of Cruz, who worked with difficult-to-use paints and woods and creatively made new types of carvings with little regard for their potential salability. Most of Isidoro Cruz’s friends and relatives in San Martín preferred to model their pieces after those of Manuel Jiménez. The difference in the willingness of Jiménez and Cruz to share their methods is in part related to their ideas about the advantages and disadvantages of cooperation with other members of their community. Jiménez regarded other artisans as competitors for a product for which there was a limited market. He was no more willing to show his methods to his neighbors than a company would be to give away trade secrets. Cruz thought that all the wood carvers in San Martín would benefit if there were numerous well-known artisan families in the community. If the town could establish itself as a wood-carving center, more tourists and collectors would visit and business would be better for everyone. His reasoning resembled that of entrepreneurs opening a particular type of ethnic restaurant in a large city in a neighborhood that already has many such restaurants. Such entrepreneurs assume that they will succeed because of the customers attracted to the area. Manuel Jiménez and Isidoro Cruz differed in their approach to cooperation and competition because of both their personalities and their positions in the wood-carving trade. Manuel was much more suspicious

96 Anthropology, Economics, and Choice

of the activities of others than was Cruz. Jiménez was gregarious in his dealings with tourists, but this extroversion did not extend to most residents of Arrazola. Although Jiménez lived in a town where participation in community organizations was a central part of daily life and where social events such as fiestas and weddings were of great importance, he largely kept the life of his immediate family separate from that of his neighbors. Cruz, in contrast, was a pillar of the community and one of the best-liked men in San Martín. These personality differences, however, are only part of the story. Manuel Jiménez reasonably feared that copying by others would threaten his position as the most famous artisan in Arrazola. For many years he had been the only important wood carver in the entire state of Oaxaca. If others created pieces similar to his and charged lower prices, Jiménez might lose both economic clout and prestige. Isidoro Cruz was the preeminent wood carver in San Martín for only a short time. Furthermore, he quickly found out that other artisans were more interested in Jiménez’s methods than his own. The costs and benefits of openness and secrecy were quite different for the two men. Even ignoring their personalities, Cruz had more reasons to share his methods than Jiménez. Over the years Oaxacan wood carvers have developed specialties in their efforts to appeal to a diverse clientele. Some artisan families make expensive, labor-intensive carvings for collectors; others churn out cheap pieces for gift shops in the United States and tourists seeking souvenirs. Artisans vary in their styles of painting and carving and in the size of their pieces. They make animals, human figures, devils, angels, frames, chairs, tables, and ox carts. There are carvings of Benito Juárez, subcomandante Marcos (the Zapatista leader), chupacabras (imaginary beings that eat goats), Martians, mermaids, and helicopters. When artisan families create successful market niches for their carvings, they know that they are likely to suffer economically if their neighbors create similar pieces at lower prices. It is not surprising that many artisans resemble Manuel Jiménez in their secrecy and complaints about imitators. Artisans go to great lengths to develop skills that enable them to create pieces that cannot be easily copied. Nowadays most artisans share Isidoro Cruz’s perspective that some degree of cooperation is necessary to attract tourists and dealers to their communities. In the past the extent to which artisan families cooperated to sell their pieces was limited because of their competition with one another for the attention of dealers. Since 2000 the demand for Oaxacan wood carvings has weakened significantly because of changing tastes

Experimental Games and Choices about Cooperation 97

in folk art and a decline in tourism because of political problems in the state. Many wood carvers have abandoned the craft, relying on income from other kinds of work and money earned through temporary migration to the United States. Those who have remained artisans have been more willing to participate in organizations that collectively market wood carvings. This necessarily means some loss in secrecy about techniques and a greater willingness to share knowledge about ways to make salable pieces.

Advantages and Disadvantages of Artisans Sharing Information Despite the emphasis on altruistic sacrifice by evolutionary biologists and economists, most cooperative behavior provides some benefits to all parties involved. When Isidoro Cruz and other wood carvers share information about methods with their neighbors, they are doing more than just contributing to good community relations. They are also improving the overall quality of artisanry in their community, which is likely to attract more potential buyers such as tourists and folk art dealers. Individual Oaxacan wood carvers sharing techniques that they have originated must weigh such benefits against the cost of losing competitive advantage in the marketplace. For Cruz the benefits clearly outweighed the costs; his innovations turned out not to be ones that his neighbors wished to emulate. Manuel Jiménez, however, worried with reason that his techniques and artistic styles would be copied by neighbors who would sell similar pieces for lower prices. But in the long run it was impossible for Jiménez to hide his methods, and his secrecy resulted in his losing popularity within his community. By this time Jiménez may not have cared too much about his local reputation; his self-esteem rested on the praise he received from collectors and art critics in the outside world. Jiménez’s economic fears about the results of copying seem to have been misguided. His family did better than ever economically after Arrazola became a noted wood-carving center with many skilled artisans. The Oaxacan example shows why in many cases neither insiders nor outsiders can easily judge whether a real-life decision involves a sacrifice. To call the sharing of an artistic innovation “altruistic behavior” would be misleading in rural Oaxaca. It is also difficult to unambiguously identify “selfishness” among Oaxaca wood carvers. Manuel Jiménez’s neighbors thought his secrecy was selfish because it prevented them from improving their miserable standard of living. He regarded their copying of his

98 Anthropology, Economics, and Choice

methods and selling of pieces similar to his at lower prices as a selfish lack of respect for the originator of an artistic tradition. From Jiménez’s perspective, his successors were freeloaders piggybacking on his hard work. The decisions that Jiménez, Cruz, and other Oaxacan wood carvers have made about openness and secrecy illustrate well the multiple risks and uncertainties associated with many choices about cooperation and competition. Artisans sharing a particular innovation cannot predict the economic effects of their generosity. The new technique may turn out not to be of much interest to potential buyers. If the innovation does attract more visitors to the community’s workshops, the sales of the originator could conceivably either increase or decrease. Sharers of a new technique also cannot know if the recipients of their generosity will be able to replicate or improve on their innovation. Artisans refusing to share their techniques similarly cannot predict with certainty what the effect will be on their reputation. Wood carvers understand the desires of artisan families to establish niches in the market by specializing in certain techniques and styles. Although many artisans are happy to copy successful innovations, they know that a certain amount of secrecy is economically necessary. Local social norms do not unambiguously favor sharing; wood carvers who freely give out information about all their methods are regarded as foolish. Nonetheless, an artisan who is unwilling to share any information about techniques, styles, and potential buyers is disdained as being too selfish. Artisans therefore walk a fine line between sharing too much and sharing too little and cannot predict exactly what will be the reactions to a particular instance of openness or secrecy. Wood carvers must consider both the short-term and long-term consequences of openness and secrecy. In the short term, secrecy may be preferable because it forestalls competition. But the long-term consequences of secrecy by particular artisans may be their lack of access to innovations by others. Artisans are most likely to share techniques with those who reciprocate. Advocates of economic experimentation in anthropology argue that we can improve our understanding of cooperation and selfishness through carefully designed games played in quasi-laboratories in field settings. A key question in assessing such claims is whether the essential features of complicated decisions such as those faced by Oaxacan wood carvers can be captured in such experiments.

Experimental Games and Choices about Cooperation 99

Evolutionary Theory and Cross-Cultural Economic Experiments When I first learned that sociocultural anthropologists were running economic experiments, I assumed that the inspiration for this work came from mainstream economics and economic anthropology. As an economic anthropologist, I knew that there was a long tradition within our subdiscipline of using ethnographic data from diverse societies to examine the cross-cultural applicability of economic theories. These experiments, I guessed, were attempts to test economists’ generalizations about human behavior. My assumptions about the impetus behind this research turned out to be only partly right. The Ultimatum Game and related experiments about public goods did originate as attempts to test economists’ assumptions about rational choice. Furthermore, some of the anthropologists involved in cross-cultural experiments (for example, Ensminger 2004) are explicitly motivated by classic questions in economics and anthropology such as debates over the effects of the development of markets on individuals’ willingness to treat strangers fairly. Nonetheless, the primary theoretical influences on most of these experimenters come from evolutionary ideas in biology and anthropology. The experimenters argue that the evolution of the capacity for culture has allowed individual human beings to be more able than members of other species to incur costs in order to help others and to punish selfish behavior. Their experiments are aimed at determining the extent to which cultural ideas about cooperation influence altruism and the punishment of selfishness. The advocates of economic experimentation in anthropology differ in the extent to which they attribute altruistic behavior to genetic propensities that evolved when all human beings were foragers. Some seem comfortable with conventional kin selection explanations from evolutionary biology. Most, however, stress the importance of the cultural transmission of ideas and culture-gene coevolution. Evolutionary Biology and the Problem of Altruism When evolutionary biologists discuss cooperative and altruistic acts, they almost always refer to behaviors that entail a cost to those performing them but that benefit others (Dugatkin 2006:28).1 This differs from the ordinary-language definition of cooperative behavior, which usually refers to acts that benefit everyone involved. Questions about altruism

100 Anthropology, Economics, and Choice

have troubled evolutionary biologists ever since the field was founded. Darwin worried about how his theory of natural selection could explain the existence of castes of insects such as bees that never reproduce but protect the lives of those that do. He speculated that the solution to this dilemma was that natural selection acted on families rather than individuals (Richards 1987:145). Darwin also opened up for debate the more general question about whether evolutionary pressures to sacrifice oneself for others extended beyond blood relatives. As Darwin suggested, explanations of altruism are inextricably intertwined with debates over the levels at which natural selection is thought to act. According to “trait group” models that developed in the 1970s, natural selection can operate both within and between groups. These models suggest that within-group selection acts against altruists. Selfish individuals receiving benefits from altruists but suffering none of their costs are more likely to pass their genes on to subsequent generations. Betweengroup selection, trait-group theorists say, can favor cooperation if groups with large numbers of altruists reproduce more than competing groups with fewer altruists. Most evolutionary biologists, however, agree with William Hamilton, who in an important article (Hamilton 1963) flatly rejected claims that altruism existed for the preservation of entire species. Building on ideas earlier developed by J. B. S. Haldane, Hamilton worked out the mathematics underlying kin selection.2 The simple idea underlying his not-so-simple equations was that the closer the degree of relationship of two individuals, the more likely they were to cooperate. The existence and behavior of members of nonreproducing insect castes could be understood by interpreting their altruism and sterility as maximizing their inclusive fitness, the chances of their genes being passed on to future generations. Selection could be best thought of as operating at the level of the gene rather than the individual or the group. Hamilton’s ideas became well known after they were popularized in two extraordinarily influential books about evolution, Richard Dawkins’s The Selfish Gene (1976) and E. O. Wilson’s Sociobiology (1975).3 Evolutionary biologists have questioned whether Hamilton’s approach can explain spiteful behavior. Spiteful acts lower the inclusive fitness of individuals committing them but decrease the fitness of recipients even more. The conundrum posed by instances of spite is that individuals who are not spiteful should have higher inclusive fitness than either initiators or recipients of punishment. Although instances of spite in nonhuman animals are rare, they are common among humans. In particular, indi-

Experimental Games and Choices about Cooperation 101

viduals and groups often punish those who fail to engage in altruistic behavior such as food sharing. Hamilton is also famous for his later work with the political scientist Robert Axelrod on cooperation among nonrelated individuals (Axelrod and Hamilton 1981, Axelrod 1984). Hamilton and Axelrod observed that in real life individuals often interact with one another repeatedly. In such cases, it would be an oversimplification to characterize any particular act as “selfish” or “cooperative.” Instead, individuals might develop different strategies toward those who had shown themselves willing to cooperate and those who were consistently selfish. Through models and computer simulations, Hamilton and Axelrod showed that in many circumstances the tit-for-tat strategy was optimal. Using this strategy, individuals cooperate with others as long as they remain cooperative. When others defect (become noncooperative), an individual using the tit-for-tat strategy also defects. Although such strategies have been effectively modeled on computers, they can only occur in the world among species in which individuals have the capacity to come up with and carry out ideas such as “tit for tat.” The work by Hamilton and Axelrod therefore suggests that cooperation (reciprocity) among nonrelated individuals might be more prevalent among humans than among other species. Culture and Evolutionary Theory The central theoretical and empirical questions underlying many economic experiments in anthropology are directly tied to evolutionary theory in biology. The experimenters want to solve what they regard as the “puzzle” (Henrich and Henrich 2007:4) of why humans cooperate. Their experiments are designed to contribute to cross-cultural comparisons of the conditions in which people are willing to cooperate, even if this involves incurring costs. They seem particularly interested in spiteful behavior toward “free riders” who are unwilling to reciprocate. The theoretical basis of such experiments is perhaps most thoroughly explained in the book Why Humans Cooperate (2007) by the anthropologists Natalie Henrich and Joseph Henrich. Unlike many evolutionary biologists, Henrich and Henrich emphasize the distinctiveness of cooperation among humans. They observe (40–41) that human cooperation differs from that of nonhuman primates in both its scale and the nature of its variability. Although the scale of cooperation in other primates rarely exceeds two or three individuals, humans sometimes cooperate on scales

102 Anthropology, Economics, and Choice

involving hundreds or even thousands of individuals. While nonhuman primate species typically show little variation in the behavioral domains of cooperative behavior, human social groups vary greatly in the activities in which they cooperate. Henrich and Henrich observe (55) that the existence and variability of reciprocity among human groups is made possible by certain features of our capacity for culture: cultural evolution occurs at a much faster rate than genetic evolution; cultural learning processes allow populations to rapidly adapt to novel situations without a genetic change; and cultural transmission can, to a degree, construct its own environment. Although these ideas might conceivably be novel for some evolutionary biologists, they are so commonplace in anthropology that they are covered at the beginning of most introductory courses. The ideas of the proponents of cross-cultural economic experiments (including some nonanthropologists) about the reasons for the evolution of cooperative behavior are more interesting. At times Henrich and Henrich argue that there is a gene for cooperation (!) and uncritically accept the basic ideas of kin selection: [N]atural selection will favor the evolution of psychological mechanisms that allow cooperators to focus their benefits on other individuals who are likely also to be cooperators . . . “Blood relatives” have many characteristics in addition to the all-important cooperation gene(s) and natural selection can take advantage of these in building psychologies that can preferentially direct benefits at other individuals likely to have cooperation genes. For example, close kin, such as siblings, may have a similar appearance or smell that natural selection may use to build a psychology such as “help those who look and smell like you.” These similarities may be related to sharing some of the same genes (ones not related to cooperation). (44)

Karl Sigmund, Ernest Fehr, and Martin Novak assert with respect to spiteful behavior that our “emotional apparatus” was shaped by years of living in foraging groups in which keeping secrets was hard: If others know that I am content with a small share, they are likely to make me small offers [presumably of food]; if I am known to become angry when facing a low offer and to reject the deal, others have an incentive to make me high offers. Consequently, evolution should have favored emotional responses to low offers. (2002:85)

Experimental Games and Choices about Cooperation 103

Such spitefulness, they go on to say, is based on “universal, biologically rooted capabilities” (87). These genetic explanations seem to be at odds with the emphasis on cultural transmission as being essential for the evolution of reciprocity and other forms of cooperation among both kin and non-kin. Henrich and Henrich and other economic experimenters deal with this apparent contradiction by endorsing a highly mathematical “dual inheritance theory” developed by Robert Boyd and Peter Richerson and others (Boyd and Richerson 1985, Durham 1991, Richerson and Boyd 2005). Stripped of the equations and jargon, this theory essentially suggests, in agreement with long-accepted, less formalized anthropological ideas (as in Steward 1955), that the genetically evolved capacity for culture that was adaptive in ancestral human lineages allows a diversity of behaviors in different situations. The proponents of cross-cultural economic experiments for the most part avoid controversies in evolutionary biology about levels of selection. Their brief statements about this question (for example, Bowles et al. 1997:9, Henrich and Henrich 2007:66) ordinarily suggest that group selection plays an important role in human cultural evolution. They argue that the groups that developed the most successful guidelines for pursuing fair interactions were evolutionary successful. The implicit or explicit assumption is that these groups developed genetic traits—perhaps merely the capacity for culture, perhaps something more like a “cooperation gene”—that helped them make equitable exchanges and punish free riders. Such claims are made even though most biologists are skeptical about group selection arguments to explain altruism in nonhuman species.

Experiments in Economics In 1948 Edward Hastings Chamberlin noted that economics has not historically been an experimental science: It is a commonplace that, in its choice of method, economics is limited by the fact that resort cannot be made to the laboratory techniques of the natural sciences. On the one hand, the data of real life are necessarily the product of many influences other than those which it is desired to isolate—a difficulty which the most refined statistical methods can overcome only in small part. On the other hand, the unwanted variables can-

104 Anthropology, Economics, and Choice

not be held constant or eliminated in an economic “laboratory” because the real world of human beings, firms, markets, and governments cannot be reproduced artificially and controlled. (95)

Nonetheless, in the past six decades an increasing number of economists, including Chamberlin himself, have conducted experiments in laboratories, usually with university students as participants. The diverse topics that have been examined include public goods, bargaining, industrial organization, auctions, and individual decision making (Roth 1995). The idea behind these experiments is to examine the extent to which participants’ behavior corresponds to the predictions of economic theories. Experiments often test assumptions about rational choice, especially those derived from game theory (von Neumann and Morgenstern 1944). The economists’ experiments have shown that notions of fairness, cooperation, and trust play a crucial role in economic transactions (at least in laboratories) and sometimes lead to outcomes different from those expected by rational choice theory (Chaudhuri 2009:17). Although such findings are not at all surprising to most anthropologists, economists have been impressed by the conclusion that “tendencies towards rampant selfinterest are [very often] moderated by notions of fairness” (ibid.:71). The interests of experimental economists overlap with those of evolutionary biologists. Both groups of scholars see cooperation as a puzzle not readily explained by prevailing theories in their disciplines, and they draw inspiration from game theory. They share an interest in spiteful behavior in which individuals incur costs to punish noncooperators. Experimental economists and evolutionary biologists, however, differ in the ways in which they have thought about the “puzzle” of cooperation. Biologists have been primarily concerned with the mechanisms underlying altruistic behavior, focusing on kin selection (often among nonhuman species). Economists, who ordinarily restrict their analyses to humans, are most interested in cooperation among nonrelated people in contemporary industrial societies. Unlike biologists (and anthropologists), most economists are relatively uninterested in how and why cooperative ideas and behaviors develop and change over time. They ordinarily emphasize instead what these ideas and behaviors are in different societies and how they affect strategic decision making. Economists and evolutionary biologists also differ in the amount of attention they give to altruism and notions of fairness. Perhaps because ideas about fairness are unimportant or nonexistent in nonhuman species, biologists focus on altruistic behavior. Economists, possibly because they find pure altruism so contrary to

Experimental Games and Choices about Cooperation 105

the fundamental tenets of their discipline, are much more concerned with social norms about fairness. Writers about the history of institutions have for centuries been interested in the relationship between markets and ideas about fairness. Some writers, notably Karl Marx, assert that markets undermine the moral foundations of societies by encouraging self-interested behavior. Others (including Baron Montesquieu) argue that markets encourage fairness because of the premium on honesty and reputation required in repeat dealings (Ensminger 2004, Hirschmann 1982). This latter idea was a forerunner of mathematical theories about reciprocity that note the advantages of strategies such as tit for tat in certain circumstances. The New Yorker economics columnist James Surowiecki elaborates on the argument that capitalism encourages fairness because much business is transacted between strangers: Modern capitalism made the idea of trusting people with whom you had “no prior personal ties” seem reasonable, if only by demonstrating that strangers would not, as a matter of course, betray you. This helped trust become woven into the basic fabric of everyday business. Buying and selling no longer required a personal connection. It could be driven instead by the benefits of mutual exchange. (2004:123)

The reason participants in market economies are willing to be fair in this limited sense, Surowiecki says, is “a greater emphasis on the accumulation of capital over the long run as opposed to the merely short-run, an emphasis that has been arguably a defining characteristic of modern capitalism” (122).

Economic Experiments by Anthropologists The Ultimatum Game is the most common economic experiment that anthropologists have carried out in their cross-cultural studies. The game is designed to measure the extent to which responders will sacrifice their own money to punish a proposer who has been unfair. A key feature of the game is that it is a one-shot affair played between anonymous participants; one’s behavior in the game (unlike much interaction in real life) has no influence on future socioeconomic relations. The Ultimatum Game, therefore, is supposed to provide some indication of abstract society-wide values concerning fairness in socioeconomic exchange be-

106 Anthropology, Economics, and Choice

tween strangers. The first Ultimatum Game experiment took place two decades ago (Güth, Schmittberger, and Schwarze 1982); since then hundreds have been carried out (Oosterbeek, Sloof, and van de Kuilen 2004). Social scientists have shown so much interest in the Ultimatum Game because participants’ behavior seems to violate the principles of rational choice. Economists generally assume that decision makers seek to either maximize gains or minimize losses of some type. In many situations, figuring out which behavior is “most rational” is far from obvious, and we cannot easily say whether individuals are acting as economic theory would predict. The best strategy for the responder in the Ultimatum Game, however, seems straightforward—accept whatever the proposer offers. Otherwise, the responder gets nothing. The best game theory strategy for the proposer is only a bit harder to grasp. Because the responder should accept any offer, the proposer should suggest the division of the stake that leaves the responder with the least possible amount of money. But in experiment after experiment with university students, many respondents have been willing to reject what they regard as unfair offers. Proposers, apparently aware of the possibility of rejection, typically make offers that leave respondents with a substantial portion of the stake. Accounts of these results often regard rejections of low offers in the Ultimatum Game as a puzzle to be explained. Why should participants care about fairness in this context? Furthermore, even among university students there is considerable variability in different countries in both the average size of offers and responses to similar-size offers. Are there intercultural differences in “selfishness”? If so, what are the causes of these differences? Psychologists and economists have given two principal interpretations of the results of Ultimatum Games. Some scholars (including Fehr and Gachter 2002 and Nowak, Page, and Sigmund 2000) emphasize similarities in results, seeing them as evidence that our hunter-gatherer ancestors evolved an emotional apparatus favoring generosity in situations where sharing was necessary for survival. Others (for example, Roth et al. 1991) emphasize variability in results, seeing them as evidence that cultures differ in their emphases on the importance of sharing and fairness. The behavior of the Machiguenga in Henrich’s experiment in the summer of 1996 differed considerably from that of the university students who had previously played the Ultimatum Game. Students often rejected offers; proposers typically suggested almost equal divisions of the stake. Among the Machiguenga an offer was only once rejected; the proposers’

Experimental Games and Choices about Cooperation 107

average offer to the recipient was only 15 percent of the stake. These differences cannot be facilely explained by the greater importance of money for the Machiguenga. The stake in the Machiguenga experiment was the equivalent of a day’s wages, the same as that offered in most games played by university students. Henrich’s experiment attracted attention because the large differences between how the Machiguenga and university students played the game suggested that culture influenced notions of fairness considerably more than had previously been thought. Moreover, ideas about the innate generosity of hunter-gatherers might be questioned since the Machiguenga (though primarily shifting cultivators rather than foragers) made such low offers. Funding agencies were therefore receptive when anthropologists sought support for comparative studies of economic experiments in settings very different from where the great majority of Ultimatum Games had previously been played. The economic experiments by anthropologists aim at providing crosscultural and cross-species perspectives on theories about selfishness by biologists and economists. They therefore have focused on • cross-cultural variations in generosity (conceived of as either “altruism” or “fairness”) • the extent to which individuals in different societies are willing to incur costs in order to punish selfish behavior • differences in the amount of generosity exhibited by participants in the Ultimatum and related games between foraging societies where reciprocity is the principal means of exchange and state societies with well-developed markets • the degree to which “prosocial” cultural activities such as food sharing, communal labor, and contributions to public feasts are reflected in behavior in experimental games.

The best summary of cross-cultural economic experiments can be found in Foundations of Human Sociality (Henrich et al. 2004), an edited collection with contributions from most of the major researchers in this field. The research sites examined in the book are described in its “Overview and Synthesis” (10) as consisting of “fifteen small-scale societies (from twelve countries on four continents and New Guinea) . . . [that include] three foraging societies, six that practice slash-and-burn agriculture, four nomadic herding groups, and two sedentary, small-scale agri-

108 Anthropology, Economics, and Choice

cultural societies.” This is an oversimplified classification of economies in places where people have multiple sources of food and income and have been affected to varying extents by colonialism, international trade, and globalization. Although this simplification is doubtless related to the researchers’ goal of placing their findings in an evolutionary framework, many readers might question, for example, whether the “foraging groups” described can really be regarded as representative of “hunter-gatherer economies” in the past. The three “foraging societies” appear to be the Hadza of Tanzania, the Au and Gnau of New Guinea, and the Lamalera of Indonesia. Only the Hadza are characterized as being purely foraging. The Au and Gnau are described as “foraging/horticulture,” while the Lamalera are listed as “foraging/trade.” (The Ache of Paraguay are described as “horticulture/foraging.”) Of the 1,000 Hadza in the study region, 300–400 were full-time foragers. The Ultimatum Game was played with the full-time foragers, but they clearly had contact with their nonforaging neighbors. The principal dietary staple of the Au and Gnau is starch extracted from sago palm raised in semi-wild stands. The Au and Gnau also practice pig husbandry and raise cocoa as a cash crop. The Lamelera trade whale meat and fish for agricultural goods, pots, and cloth. Tourism and cash remittances from wage laborers working elsewhere are significant income sources. To state the obvious, our hunter-gatherer ancestors did not raise crops and animals, trade with nonforagers, work as wage laborers, receive cash remittances, or welcome tourists. In each case study experimenters compared how different groups played the Ultimatum Game, usually attempting to test hypotheses about intergroup variation. The hypotheses and groups selected for analyses varied greatly. Criteria for assigning participants to groups included ethnicity, place of residence, degree of market integration, occupation, and whether one lived in a community that had recently been resettled. The researchers often attempted to see if the results of experimental games reflected local norms concerning exchange. Jean Ensminger nicely expresses the reasoning: It is conceivable that in the context of many small-scale societies, guaranteed perfect Anonymity is a rare event—so much so that people behave in many contexts as if their behavior will be known. Thus, we might find behavior in one-shot games consistent with behavior more appropriate to repeated games. Put simply, people invoke “rules of thumb” cued by reminders of familiar circumstances when faced by a completely novel situation, as might occur in an economic experiment. (2004:358)

Experimental Games and Choices about Cooperation 109

Not every contributor to the volume accepts this argument. Michael Gurven, unable to find a relationship among the Tsimane of Bolivia between market exposure and behavior in two experimental games, writes: The endowment of the games represents a “windfall”—a large sum obtained without any cost. It is unclear whether rules of thumb or cultural norms designed to direct costly acts of cooperation apply for cooperative acts that are relatively costless. Deciding how much of a windfall should go to others more accurately requires an economics of etiquette or manners. (2004:226)

Two examples illustrate the parallels that some researchers found between local patterns of exchange and the results of experimental games. Under Ensminger’s supervision, the Orma of northern Kenya played the “public goods game.”4 Many participants saw a resemblance between this game and harambee, a local institution of village-level contributions for community projects such as school building. The Kenyan government for many years has encouraged harambee fund raising as a method of community development. The norms associated with harambee, Ensminger says, were applied to the public goods game. Michael Alvard found that the Lamalera of Indonesia tend to make “fair” (relatively equitable) offers in the Ultimatum Game. The Lamalera obtain much of their food from cooperative whale hunting. Alvard argues that in big-game hunting societies the rewards for fairness in meat distribution are clear. He says, “To the extent that achieving such fairness is critical for subsistence, we expect fairness to be expressed in the Ultimatum Game” (2004:428).5 Because many of the contributors to Foundations of Human Sociality are sympathetic to evolutionary psychology, they often frame their discussions in terms of the supposed pan-human genetic universals suggested by the book’s title. Only two case studies (by Abigail Barr and Jean Ensminger) make no references to theories about such cognitive universals. Such theories are always respectfully considered, though some experimenters (especially Richard McElreath) note their limitations. In my view, the emphasis on evolutionary psychology is the weakest theoretical aspect of the project because a pan-human cognitive architecture obviously cannot explain intracultural and intercultural variation in experimental results. The contributors’ contorted attempts to escape this dilemma are unconvincing. Here are a few of many examples that could be cited in which authors allude to evolutionary ideas in ways that do little to aid our understanding of variability in the experimental findings:

110 Anthropology, Economics, and Choice

Are there innate social grammars for acquiring contextually specific rules and cues about fairness, cooperation, and punishment? (Henrich and Smith 2004:164–165) Understanding how cooperation works under realistic conditions is the key next step of understanding the variation seen in the results of the Ultimatum Game. Indeed, in retrospect, given the costs in maintaining and enforcing norms, it makes evolutionary sense that such norms not be fixed but rather vary in adaptive ways. (Alvard 2004:431–432) In many situations it is probably adaptive to assess welfare relatively since Darwinian fitness is relative. Spite makes sense if the endgame is relative success. People may use a conditional strategy to evaluate their welfare in relative or absolute terms depending on the circumstances. When the benefit of an absolute gain would ultimately add less to one’s fitness than preventing another from gaining the upper hand, people may behave in a way that only appears to be spiteful, but is actually rational. (Marlowe 2004:190–191)

If our genetic emotional apparatus allows variation in the degree of cooperative behavior in different circumstances, how can evolutionary psychology aid in our understanding of generosity and selfishness in the Ultimatum Game? The authors of the “Overview and Synthesis” of Foundations of Human Sociality had a hard task. The heterogeneity of the case studies, with the only constant the Ultimatum Game, made generalization difficult. The experimenters often (admirably) report findings that do not support hypotheses underlying the project. Here, for example, are two cases in which the game results seem unrelated to cultural influences: [R]esponses in the games shows [sic] no significant association with real-life behaviors such as measured time preference, previously measured food-sharing generosity, or variation in food production. (Hill and Gurven 2004:408) The Hadza made lower offers in . . . the Ultimatum Game . . . than is typical of complex societies. The result is especially interesting given the Hadza can only be described as extremely egalitarian, with a strong sharing ethic. . . . Perhaps the more frequently one must share, the more

Experimental Games and Choices about Cooperation 111

weary one grows of it, and the more one looks for any opportunity to escape it. (Marlowe 2004: 187)

Furthermore, correlations between variables that are positive in some places are nonexistent in others. David Tracer finds a positive relationship between degree of market integration and size of offer in the Ultimatum Game among the Au and Gnau of New Guinea (2004:253); Michael Gurven finds no such relationship in lowland Bolivia (2004:210–212). Richard McElreath finds that age correlates positively with size of offer among the Sangu of Tanzania (2004:340); Joseph Henrich and Natalie Smith find no relationship between these two variables among the Mapuche of Chile (2004:136–139). Despite these problems, the authors of the synthesis chapter say (10– 11) that the results of the project can be summarized in five points: 1. There is no society where people play the Ultimatum and related games as economics textbooks would predict (as income maximizers). 2. There is much more variation among groups in game playing than had previously been reported. 3. Differences between societies in market integration and the importance of cooperation explain a substantial portion of the behavioral variation between groups. 4. Individual-level economic and demographic variables do not explain behavior within or between groups. 5. Experimental play often mirrors patterns of interaction in everyday life.

Four of these findings are either not particularly interesting or only weakly supported by the case studies. Because income-maximizing behavior has not often been found among university students playing the Ultimatum Game, there would be no particular reason to expect decision makers anywhere else to make what classical economics would consider to be “rational” offers. Surprisingly, a table in the synthesis chapter (25) clearly shows that the mean offers of three groups (Hadza, Orma, and Sangu farmers) were in fact income maximizing, given the rate of rejections. However, the rejection rate in all the case studies was higher than would be predicted by rational choice theory. One of the principal objectives of the researchers in Foundations in Human Sociality was to conduct experimental games among a greater

112 Anthropology, Economics, and Choice

diversity of groups than had been represented in previous studies. It is therefore hardly startling that intergroup variation was higher than had previously been reported. I think it is more interesting that individuallevel economic and demographic variables did not for the most part explain game-related behavior within or between groups. Such variables have been shown to have significant effects on economic decision making in many settings and are often the starting points for analysis. However, the authors of the case studies rarely present theoretical reasons why the particular individual-level variables isolated for analysis (age, gender, relative wealth) should affect behavior in the games they supervised. In the absence of plausible connections among particular individual-level variables, offer sizes, and rejection rates, we cannot assess the significance of negative results. The evidence that experimental play often mirrors interaction patterns in daily life is soft by the rigorous standards that the project contributors extol. As noted, in some cases behavior in experimental games did not reflect such patterns in hypothesized ways. Even when the behavior in games did seem to have parallels to real-life exchanges, the evidence presented is qualitative and descriptive. Such a mode of explanation is entirely satisfactory to many sociocultural anthropologists (including myself ) but certainly falls short of the type of “science” advocated elsewhere by project personnel. In any case, the authors of the book’s overview are uncertain about the nature of the relationship between everyday life and game playing: [W]e are unclear about some important details of how local situations influence behaviors. Two plausible interpretations come to mind. Perhaps different social and physical environments foster the development of differing generalized behavioral dispositions that are applicable across many domains . . . In contrast, our abstract game structure may cue one or more highly context specific behavioral rules . . . According to this interpretation, our subjects were first identifying the kind of situation they were in, seeking analogs in daily life, and then acting in an appropriate manner . . . These two approaches are difficult to distinguish empirically and our dataset does not help us judge their relative importance. (48)

An assessment of Foundations in Human Sociality rests on claims that the economic experiments described in the volume help us understand fundamental questions about human nature related to selfishness and cooperation and show that people living in societies highly integrated into

Experimental Games and Choices about Cooperation 113

market economies are more willing to be fair to strangers in economic exchanges than people living in societies less integrated into market economies. The case studies in the book tell us little about human propensities to selfishness or cooperation. Although many of the experimenters relate their findings to evolutionary psychology, they usually do so in a perfunctory way. They are particularly bedeviled by the finding that foragers— assumed by evolutionary psychologists to have developed cooperation among close kin to increase their chances of propagating their genes—are more selfish in their game playing than their counterparts in more marketoriented societies. In practice, the experimenters place more weight on parallels between the results of the games and local institutions and cultural conventions related to economic exchange. The relationship between immersion into markets and norms of reciprocity with relatives, friends, acquaintances, business partners, and strangers is a question of central interest to many historians and social scientists. The contributors to Foundations of Human Sociality find that economic experiments like the Ultimatum Game can improve our understanding of this relationship. Most anthropologists would likely prefer to find out about norms concerning economic exchange via more conventional ethnographic methods such as interviews, observations, case studies, and surveys. While some of these methods (case studies, observations) would not readily allow cross-cultural comparisons, others (structured interviews, surveys) would. Furthermore, as many of the experimenters point out, it is not at all clear that participants in the games were treating their partners as “strangers.” Even if we accept the project’s methodology and conclusions, should we be startled by a finding that greater market integration is associated with greater willingness to treat strangers fairly in economic exchange? Anthropologists have long emphasized the importance of kinship in economic exchanges in places where markets are either absent or of limited importance. In such places kin are usually treated better in economic exchanges than strangers (Sahlins 1972:196–210). Because most exchange is with kin, a lack of trust of strangers does not threaten one’s economic livelihood. In larger-scale societies, people spend more time in economic exchange with either distant kin or complete strangers. In such exchanges, it becomes an economic necessity to have a certain amount of trust in particular types of strangers, and generalized norms of fair play develop. More generally, anthropologists such as Robert Redfield (1947) and sociologists such as Louis Wirth (1938) argued many years ago that social relations are fundamentally different in small-scale and large-scale

114 Anthropology, Economics, and Choice

societies. In small-scale societies, norms of behavior between individuals are governed by particularistic rules often based on kinship relations. In large-scale societies, where interactions with strangers are common, more generalized (universal) norms govern interactions. The findings reported in Foundations of Human Sociality would not surprise Redfield and Wirth. Although behavioral economics continues to flourish, anthropologists seem to have become less enthusiastic about the promise of economic experiments in the years since the publication of Foundations of Human Sociality. Fewer articles on this topic are appearing in major anthropological journals. Furthermore, those anthropologists continuing to conduct economic experiments have been placing more emphasis on the frequent lack of correspondence between how people act in games and their behavior in real-world situations. While questions about the relationship between decision making in laboratory experiments and real-world choices would occur immediately to most anthropologists, behavioral economists have published few relevant empirical studies. In a recent article, however, two prominent economists discuss in detail problems associated with drawing inferences about the real world from experimental studies. Steven Levitt and John List (2007) argue that behavior in laboratories is influenced by moral and ethical considerations, the nature and extent of the scrutiny of participants’ actions, and the context in which decisions take place. They say these problems are especially relevant for experiments like the Ultimatum Game that have a strong moral component. Levitt and List suggest that participants interacting with anonymous strangers in laboratory experiments are likely to be less prosocial (cooperative) than they are in daily economic exchanges with kin and friends. On the other hand, participants in laboratory experiments may be more prosocial than they are in realworld dealings in financial markets. Levitt and List’s discussion is mostly conjectural; they present only a few sketchy cases in which researchers attempt to compare laboratory and real-life decision making. Three recent anthropological studies examine this issue in more depth. Polly Wiessner (2009) played the Ultimatum and Dictator games with the Ju/’hoansi Bushmen of southern Africa. In the Dictator Game, proposers have a fixed amount of money, from which they are obligated to give some—as little or as much as they want—to responders. Responders must accept whatever offer is made. The Ju/’hoansi proposer offered relatively low sums of money in both games; responders in the Ultimatum Game rarely punished low offers by declining them. Wiessner then observed what participants in the experiments did with the money they

Experimental Games and Choices about Cooperation 115

earned. Participants readily shared food they bought with their winnings with many of their neighbors. They would not, however, share food with people with whom they had recently quarreled. Wiessner concludes, as Levitt and List would guess, that the Ju/’hoansi were much more likely to be “highly self-regarding” (selfish) when there were no faces of loved ones or friends present to make them want to share, especially in those experimental situations when the social conventions that influence everyday sharing and the punishment of free riders seemed irrelevant. Michael Gurven and Jeffrey Winking (2008) conducted a more detailed study along the same lines among the Tsimane of Bolivia. The Tsimane played the Ultimatum and Dictator games and the related ThirdParty Punishment Game. The Tsimanes’ offers and rates of rejection were low, similar to those of the Ju/’hoansi. Gurven and Winking also examined the extent to which the Tsimane exhibited prosocial behavior in activities such as digging a communal well, sharing food, and contributing to a public feast. They found that the Tsimane, again resembling the Ju/’hoansi, were more prosocial in real-life situations than in experimental games. However, the amount of prosociality the Tsimane exhibited in real-life situations varied considerably, with people overall being more cooperative in some contexts than in others. Natalie Henrich and Joseph Henrich (2007:171–173) compare the actions of Arab-American Chaldeans in Detroit in the Ultimatum and Dictator games with their cultural attitudes toward business dealings and charitable giving. The Chaldeans made more generous offers in the Dictator Game than in the Ultimatum Game. This seems counterintuitive because responders in the Ultimatum Game, unlike those in the Dictator Game, have the option of refusing an offer. Henrich and Henrich interpret this surprising result in terms of cultural models evoked by the two games. The Ultimatum Game, they say, triggers a business model in which there is motivation both to cooperate and to pursue self-interest. The Dictator Game evokes cultural norms associated with charitable giving that value generosity toward others. Gurven and Winking, Wiessner, and Henrich and Henrich all say, as would any anthropologist, that cultural norms affect the extent of cooperative behavior in real-life situations. They agree that the amount of prosocial behavior is context- dependent; people who are cooperative in certain situations will be selfish in other circumstances. Henrich and Henrich point out that when experimental games resemble real-life situations, the relevant cultural norms will influence decisions such as the amount to offer in a game. Wiessner and Gurven and Winking empha-

116 Anthropology, Economics, and Choice

size differences between experimental games and real-life situations and say that cultural norms are often of little use in predicting game behavior. Each of these anthropologists has added to our knowledge of how and when cultural norms related to prosociality affect behavior in the Ultimatum Game and other experimental games. But the goal of anthropology is not to learn how people play such games; it is to describe and understand their behavior in the world outside the laboratory.

Conclusions Social scientists and biologists have long been interested in the reasons people cooperate with one another. The evolution of the capacity for culture has allowed humans to work together in ways that are impossible for other animals. Such cooperation is a major reason our species dominates the planet. However, cooperation often entails short-term costs for some participants in exchanges. For this reason cooperation is not inevitable; free riding and other selfish behavior is common. The principal explanations offered by biologists and economists for why people sometimes cooperate have serious shortcomings. The biologists’ focus on kin selection ignores the immense amount of human cooperation with nonrelatives. The economists’ rational choice theories work well when cooperative behavior is useful for all parties involved, but they falter when trying to explain why some individuals make costly sacrifices to help others. Anthropologists usually explain cooperative activities with reference to their usefulness for society as a whole (the “public good,” in the language of political scientists and economists). Such grouplevel explanations are unpalatable to the many evolutionary biologists and economists whose analyses examine the advantages and disadvantages for individuals of particular types of behavior. Anthropologists also sometimes attribute cooperation to social norms without paying sufficient attention to how and why such cultural rules develop and change over time. Many of the goals of anthropologists conducting economic experiments are consistent with the historical roots of their discipline. By carrying out experiments in diverse societies, they provide a cross-cultural perspective on previous research done in western industrial countries. These anthropologists, in marked contrast to most of their counterparts in other fields conducting economic experiments, emphasize how culture influences decision making both inside and outside the laboratory. They show how cultural rules about cooperation in any particular society vary in

Experimental Games and Choices about Cooperation 117

different situations. Individuals may cooperate more with kin and close friends than with strangers. Cooperative behavior may be strongly prescribed in some contexts (in an army unit) and loosely desirable in others (charitable giving). Anthropologists running economic experiments in the field sometimes explain behavior in their quasi-laboratories by pointing out resemblances between the games played by participants and particular real-world institutions. In other cases they argue that the lack of cultural similarities between these games and real-life situations results in behavior in experiments that seems inconsistent with social norms. Anthropologists conducting economic experiments have focused rather narrowly on certain types of cooperative behavior. They have rarely looked at situations, common in the world outside the laboratory, in which cooperative behavior is useful in the short run to all parties involved. Instead, their experiments usually involve situations in which cooperation incurs costs for some participants. They are particularly interested in the extent to which individuals are willing to incur costs in order to punish selfish behavior. These emphases come from theoretical debates about altruism in evolutionary biology and psychology. The underlying assumption of economic experimentation in anthropology and other disciplines is that behavior in laboratories can tell us something about the wider world. Anthropologists, economists, and psychologists writing about this issue have concentrated on the correspondence between abstract experiments and real-life situations. Their emphasis is on the extent to which social norms in the wider world are reflected in decision making by participants in experimental games. While such discussions are useful, they ignore the principal difficulty with such games. Economic experiments intentionally create situations in which variables are carefully defined and measured and the confounding effects of extraneous factors can be controlled for through manipulation of variables. In my view such “rigor” results in experiments that diverge too much from real-life situations to be of much use in understanding social norms about cooperation and selfishness. The countless economic experiments by anthropologists and other social scientists do not, for example, aid in my understanding of when and why Oaxacan wood carvers share information about their techniques, styles, and customers. Their willingness to cooperate with their kin and neighbors in certain circumstances can only be understood through ethnographically informed analyses of the history and economics of the wood-carving trade.

CHAPTER 5

Who Makes Household Economic Decisions?

Anthropologists and economists analyzing decision-making situations in particular times and places need to specify who exactly has the power to make choices. Before around 1960 economists ordinarily assumed that decision-making units were either individuals or groups acting as a single entity. This assumption simplified their models of choice based on neoclassical economic theory. In the real world, however, many group decisions are the outcomes of discussions between people with conflicting goals. For example, the firms that are the subject of extensive economic theorizing have numerous decision makers who may disagree about what their companies should do in particular situations. Some of these individuals may exert great amounts of influence in certain types of choices but have little power to determine the outcome of other kinds of decisionmaking situations. Over the past fifty years anthropologists and economists have become increasingly aware that many important economic decisions are made within households whose members pool some of their resources. Although anthropologists have written extensively about problems in defining and delineating “households,” economists for the most part seem unconcerned about these issues. Both anthropologists and economists, however, have written extensively about the difficulties associated with regarding households as decision-making units. They understand that members of households negotiate and compromise over the use of shared resources. Anthropologists’ work on intrahousehold negotiations has mostly been descriptive; economists, in contrast, have developed game theory models of bargaining among household members. Feminist scholars in the two disciplines have been particularly concerned with how society-wide gender relations affect the relative influence of women and

Who Makes Household Economic Decisions? 119

men in household decision making. They have also emphasized the considerable autonomy that women and men often have in different spheres of household economics. This chapter examines similarities and differences in the ways anthropologists and economists discuss household-level choices. My emphasis is on how scholars in the two fields define and delineate decision-making units. I begin with a discussion of problems I encountered in delineating “households” in the three places I have carried out extended fieldwork— Belize, the Peruvian Amazon, and Oaxaca. These case studies illustrate the difficulties that both anthropologists and economists face when attempting to specify who exactly is influencing important household-level decisions.

Rural Belize in the Early 1970s When I went to the Stann Creek District in southern Belize in 1971, my primary goal was to find out how individuals made decisions about how to allocate their labor time among wage labor, cash cropping, and agricultural production for home consumption. I assumed (without much thought) that my fieldwork would focus on the choices men made because rural women in the area rarely participated in wage labor. Although my assumption about the activities of women turned out to be right, I quickly learned that the economic decisions of individual men could not be understood without an examination of what other people in their households were doing. Families supported themselves through a diverse mix of activities that included growing rice, corn, beans, plantains, cassava (manioc), and citrus crops, raising chickens and pigs, fishing, hunting, and doing migratory wage work in construction and on commercial estates. Men considering leaving home to do wage work relied on other members of their households to maintain their fields, tend to chickens, take care of children, prepare food, mend clothes, and do other domestic tasks. Furthermore, the total amount of time men spent on wage work and farming each year depended in part on how much food and income they needed to support their families at a culturally acceptable standard of living. Men with large families needed more income and food (whether grown or purchased) than men with smaller families. Men who were the only wage workers in their households needed higher incomes than those who had adult sons in their households who shared some of the money they earned.

120 Anthropology, Economics, and Choice

It became clear to me that an exclusive focus on individual decision making was misguided. Because the economic activities of various members of a household were interrelated, analyzing the choices any one person made could not be done without looking at the choices of others. More fundamentally, I wondered about the extent to which the economic activities of rural Belizean households could be regarded as the outcomes of independent decisions made by individual members. There was clearly some group decision making in households that affected some activities of individual members. However, neither the household as a group nor any one member could dictate all economic activities. My original assumptions about decision making were not entirely wrong; individuals did have considerable autonomy in allocating their labor time. Despite my eventual recognition of the importance of household decision making, there can be no question that my Belizean research paid far too little attention to the influence of women on economic decisions. This was doubtless partly because the fieldwork was carried out before feminist critiques of male-centered ethnography became widely accepted. But it was also because as a young single male I had easier access to the world of men. Some of the data I collected in Belize came from a household survey I conducted in three villages in the spring of 1972. Although I lived some of the time in one of these villages and knew many people there, I was less familiar with the other two communities. I conducted a survey of households because they were widely regarded as important socioeconomic units by anthropologists, sociologists, historians, and census takers in diverse times and places. I knew, however, that there could be difficult problems defining and delineating households, especially in the Caribbean cultural area where Belize was ordinarily included. I worried about such problems before and during the survey and while attempting to analyze my data after returning from the field. Many of these definitional and boundary problems arose from my efforts to collect a basic piece of information, the number of people occupying particular dwellings. Simply asking how many people lived in a place did not work all that well. Some people spent part of their time in the residence I was censusing and part of their time somewhere else. These included migrant laborers, students at a high school in a nearby town, and people who moved back and forth between their communities and other parts of Belize. There were also many young men (and some not-so-young men) who lived most of the time with their parents but had longtime relationships with women in other dwellings. Many of these

Who Makes Household Economic Decisions? 121

men had children with the women, provided them with money, and spent significant amounts of time at their place. Such arrangements were common among the Creoles and Caribs (Garifuna) in the area, as they are in other Afro-Caribbean communities. They were, however, rare among the Mayas who comprised about a quarter of the population in one of the villages where I conducted the survey. Boundary issues were especially problematic in compounds of two or more dwellings occupied by related families. One house in these compounds usually consisted of an older couple, one or more of their children, and perhaps also grandchildren. Adult sons and daughters lived in nearby houses with their spouses and children.1 There was considerable variation in the extent to which the residents of different dwellings within a compound shared land, helped one another economically, ate together, and in other ways pooled resources. Because such cooperation and sharing was always partial, I struggled in my efforts to determine how many households there were in a compound. In an appendix to my dissertation I note some of these difficulties: My original intention was to have a sample of 60, 20 from each village. Sixty interviews were made, but after they were compiled it was found in a number of cases that errors had been made in regarding certain households as independent economic units. The most common mistake discovered was where men and their fathers lived in different households but worked together in some way. For example, in one case in Silk Grass a father provided the capital and land for a citrus orchard and the son provided the labor. In addition both father and son did a little milpa [shifting cultivation] farming for their own household. In such cases I usually regarded the father and son as belonging to the same economic unit and combined the households in the survey. In addition, a few old men had to be dropped from the sample since although they did do some farming, their household was not really an independent economic unit since the bulk of its income came from gifts from their children. (Chibnik 1975:250)

If I were to rewrite this today, I would phrase my findings differently. What I wrote implies that finding households that were independent economic units was unproblematic in cases other than the ones mentioned. Surely there were many other cases of interdependent households in which some pooling of resources occurred. But I cannot be too critical of these analytic choices I made many years ago. If I wanted to compare

122 Anthropology, Economics, and Choice

household economic strategies, I had to make decisions, however imperfect, about which people belonged to particular households. Although I did not think much about it at the time, in some cases there were also difficult problems regarding all the residents of particular dwellings as being members of resource-pooling households. This analytic quandary arises in many places but is especially important in Belize where unmarried adults (with or without co-resident children) often live in houses with their parents. The degree to which such adults pool resources with their parents may not be very different from situations in which parents and children occupy separate houses in compounds. Nonetheless, in practice I always regarded all the residents of a dwelling as belonging to the same household. Twenty-five years after I conducted this research, Richard Wilk and Steven Miller wrote a perceptive article (1997) in which they defined four categories of household membership in Crooked Tree, an AfroCaribbean Belizean community similar to two of the three villages where I conducted my surveys: 1. Full-time residents sleep in the unit every night, and contribute a major portion of their income to a domestic fund. However, they do not necessarily eat all their meals there. 2. Part-time residents have another residence outside the community, and sleep in their Crooked Tree household a minimum of once or twice a month. They are expected to contribute a smaller portion of their income to the domestic fund, and have the right to become full-time residents in the future . . . 3. Absent members live full-time in another community, but retain the right to return and become residents in the Crooked Tree household . . . 4. Visitors are people who do not sleep in the household, but who have a specific long-term economic and social relationship with it. Most often they are what we called “people who eat but don’t sleep.” . . . Young men or women may also become visitors in a household where they have a girlfriend or boyfriend and/or children. A visiting member may receive other services like clothes washing, use of tools, and storage of personal effects, and they may give contributions of labor, goods, and cash to the household. (67–68)

This is certainly a more accurate way to categorize residents of Belizean households than simply describing them as members or nonmem-

Who Makes Household Economic Decisions? 123

bers. But even if I had collected this information I am not sure how much it would have helped me make statistical comparisons of the economic activities of households of different types. I still would have had to confront the fundamental problem of counting the “number of members” in each household. I suppose that some formula could have been devised to count the part-time, visiting, and absent members as fractional people, but any such methods would have been quite arbitrary. In any case, two households having, say, 5.65 members might differ considerably in their proportion of full-time, part-time, absent, and visiting “residents.” Furthermore, the alternative of abandoning “number of members of a household” as a survey category would have precluded all kinds of useful comparisons.

The Peruvian Amazon in the Mid-1980s When I carried out fieldwork in the Peruvian Amazon in the mid-1980s, I was more sophisticated about the complexities of delineating households as economic units than I had been during my Belize research in the previous decade. In the intervening years, anthropologists and historians had written extensively about household definitions and boundaries (among them Netting, Wilk, and Arnould 1984 and Wachter, Hammel, and Laslett 1978). I had contributed in a small way to this literature, writing an article about the relationship between household composition and agricultural production (Chibnik 1984) and helping to organize a conference in Iowa City in 1983 on “household economies and their transformations” that eventually resulted in an edited collection (Maclachlan 1987). An important part of my research in Peru was an examination of the economic strategies that households of farmers in three floodplain villages used to cope with the annual rise and fall of the Amazon River. To gain an understanding of these strategies I decided to survey every farming household in the villages about their economic activities in 1984 and 1985.2 I was, of course, by now hyperconscious about whether households were really the best economic units with which to conduct my analysis. I was reassured by an offhand comment by my friend Christine Padoch, an anthropologist affiliated with the New York Botanical Garden who was in the midst of long-term research on ribereño cultural ecology. Contrasting the ribereños with the self-identified Indians (indígenas) of the region, Christine remarked that “they [the ribereños] are really classic peasants.” What she meant by this is that ribereños lived in discrete, landowning households that pooled resources, grew most of their food for

124 Anthropology, Economics, and Choice

home consumption, sold some crops, and were integrated into the Peruvian state. As Christine had suggested, delineating household boundaries was much less problematic in Peru than it had been in Belize. In the book I later wrote about my Peruvian research, I devoted only a paragraph to this topic: In many parts of the world, households cannot be clearly delineated, because units of production and consumption may not coincide with one another (Netting et al. 1984; Wilk 1989). Some resources may be shared among people in two or more dwellings, while other resources may not be shared within a dwelling. In the ribereño communities of the Peruvian Amazon, a researcher who assumes that residents of a particular dwelling constitute one and only one independent household can sometimes get a misleading picture of economic relations. Such an assumption is especially problematic for dwellings inhabited by more than two adults. Nonetheless, the residents of most dwellings in riverine villages can be regarded as members of more-or-less discrete households, which are usually units of both production and consumption. Residents eat together, share possessions, and pool the results of their labor. Most dwelling units are economically self-sufficient. They include an adult male and an adult female, produce most of what is consumed, and receive only occasional monetary infusions from relatives living elsewhere. I . . . [therefore] assume [in the rest of the book] that the residents of a dwelling constitute a household. (Chibnik 1994:102–103)

Looking at these printed pages nowadays, I can recall my sigh of relief while writing this paragraph that I would not have to deal with this issue anymore. Nonetheless, the situation was not quite as simple as I suggested. Of the 82 households I surveyed, 33 included more than two adults. Eleven of these households included two or more couples. There were an additional 3 households without any couples, all of which relied heavily on interhousehold pooling of resources. Thus 33 of the 82 dwellings had compositions suggesting that pooling of resources was likely to be only partial; another 3 were clearly not autonomous. My paragraph also overstated the self-sufficiency of ribereño households. In all three communities, cooperative work groups were used in agriculture. These took two forms. In exchange labor groups, farmers worked on each other’s fields on a rotating basis. The food and drink served to guests was not much better than ordinary, and the amount of

Who Makes Household Economic Decisions? 125

work each group member provided for others was reciprocated almost exactly. In festive work parties, farmers issued an invitation to neighbors to work on a particular task such as clearing a field or harvesting. The host family was expected to provide extraordinary food and drink but had little or no obligation to attend future work parties called by guests. Each village also had communal work days when households had to either send a representative or pay a small fine. Typical tasks were cleaning paths, clearing the schoolyard and soccer fields, constructing a community center, and working a small plot of collectively farmed land.

Artisan Communities in Oaxaca since 1994 Although I have spent some time examining economic strategies in woodcarving households in rural Oaxaca (for example, Chibnik 2003:80–93), this has not been a major focus of my fieldwork. My research instead has emphasized socioeconomic relations between wood-carving families and the tourists, store owners, and wholesalers who buy their pieces. My publications on Oaxaca have been more descriptive and less quantitative (and quite likely more readable) than my earlier writings about Belize and Peru. Because of the qualitative nature of much of my Oaxacan fieldwork, I have not had to confront problems of household boundaries as directly as I did in my previous studies that depended more on information collected in surveys. I did, nonetheless, carry out an unsystematic survey of thirty-two artisan households that forced me to consider boundary issues. I have also collected numerous life histories and economic case studies that include information on household composition. If one created a tight–loose scale measuring the extent to which households (however delineated) are independent economic units, Oaxacan artisans would occupy an intermediate position between ribereños and Belizeans. In the two wood-carving communities where I conducted most of my fieldwork, one or more married adult sons or daughters (many with their children) often lived in compounds alongside their parents. Although the younger couples and their parents sometimes lived in separate dwellings, they more commonly occupied rooms in horizontally expanding concrete houses. The economic relationships in these compounds were complex and variable. Younger couples sometimes ate with their parents but usually took meals separately. Grandparents cared for grandchildren; sons and daughters cared for elderly parents. Adult children worked in fields along-

126 Anthropology, Economics, and Choice

side their parents. There were often two or more semi-independent woodcarving workshops in compounds where knives, brushes, and paints were shared. Although the Oaxacan communities where I have conducted research over the past two decades are far from wealthy, households there have a much higher standard of living (however measured) than the Belizeans and Peruvians in my previous field settings. Because the Oaxacans have more money, economic relationships between parents and children (whether living in the same dwelling or different dwellings) are more complicated than those I observed in earlier research. Women and teenage children in contemporary Oaxacan artisan communities have many more opportunities for earning income than their counterparts in rural Belize in the 1970s and the Peruvian Amazon in the 1980s.3 They can paint or carve on a piecework basis for artisans in other households, make tortillas for sale in marketplaces, and commute by bus or collective taxi to jobs in the nearby city of Oaxaca. Women and children also do more than half of the labor on most Oaxacan wood carvings. Money from artisan sales is usually pooled, with wives having considerable control over how it is used. The money that household members earn in other ways (including wage work by adult men) may or may not be shared. Cooperative agricultural work parties composed of members of different households are rare in rural Oaxacan communities. Nonetheless, overall there is more interhousehold cooperation in Oaxaca than in either Belize or the Peruvian Amazon. Communal service on public works (locally called tequio) is similar to analogous arrangements in Peru but more frequent and time-consuming (Cohen 1999:114–118). In addition, many communities (including all the places I did fieldwork) have a local political system in which all adult men (and in some places also adult women) are required to spend years in unpaid governmental positions (Clarke 2000:192–197). Most communities also have guelaguetza, an institutionalized way in which community members can borrow money from relatives and neighbors to pay for food, drink, and music at saints’ day celebrations, weddings, graduation parties, and other ritual events (Cohen 1999:190–191, Stephen 2005:268–274). The biggest difficulty in delineating Oaxacan households is extensive temporary and permanent migration to the United States (Cohen 2004; VanWey, Tucker, and McConnell 2005).4 These migrants send large amounts of money to Oaxaca to their families. Whether to count such migrants as household members in censuses poses difficult analytic prob-

Who Makes Household Economic Decisions? 127

lems because migrants vary greatly in the degree to which they retain economic ties with their families at home. There have been numerous anthropological studies of rural and urban Oaxaca based in part on household surveys. An examination of three such studies shows the variability in the extent to which anthropologists conducting research in Oaxaca consider the complexity of delineating households. Perhaps the most comprehensive socioeconomic survey conducted by anthropologists in Oaxaca took place between 1978 and 1981 among 952 rural households in the Central Valleys. The book based on this research (Cook and Binford 1990) includes twenty-eight informative tables giving data on economic activities in households supporting themselves through combinations of farming, artisanry, and wage work. Despite the authors’ perceptive, sometimes harsh critiques of the shortcomings of previous analyses of the economics of “peasant” households, Cook and Binford do not devote any space at all to discussing how they defined the “households” included in their tables. In Social Inequality in Oaxaca (1991), Arthur Murphy and Alex Stepick show a sophisticated understanding of some of the problems involved with treating households as economic units: We use both the words household and family, because they overlap considerably. To Oaxacans . . . family is the more important culturally. The family is . . . under ideal situations, also coincident with the household. All family members should live within the household, yet for numerous reasons, some imposed and others from individual choice, family members often do not live together. . . . Households are more constant and concrete than families, which may become dispersed, move apart, and later reunite. For these reasons, most of the subsequent discussion refers to households rather than families. (139)

Murphy and Stepick proceed to define a household as “a social group living in a single dwelling area, such as a house lot,” recognizing that “there may be more than one occupied house or structure on a lot, all under the control of a single head of household” (140). They do not indicate how they determined when people living in separate structures should be considered as parts of the same household. Their use of the term “control” is unfortunate, since individual “heads of household” rarely can dictate all decision making even in social groups living under

128 Anthropology, Economics, and Choice

the same roof. Although Murphy and Stepick know this, they nonetheless regard households as more or less unitary decision makers, downplaying the autonomy of individual members: Households, not individuals, are . . . the budgeting units . . . Everyone commits his or her share to the collective and the collective determines how it will be expended. The decisions about how to spend it may be made by a patriarchal autocrat or by democratic consensus; regardless, the household collective forms the basic unit, not the individuals who compose and contribute to it. (Ibid.)

Jeffrey Cohen more recently (2004) surveyed 590 households in twelve communities in the Central Valleys in a study of migration from Oaxaca to the United States. Although Cohen explicitly recognizes definitional problems and the diversity of household composition, he never says how exactly “households” were defined for the purpose of his survey: A household approach meant that we collected data on all members of the domestic unit. In rural Oaxaca, the household could be difficult to define. In general, most rural households (63%) were nuclear units living in independent compounds, quite like their U.S. counterparts. In other words, the household included members of two generations living in a single homestead and pooling the resources and skills of their members. The senior generation consisted of a legally married couple; the junior generation included the offspring of the seniors. However, some households in the central valleys included more than two generations and were better thought of as extended units (35%). Typically, we found extended units organized around a married couple and their children, with the addition of a grandparent. Sometimes a household appeared to be an extended unit but was in fact a series of independent nuclear units that shared a common area or patio. Finally, some households defied classification (2%) and included odd collections of members. (24)

I cannot criticize any of these anthropologists for being imprecise about how they defined “households” for the purposes of their surveys. In the majority of cases, people living under one roof formed a cohesive, relatively independent socioeconomic unit. In more complicated situations, there may often be no good way to make a list of people who indisputably should be regarded as members of the same household.

Who Makes Household Economic Decisions? 129

Table 5.1. Importance of problems in delineating households in three field sites Characteristic

Belize

Peru

Oaxaca

Fluid composition of households

High

Medium

High

Resource pooling among kin in two or more dwellings

High

Low

High

Incomplete resource pooling within dwellings

High

Low

Medium

Autonomy in decision making by individuals within dwellings

Medium

Low

Medium

Resource pooling with nonrelatives

Low

Medium

High

Common Problems In my attempts to delineate households as economic units while conducting fieldwork in varied places, I encountered five general problems: 1. Resource pooling among family members was not restricted to the residents of single dwellings. In Belize and Oaxaca, for example, there were compounds of two or more dwellings where residents shared some, but not all, of their resources. 2. Within dwellings, residents did not completely share resources. 3. Residents of different households had varied degrees of autonomy in their individual decision making. 4. It was difficult to count the number of people in many households because of their fluid composition. This was especially difficult in field settings where temporary migration to other places was common or where people split their time and resource pooling among two or more dwellings. 5. In some places (particularly Peru and Oaxaca), there was significant resource pooling in cooperative work parties and required community service among unrelated people living in different dwellings.

These problems differed in their importance in Belize, the Peruvian Amazon, and Oaxaca (table 5.1). In all three places, however, one or more of these analytic problems prevented me from unequivocally regarding residents of particular dwell-

130 Anthropology, Economics, and Choice

ings as households that pooled resources and made decisions as an economic unit. Regarding dwellings as household economic units was less problematic in some field sites than in others. In the Peruvian Amazon the residents of dwellings pooled most of their resources. Moreover, ribereño communities largely lacked the compounds of related families that were common in my other field sites. In Belize, in contrast, delineating households was difficult because of migration and a social structure in which many men lived in one residence but supported women and children living elsewhere. In Oaxaca resource pooling within dwellings was greater than in Belize but less than in the Peruvian Amazon. In Oaxaca the major problem than delineating households came from extensive temporary migration to the United States. Despite these complications, I found it useful to collect householdlevel economic data in each of my field sites. For the most part, residents of dwellings in Belize, the Peruvian Amazon, and Oaxaca did pool many of their resources. In all three places, the household acted more or less as a group in many important economic decisions.

Anthropological Approaches Before around 1960, anthropologists rarely placed much emphasis on households as economic units (Bender 1967; Netting, Wilk, and Arnould 1984). Some of their best-known comprehensive discussions of social organization (such as Lowie 1920, Murdock 1949) used the term “household” infrequently. These discussions concentrated on institutions such as lineages that ordinarily involved related family members living in more than one dwelling. Economic activities were usually regarded as embedded in social relations rather than being the results of conscious decision making by individuals or households weighing the costs and benefits of alternate actions. Anthropologists’ general lack of interest in decision making during this period (with such notable exceptions as Raymond Firth) therefore prevented them from examining households as unitary economic actors making choices about how to allocate resources. Analyses of economics by anthropologists focused on kinship obligations, production methods, exchange systems, and culturally approved divisions of labor by age and “sex” (now more often called “gender”). The types of social relations in the field settings of most anthropologists prior to 1960 may have led to the development of theoretical approaches that paid relatively little attention to household-level decision

Who Makes Household Economic Decisions? 131

making. The best-known anthropological research during this time took place in non-state societies where economic exchanges were ordinarily structured by kin relationships. Anthropologists therefore concentrated on creating classifications of the bewildering variety of kinship and residence rules governing such exchanges, focusing less on everyday decision making by individuals and households. When anthropologists did conduct fieldwork in state societies, their research often took place in areas where suprahousehold social units such as lineages and castes seemed of more economic importance than households. By the late 1950s many anthropologists were conducting research in areas such as Latin America and western industrial countries where much important economic decision making clearly took place within households. Anthropologists carrying out research in these and other locations began placing more emphasis on the distinction between “households” and “families.” In an introductory textbook, for example, Felix Keesing wrote (1958:271) that families were organized on the basis of kinship, while households were defined by locality. Paul Bohannan argued (1963:86) that “kinship and propinquity do not even belong in the same universe of discourse.” Donald Bender (1967) provided an extensive discussion of the concept of “household” in the American Anthropologist, the flagship journal of the American Anthropological Association. He suggested that this concept referred not only to people occupying a common residence but also to their carrying out what he called “domestic functions” (providing food and shelter and raising children). Anticipating ideas that became widely accepted later, Bender wrote: One is dealing, then, not with two distinct social phenomena—families and households—but with three distinct social phenomena, families, coresidential groups, and domestic functions. All three frequently correspond both ideally and in fact . . . The three also can and sometimes do vary independently. (495)

Household-level economic studies by anthropologists peaked between 1970 and 1990. There were numerous analyses (for example, Barlett 1980) of the economic strategies that households, especially in rural areas of Africa, Asia, and Latin America, used to reduce risk and uncertainty, take advantage of the varied skills of their members, and deploy efficiently their land, labor, and capital. This proliferation of household-level studies was doubtless in part a reaction to economists’ focus on commercial enterprises and individual consumers and producers. Many anthropologists

132 Anthropology, Economics, and Choice

thought that this neglect of household-level decision making downplayed an important aspect of local, regional, and national economies. Economic anthropologists were influenced as well by Chayanov’s ideas in the 1920s, later available in English (1966), about the effects of the demographic composition of a household on its agricultural practices. Chayanov placed particular emphasis on how the consumer/worker ratio of a household changed over time as children were born, grew up, and left. Chayanov’s work carried out in the 1920s and 1930s was brought to the attention of many anthropologists by Marshall Sahlins, who emphasized (1972) how the social organization in particular cultures led households to exchange agricultural products in ways that evened out imbalances brought about by their differing consumer/worker ratios. This was not, however, how most economic anthropologists interpreted Chayanov. Instead, they created models (for example, Durrenberger 1984) and carried out statistical analyses (for example, Chibnik 1984) of how the consumer/worker ratio affected household agricultural behavior in varied times and places. Chayanov’s ideas were based on conditions in Russia that differed from those in many other parts of the world. He assumed that land was freely available, that there was no market for hired labor, and that most production was for home consumption. He also assumed, in contrast to the ways his ideas later were adapted by Sahlins, that households were more or less discrete economic units. Nonetheless, Chayanov’s basic idea that household demographic composition affects agriculture production has been influential in the analysis of economic decision making in parts of the world where these assumptions are clearly inapplicable (Netting 1993:295–319). Anthropologists quickly became aware of the diverse problems associated with regarding households as independent economic units (Netting, Wilk, and Arnould 1984; Wilk 1989; Yanagisako 1979). The major problems they raised have already been mentioned in the discussion of the difficulties delineating household economic units in my field sites. These include drawing boundaries between households when there is some resource pooling among people living in different dwellings, the limited extent of resource pooling among people living in the same dwelling, the fluidity of household membership, and the difficulty in specifying household membership for certain people. Nonetheless, most economic anthropologists in the 1970s and 1980s agreed with Peter Kunstadter that despite the existence of a few ethnographic anomalies, most people in most societies at most times live in households, membership in which

Who Makes Household Economic Decisions? 133

is usually based on kin relationships of marriage and descent, which are simultaneously a combination of dwelling unit, a unit of economic cooperation (at least in distribution and consumption), and the unit within which most reproduction and early childhood socialization takes place. Thus, the kin-based household seems to be an important social unit in all (or almost all) societies, with widely proliferating implications for many important aspects of behavior. (1984:300)

The rise of household studies in economic anthropology eventually led to critiques that altogether rejected the idea of households as unitary decision makers. A surprising, if idiosyncratic, critique came from Eugene Hammel, who had been a leading figure in anthropological and historical studies of households. He argued in 1984 that interminable arguments about the definition and delineation of anything at all took place because “often those engaged in [such] conversation[s] do not know what they are talking about” (40). The household, according to Hammel, was “simply not a good unit of observation.” Hammel’s proposed remedy for the ignorance of his colleagues is, to say the least, opaque: The household in any society, I suggest, is that social group larger than the individual that does not fail to control for its members all those resources that any (adult) member could expect to control for himself. It maximizes size and corporacy simultaneously. (41)

I have no idea how any anthropologist conducting fieldwork could use this definition to delineate households. More enduring critiques came from feminist anthropologists concerned with the relative autonomy and power of women in different cultures (for example, Clark 1989; Guyer 1981, 1988; Lockwood 1989). These anthropologists argued that analyses treating households as economic units made the implicit assumption that one person, usually a man, could to a large extent dictate the economic activities of other members. Their ethnographic research in diverse societies showed convincingly that the other members of households—in particular, adult women— often exerted considerable control over how they allocated their labor and what they did with their earnings. Concentrating on household decision making, in their view, was a fundamental error that obscured power relations between men and women. They argued instead for a focus on the extent to which such power relations allowed autonomous decision making by men and, especially, women.

134 Anthropology, Economics, and Choice

In recent years anthropologists have not written much about the extent to which households can be regarded as unitary, discrete decision-making economic units. There is general agreement that there are difficulties regarding members of single dwellings as isolated resource-pooling entities. Nonetheless, the household—sometimes loosely defined, more often undefined—continues to be an important unit of analysis when economic anthropologists discuss production, exchange, and consumption. Despite the cogency of feminist critiques, most anthropologists agree with Kunstadter’s pragmatic view that residents of dwellings in most societies pool some resources and make certain decisions together. Moreover, the world outside of academia obviously thinks that the household is an important economic category. Governments collect census data at the household level. People discussing the economic choices of their “families” ordinarily are talking about relatives with whom they share living space. As more and more anthropologists conduct work in urban areas in industrial societies, the ways in which they examine household economics are changing. Many people in such field settings earn most of their incomes from jobs outside the home. Anthropologists therefore have been increasingly interested in how members of households in industrial societies make choices between paid work outside the home and unpaid domestic work. Economic anthropologists are now paying less attention to household production strategies than previously and more to decisions about resource pooling and consumption. Overall, contemporary anthropologists are focusing less on household economics than they did in the 1970s and 1980s. They instead are spending more time examining both individual-level decision making and the activities of nonhousehold socioeconomic organizations such as firms, cooperatives, and credit associations.

Economic Approaches There are certain resemblances in the ways that economists and anthropologists have looked at households. Prior to the 1960s relatively little attention in either discipline was given to households as economic units. Economists and anthropologists then began to analyze many important choices made at the household level. In the past three decades feminist scholars in the two fields have emphasized ways individual members of households, especially women, exercise varying degrees of autonomy in their economic decision making.

Who Makes Household Economic Decisions? 135

Despite these similarities, no one would mistake the publications of economists on this topic for those of anthropologists. Economists as a group are much more concerned with devising formal, mathematically tractable models of household-level decision making. Anthropologists, with their “bias towards detailed empiricism” (Chen and Dunn 1996:5), are more likely to examine on-the-ground activities. Economists are not much concerned with intracultural and intercultural variation in household forms; anthropologists usually begin their analyses by describing such variations. Anthropologists write extensively about the fluid membership of households and resource pooling among kin living in different dwellings; economists rarely consider such complexities in their models. The standard models of neoclassical economics treat production and consumption separately. The assumption is that business firms and public enterprises produce and that individuals and households consume. Households are treated as unitary decision makers that purchase everything they consume (Chen and Dunn 1996:12–13, Pollak 2003:122–123). Such models seem clearly inapplicable in the many situations, especially common in parts of the world where anthropologists often do fieldwork, in which households produce goods sold in markets. They also ignore the enormous amounts of unpaid work in households in all societies on tasks such as child care, cooking, and growing subsistence crops. Some neoclassical economists long ago were aware of the difficulties associated with treating households as decision-making units. They wondered how the consumption behavior of a household could be modeled if members differed in their goals (“preferences” in economic jargon). The attempt of the famous economist Paul Samuelson illustrates how the idea of unitary households was nonetheless justified: [We] might try to save the conventional theory by claiming that one titular head has sovereign power within the family and all of its demands reflect his (or her) consistent indifference curves. But as casual anthropologists, we all know how unlikely it is in modern Western culture for one person to “wear the pants.” It is perhaps less unrealistic to adopt the hypothesis of a consistent “family consensus” that represents a meeting of the minds or a compromise between them . . . the family acts as if it were maximizing their joint welfare function. (1956:9–10)

There were some heterodox economists prior to the 1960s who rejected some of the assumptions of the neoclassical models. During the 1920s and 1930s, as noted earlier, Chayanov published extensively on the

136 Anthropology, Economics, and Choice

interrelationship between household production and consumption. At the same time, Hazel Kyrk (1933) and Margaret Reid (1934) introduced household economics as an academic specialization in the United States. They placed particular emphasis on the economic value of time devoted to unpaid household work (Grossbard-Schechtman 2001:104, Ironmonger 2001, Yi 1996). Reid argued (1934:11) that if an activity such as cooking for one’s family might be delegated to a paid worker, it should be regarded as “productive.” Although these scholars’ ideas were influential within the field of home economics, their work was largely ignored by mainstream economists, who continued to portray households as places of only consumption and leisure (Ironmonger 2001). Classical Marxist economists regarded household production as an anomaly in advanced capitalism, thinking that over time women would be drawn into the workforce and become economically independent of men (Hart 1992:117). This view was challenged during the so-called Domestic Labor Debate of the 1970s (Seccombe 1974, Himmelweit and Mohun 1977, Molyneux 1979) by Marxists contending that women’s unpaid housework was useful to capitalism because their domestic activities allowed male laborers to be paid low wages. Both early and later Marxists downplayed the importance of household decisions, seeing such choices as largely determined by suprahousehold socioeconomic relations such as conflicts between workers and the owners of capitalist enterprises. By the 1960s economists were beginning to analyze the implications of the increasing numbers of married women in the labor force in the United States (Mincer 1962, for example) and the importance of unpaid household work in the world economy. The New Household Economics, best known through the work of Gary Becker, was developed in part to provide theoretical tools to examine these issues. Among economists, Becker is celebrated for his observation, not in the least surprising to anthropologists, that the household is a unit of both production and consumption even in advanced capitalist societies. The general aim of the New Household Economics was to provide economic explanations of various decisions made at the household level. This forced Becker to think in more detail than Samuelson had about how to analyze decision-making situations in which household members differed in their preferences. His solution might seem strange to empirically oriented anthropologists. Becker argues (1974, 1981) that as long as a household was headed by an “altruist” (ordinarily assumed to be an adult male) who “cared sufficiently” to transfer some general resources to other members, it would be in the self-interest of the altruist’s spouse and chil-

Who Makes Household Economic Decisions? 137

dren to act in ways that maximized the income and consumption of the household as a whole. Becker does not assume that the altruist exhibits equal concern for all family members; anything less than complete selfishness is sufficient for his theories to work (Pollak 2003:119). The assumption that a household is headed by a semi-benevolent semidictator allowed Becker and other New Household Economists to treat the household as a decision-making unit. Although Becker supports his argument with various mathematical models, some of his assumptions seem clearly unrealistic. How can it be assumed that households have “heads” whose decision making is so influential that all other members will be induced to act in ways that will benefit the household as a whole? We all know of families, for example, in which members who spend too much or get fired from jobs for irresponsible behavior depend on parents, spouses, or siblings for their daily expenses. Despite the “altruistic” support these members receive from the “head” or others, their behavior cannot be described as maximizing family income. And how can Becker assume that a family (household) consists of “a brood of egoistical but rational ‘kids’” (usually including the wife!) and one altruistic parent? Does it not make more sense that most households consist of a number of people whose economic motives are a combination of altruism and self-interest? Finally, Becker’s theories assume without empirical evidence that members of a household completely pool their resources (Pollak 2003:131). The implicit assumption of Becker’s models that households include an altruistic husband and an egoistic (self-interested wife) is understandably offensive to many women (and men) observing actual behavior within households. Becker makes this assumption explicit in articles such as “Altruism, Egoism, and Genetic Fitness: Economics and Sociobiology” (1976), in which he claims that there were “natural” grounds for existing gender-based divisions of labor and authority structures. These sociobiological ideas were used as additional justifications for the creation of models treating households as unitary decision-making units. Becker also presents views unpalatable to many in his book Treatise on the Family (1981), in which he attributes—perhaps reasonably—higher divorce rates, increased cohabitation rates, and lower marriage rates to gains in women’s earning power, leaving some readers with the impression that he thought families would be stronger if women earned less. Unsurprisingly, the number of women economists adopting Becker’s approach to household decision making dropped sharply as the years went by (GrossbardSchechtman 2001).

138 Anthropology, Economics, and Choice

The altruistic model of household decision making was eventually challenged on both theoretical grounds (for example, in Bergstrom 1989 and Bernheim and Stark 1988) and empirical grounds (as in Folbre 1994 and Hoddinott and Haddad 1994). Economists devised models suggesting that under certain conditions Becker’s assumptions about altruism do not guarantee optimal results with respect to household income and consumption. In an influential study aimed at testing ideas about unitary household decision making, Duncan Thomas showed (1990) that when women in Brazilian households controlled a high proportion of family income, child infant mortality rates were lower than when men had more control of intrahousehold income flows. This showed difficulties with Becker’s assumption that any degree of altruism by a single member would lead to economic decisions that benefited the household as a whole. Shelly Lundberg, Robert Pollak, and Terence Wales (1997) found similar difficulties with Becker’s ideas in their analyses of the results of a change in child benefit programs in the United Kingdom in the 1970s. These programs paid money to all families with children. The program had channeled money via the income tax to the primary wage earner in a household, usually at that time the husband. A new program instead paid mothers in cash at the post office. Using data from the British Family Expenditure Survey, the researchers were able to show that this method of payment led to significant changes in the types of goods that households purchased. Many economists reached the conclusion that “family decisions . . . [should be] modeled as negotiations among primarily self-interested individuals” (Bernheim and Stark 1988:1044). Some therefore created “bargaining” models (including Manser and Brown 1980, McElroy and Hornery 1981) that use concepts from game theory. The general idea is that the specific economic situations of households, especially the independent resources available to particular members, influence the extent and types of cooperation between men and women. While this idea is familiar to anthropologists and other social scientists, the language in which cooperative bargaining models are phrased can be difficult for noneconomists. Here is one of many examples that could be cited: Nash bargaining provides the leading solution concept in cooperative bargaining models of marriage. In the Nash bargaining solution, the utilities received by husband and wife depend on their threat point; the higher a spouse’s utility at the threat point, the higher the utility that

Who Makes Household Economic Decisions? 139

spouse will receive in the Nash bargaining solution. This dependence is the critical empirical implication of Nash bargaining models; the couple’s expenditure pattern depends not only on prices and the couple’s total income, but also on determinants of the threat point. (Pollak 2003:128).

Fortunately, readers do not need to know much about game theory and threat points to see that this is clearly not a model of a unitary decisionmaking household. While the cooperative bargaining theorists accept many assumptions of neoclassical economics, feminist economists provide a more radical critique of models of unitary households. Feminist economists agree with cooperative bargaining theorists that individual members of households have separate—sometimes competing—preferences, interests, and resources (Folbre 1986). The two groups of theorists also agree that the relative power of men and women affects the outcomes of bargaining over household decisions. They differ, however, in the ways they analyze power. For the cooperative bargaining theorists, power primarily consists of control over economic resources. They are ordinarily unconcerned with the reasons for power differentials between men and women; these are unquestioned preconditions that provide the starting point for their models. Feminist economists, in contrast, resemble their counterparts in anthropology by focusing also on the ideological, political, and cultural aspects of these power differentials (Hart 1992). In the 1970s and 1980s Marxist-oriented feminist economists (such as Hartmann 1979) carried out “dual systems” analyses in which many societies were regarded as consisting of two separate, interacting structures— patriarchy and capitalism. This interaction resulted in a labor market in which women’s low pay both perpetuated their dependence on men and encouraged them to specialize in housework and child care. Moreover, women’s domestic responsibilities made them less competitive in the labor market. Nancy Folbre argues that this perspective improved our understanding of intrahousehold economic negotiations, saying that “the structure of patriarchy as a system sets the stage for an analysis of bargaining power within a specific household” (1986:250). Many contemporary feminist economists are less willing than dual systems theorists to accept the universality of situations in which patriarchy and capitalism interact to keep women in subordinate positions in households. Nonetheless, there is general agreement among feminist economists that the power differentials that affect intrahousehold bargaining

140 Anthropology, Economics, and Choice

are linked in crucial ways to ideology, culture, and politics. By taking such a holistic position, feminist economists are truly heterodox, differing fundamentally from the dominant practices in their discipline.

Conclusions For many years anthropologists and economists wrote little about household-level decision making. The reasons for this neglect differed in the two disciplines. Anthropologists often failed to distinguish between “households” and “families,” using these terms interchangeably. They also downplayed the significance of decision making in general, seeing economic choices as largely determined by kinship obligations, cultural rules, and religious beliefs. When they did examine decision making, their analyses focused on suprahousehold institutions such as clans, lineages, age groups, and political organizations. Although economists regarded households as sites of consumption, they paid little attention to households as productive entities. This could be seen most clearly in their neglect of unpaid work in households on tasks such as child care, meal preparation, and growing of subsistence crops. Moreover, the methodological individualism of neoclassical economics resulted in models of consumption in which households were treated as undifferentiated units where the conflicting preferences of individual members were ignored. Starting around 1960 both anthropologists and economists began paying more attention to household-level decision making. Anthropologists emphasized the problems associated with regarding members of particular dwellings as self-contained economic entities. They observed that people moved in and out of residences over the course of a year and that there was much resource pooling among inhabitants of different dwellings. Nonetheless, between 1960 and 1985 there was general agreement among anthropologists that family members living together made important group decisions regarding many of the resources at their disposal. The emergence of the New Household Economics in the 1960s, exemplified by the work of Gary Becker, showed the increasing recognition by neoclassical economists that members of households made significant decisions about how to allocate the land, labor, and capital at their disposal. Using elaborate (and in my view unrealistic) theoretical justifications, the models of the New Household Economics treated households as unitary decision makers. The implicit, and sometimes explicit, assumption was that an adult male member of a household had the ultimate say

Who Makes Household Economic Decisions? 141

on how resources were allocated. Although the New Household Economics sometimes considered household decision making in poorer areas of Africa, Asia, and Latin America, its focus was on industrial regions of western countries. Since about 1980 many anthropologists and economists have strongly criticized the idea that households can be regarded as decision-making units. They argue persuasively that household decisions are the outcomes of negotiations among individual members who vary in their goals and influence. They also point out that the extent to which members of households pool different resources varies, allowing individuals to exert considerable autonomy in certain kinds of decisions. The most influential advocates of this position in both disciplines are feminists who emphasize the social conditions that affect the relative power of men and women in different types of household decisions. In economics, collective bargaining theorists also reject the idea of the unitary household. These theorists construct game theory models of the intrahousehold negotiations accompanying varied choices. Although I recognize the cogency of these critiques, I nonetheless think there is much to be gained through careful analyses of householdlevel decision making. It is certainly a mistake to regard households as undifferentiated groups whose members all share common goals. Nonetheless, family members sharing dwellings and pooling many of their resources constitute important socioeconomic units almost everywhere in the world. There is no reason for anthropologists to return to theories about the superorganic and for economists to return to methodological individualism. What is needed, instead, are more detailed ethnographic descriptions and analyses of how particular important household-level choices are made in different times and places.

CHAPTER 6

Is There a Tragedy of the Commons?

Four decades ago an unusual article appeared in one of the world’s most prestigious journals. Most authors of papers in Science present their findings in language difficult for nonspecialists to understand. Garrett Hardin’s “The Tragedy of the Commons” (1968), in contrast, is a breezily written essay about a group of problems said to have no technical solution. Hardin argues that resource destruction and social chaos often occur when individuals have unrestricted access to commonly owned property. His paper inspired a generation of scholars in the biological and social sciences to examine relationships between property rights and resource use. Although these writers usually reject some of Hardin’s assumptions, many would agree that his approach provides a useful starting point for discussions of environmental policies. Perhaps the most influential of these scholars, the political scientist Elinor Ostrom, shared the 2009 Nobel Prize in Economics for her research on how people in societies around the world managed common property. Hardin’s metaphor of the tragedy of the commons is a situation in which cattle raisers herd their animals on a pasture where there are no individual property rights.1 Most herders reason that the benefits of adding another animal to their herd outweigh the costs associated with increased grazing. They do so because the benefits of raising the additional animal accrue only to the herder, while the costs of overgrazing are shared by everyone using the pasture. Moreover, those herders who decide not to raise additional animals because of ecological concerns suffer from overgrazing as much as their more selfish neighbors. The tragedy of the commons arises when individual herders acting in their self-interest add so many animals to their flocks that the pasture is ruined. According to Hardin, many environmental and social problems are

Is There a Tragedy of the Commons? 143

the consequence of individuals overexploiting commonly held resources. He grudgingly concedes that commonly held resources might be feasible when population densities are low (1248). But he says the commons must be abandoned when population increases bring pressures on a resource base. Hardin also recognizes that technological changes are another important source of pressure on commonly held resources. In “The Tragedy of the Commons” Hardin gives several examples of situations in which commons deteriorate because of unrestricted access. These include crowding in national parks, depletion of fish stocks in oceans, and air and water pollution. His emphasis, however, is on the environmental and societal damage caused by overpopulation. Even if one accepts the controversial premise that “overpopulation” is a major cause of diverse world problems, Hardin’s analogy of family planning decisions to cattle raising in pastures is not well developed. For his metaphor to work, one must assume that individuals finding it advantageous to have large families cause problems for particular commonly held resources. Hardin never says what conditions make it advantageous for individuals in certain times and places to have many children. Furthermore, the “commons” allegedly destroyed by overpopulation is never specified. Presumably, Hardin is referring to the overall well-being of the planet. One of the main points of Hardin’s article is that societies must engage in what he calls “coercive” action to deal with potential tragedies of the commons. Hardin contends that moral appeals to individuals to sacrifice for the public good will inevitably fail. What is needed, instead, are restrictions on access to resources. These restrictions can sometimes take the form of private property. But certain resources such as the air or the oceans cannot be privatized. In such situations, societies need to devise incentives, punishments, and laws that prevent individuals, corporations, and governments from overexploiting common property. Hardin is not content to limit his discussion to relatively uncontroversial topics such as attempts to manage access to public parks by raising admission fees or restricting the number of people who can enter each day. He also ominously, if vaguely, says that “the freedom to breed is intolerable” and that steps must be taken to prevent this from happening (1245). The general approach of Hardin’s article is consistent with the rational choice models of economics. Hardin assumes that a limited set of variables influences the actions of individuals who make decisions that maximize their (hazily defined) “utility.” The conclusions that Hardin reaches, however, are quite different from those of advocates of free markets, including many economists, who argue that the results of individual utility

144 Anthropology, Economics, and Choice

maximization are often good for society and who oppose most types of government intervention. Because Hardin sees individual utility maximization in most commons situations as detrimental to the public good, he is a strong advocate of certain types of government interventions. Even though Hardin’s conclusions differ from those of many economists, critiques of his ideas resemble those arguments against rational choice theory that question the autonomy of individual decision makers and emphasize the importance of empirical studies examining the ethnographic complexity of local situations. Hardin assumes that in the absence of government intervention, individuals are the decision makers who matter in the exploitation of common resources. Social scientists in diverse disciplines have convincingly shown, however, that in practice unrestricted access to commons is rare. Almost every society has developed rules and cultural norms that specify who is allowed to use common resources and how these resources can be used. Individuals often do not have the power to destroy commons. Most critiques of Hardin’s ideas are therefore aimed at his assumptions about the units of economic decision making. His many critics say that Hardin fails to understand that in most places supra-individual social groups exert considerable control over access to common resources. Some critics go further and disagree with Hardin on his views about the supposed futility of what he calls “appeals to conscience” (1244). They argue that in many places cultural norms can effectively prevent people from acting in environmentally destructive manners.

Access to Barreales in the Peruvian Amazon The strengths and weaknesses of Hardin’s approach to common property can be seen through an examination of societal regulation of access to seasonal mud flats (barreales) in the Peruvian Amazon in the mid-1980s. Ideas drawn from “The Tragedy of the Commons” aid in the understanding of the land tenure conditions that led to conflict in floodplain areas. But Hardin’s pessimism about the outcome of such conflicts was not warranted. Local communities quickly recognized the need to regulate access to new types of commonly held land and were able to avert a “tragedy” of the type envisioned by Hardin. Visitors to rural areas of the Peruvian Amazon are unlikely to think that a shortage of land is a problem for the region’s residents. Population densities are among the lowest in Peru. Despite the extensive publication

Is There a Tragedy of the Commons? 145

given to deforestation and mining, large tracts of jungle are still sustainably used for foraging, fishing, and farming. Outsiders are therefore often surprised to learn that conflicts over land are common in the Peruvian Amazon because areas vary in their potential for agriculture and forestry and their accessibility to markets. Some of the most valuable land in the region is in fertile floodplains near the city of Iquitos that are used for the cultivation of rice and other crops. Over the past fifty years, there have been numerous disputes over rights to lands in the floodplains. These conflicts have been especially complex because the very best land is in impermanent barreales that form and disappear every year. When I began research in the Iquitos area in the mid-1980s I was immediately intrigued by questions associated with the appearance and disappearance of barreales. How did farmers gain access to new barreales? What did farmers do when their barreales disappeared? What kinds of disputes arose concerning access to new barreales? How were these disputes resolved? The appearance of new barreales posed an annual dilemma for the ribereño inhabitants of floodplain communities that in some ways fit Hardin’s model of the tragedy of the commons. The mud flats were valuable pieces of land that many people desired; no individuals had permanent rights to barreales. If there were unrestricted access to new barreales, serious problems would arise. These problems, however, would not be the overexploitation of a common resource that Hardin emphasized. Although fishers were depleting the region’s marine resources, ribereños lacked the technology and capital to cause permanent damage to agricultural land in floodplains. Instead, the principal difficulties likely to result from unrestricted access to new barreales would be confusion, chaos, and conflict in a free-for-all rush to grab the best pieces of land. Hardin’s observation that restrictions on access to a commons are necessary for the public good when there is sufficient competition for resources clearly applied to land rights in barreales. When I was conducting research in the Peruvian Amazon, I was not explicitly thinking about barreal allocation in terms of the tragedy of the commons. Nonetheless, the questions that Hardin and others had raised about communal property preoccupied me throughout my fieldwork. I examined the systems that the Peruvian state and local communities had devised to regulate access to new barreales, and I asked ribereños how satisfied they were with these rules. I looked at the extent to which local and state systems of regulating access alleviated commons-related problems. Finally, I was interested in a commons-related political economic ques-

146 Anthropology, Economics, and Choice

tion that was not covered by Hardin’s model—the extent to which access to good floodplain land was affected by the political power and social position of the local elites who owned commercial estates along the river. During the 1980s local branches of the Peruvian Ministry of Agriculture had responsibility for assigning land rights in floodplain communities near Iquitos. There were two types of land use permits, titles and certificates. Titles gave the possessor permanent rights to land. The ministry ordinarily gave individual titles only to owners of commercial estates and group titles only to state-recognized “indigenous” communities. (Ribereño villages were only rarely officially recognized as “indigenous.”) Disputes frequently arose when the ministry gave estate owners rights to land used by residents of a ribereño community. Certificates gave ribereños temporary use rights to land. In barreales, certificates for up to 10 hectares were given for one year, beginning in September and ending in August. Holders of a certificate maintained use rights only if their land did not change from one year to the next. As I had speculated when first thinking about the situation, disputes often occurred within and between communities concerning rights to newly formed barreales. Individuals and communities that had lost barreales often thought they should be given priority when allocations of new barreales were made. Such claims, however, were sometimes rejected. The Ministry of Agriculture rarely interfered in villages’ internal allocation decisions. During the mid-1980s many communities allocated barreales in two separate meetings. The first meeting took place before the rivers fell and barreales appeared. A preliminary division of barreales was made for the specific purpose of arranging for certificates to be given to farmers applying for rice credit from the Agrarian Bank. Once the rivers fell and villagers knew where the barreales were, a second meeting was held to make a binding division. The allocation of barreales in the area near the community of Santa Sofía in November 1985 shows how local ecology and politics affected decision making about rights to land in floodplains. When the Amazon River receded in June, new barreales had formed on an island near Santa Sofía. Although parts of the land were claimed by Santa Sofía and the nearby communities of Manatí and Capironal, the land had not been legally divided among the villages. Furthermore, customary use patterns did not provide clear guidelines for the distribution of the new barreales. The island had been farmed in 1985 by inhabitants of the three villages and the employees of a commercial estate near Manatí. Residents of the three ribereño communities all wanted the commercial estate to be prohibited

Is There a Tragedy of the Commons? 147

from using the new barreales but worried that this might not happen because of social ties between the estate owner and government officials. Farmers in Santa Sofía hoped to gain access to about twenty hectares of new land. They had been using about ten hectares of mud bar in a different location. No one farmed more than two hectares at this place, and many residents of Santa Sofía lacked access to any barreales. The older barreal, which was on higher land than the island, was unlikely to disappear in the immediate future. However, this land was regarded as less agriculturally desirable than the new barreales. After consulting with the political leaders of Santa Sofía, Capironal, and Manatí, the official of the Ministry of Agriculture assigned to allocate land on the island devised a plan satisfactory to all three communities. Village leaders were especially pleased that the estate owner was not given access to the new barreales. Santa Sofía’s farmers received the twenty hectares on the island that they had requested, and the thirty hectares of barreales that the community now had were sufficient for its agricultural needs. I attended the meeting in which Santa Sofía allocated its new land. The process was not altogether straightforward. Residents disagreed about whether farmers who were still using land in the old barreal should be given rights in the new barreal. These decisions were eventually made on a case-by-case basis. Another problem involved the allocation of plots among households. Someone suggested that the new barreal be divided into twenty one-hectare plots. However, more than twenty requests for land were made because in some households several members asked for plots.2 The problem was resolved by placing limits on the number of plots given out per household. The complex ways in which new barreales were allocated in the Peruvian Amazon in the mid-1980s illustrate some of the limitations of simplistic approaches to common property. Examinations of situations that may exemplify a “tragedy of the commons” require clear descriptions of the alleged commons and potential users. Such analyses must also say something about the possible costs and benefits of different types of attempts to restrict access to common property to both the public good and diverse potential users. In the Peruvian case described here, the commons can be specified fairly easily as the new barreales that appear in the vicinity of particular communities. It is harder to determine the potential users, who may include residents of several villages and the owners of local commercial estates. Stating the gains and losses for different individuals under alterna-

148 Anthropology, Economics, and Choice

tive methods of allocating new barreales is easy to do in a general way but almost impossible to quantify. Furthermore, there is no value-free way of comparing the costs and benefits to the “public good” of including or excluding the owners of commercial estates from access to new barreales. The version of the tragedy of the commons proposed by Hardin assumes that there are no restrictions to access to communal resources. In the Peruvian Amazon, access to good floodplain land has long been regulated by the state, commercial estates, and local communities. Hardin would approve of and perhaps expect the state’s adjudication of disputes over access to new barreales. His approach, however, does not consider the informal procedures that communities develop that limit access to common resources. In the Iquitos area in the mid-1980s, these took the form of community meetings in which decisions were made about how to allocate new barreales. These decisions were based on shared beliefs that all households in a community had the right to some access to barreales and that mud-flat land should be distributed relatively equally among potential users. As Hardin surely would have observed, these beliefs could be put into practice because the demand for barreales was limited by local agricultural methods and low population densities. Hardin assumed that tragedies of the commons arise because of technological changes and population increases that increase the demand for common property. His approach largely ignores the ways historical circumstances and local political economies determine who are the potential users of communal resources. The resolution of the dispute between the communities near Santa Sofía and the local commercial estate over new barreales in 1985, however, can only be understood in the context of a history of land conflicts. Riverine communities near Iquitos had been attempting for decades to wrest property rights over floodplains from owners of commercial estates that had come into existence during and after a rubber boom that lasted from about 1870 to 1912. A left-leaning military government in Peru had created a political environment during the 1970s in which many communities in jungle regions were able to gain land tenure rights. Although the military government ended in 1980, during the period of my research Amazonian communities were continuing to win some of their struggles for land rights. National and local politics therefore created a situation in which the Ministry of Agriculture was able to give control of new barreales to Santa Sofía and neighboring communities rather than to the owner of the nearby commercial estate. The local methods of allocating barreal land were not long-standing customs.

Is There a Tragedy of the Commons? 149

Instead, they were newly devised systems that developed after the owners of commercial estates lost control of floodplains.

Reactions to Hardin The principal goal of Hardin’s 1968 Science article was to use the tragedy of the commons as metaphor for the causes and consequences of population increases. Hardin’s ideas about overpopulation, which are not all that different from those of Malthus, have had little influence among demographers and other social scientists. His model of the tragedy of the commons, however, was adopted, elaborated, and modified by countless public officials and writers interested in relationships between property rights and environmental conservation. Many policy analysts and scholars concluded that common ownership of property was a major cause of the degradation of natural resources such as fisheries, forests, lakes, and oceans. They contended that the users of commons were caught in a dilemma from which they could not extract themselves without the imposition of rules and regulations by external authorities (Ostrom 1999:493–494). Most economic and ecological anthropologists were uneasy with this enthusiasm for Hardin’s ideas. These empirically oriented researchers emphasized the difference between “open access” and “common property.” Hardin’s model assumed that all members of a community had open access to the use of commonly held resources. Research by anthropologists and other social scientists, however, showed that in many places where there was common property, societies had devised rules and regulations specifying how communal resources could be used (Acheson 1989, McCay and Acheson 1987). These rules and regulations in many cases alleviated or even eliminated potential resource degradation. Anthropologists for the most part focused on providing ethnographic descriptions of the diverse social institutions that have regulated resource use in commonly held property at different times and places. Historians also have presented numerous relevant examples. Although scholars in other disciplines such as political science and resource economics have appreciated the empirical evidence reported by anthropologists and historians, they consider it important to go beyond analyses of particular cases. Social scientists such as Elinor Ostrom (1990, 1999) and Arun Agrawal (2001) have wanted to come up with generalizations about which par-

150 Anthropology, Economics, and Choice

ticular conditions and institutions favor the environmentally sustainable use of public property. They emphasize theory building and multicase comparisons; their evidence includes laboratory experiments (Ostrom 1999:501–508) as well as fieldwork-based studies. Craig Johnson has argued that there is a fundamental tension within scholarship about the commons between “a social science which seeks to build theory on the basis of scientific empiricism and an ethnography which rejects the universalism that underlies the scientific approach” (2004:428). While there certainly are disciplinary differences in research styles and goals, Johnson misunderstands the work of the anthropologists writing about common-property situations. Most of the anthropologists known for the expertise on the commons—in particular Bonnie McCay and James Acheson—fall squarely on the “science” side of the science/ humanities divide in their field. Furthermore, these anthropologists have often worked closely with generalizers from other disciplines, notably Elinor Ostrom. Nonetheless, Johnson makes a useful suggestion about how to look at research on the commons: Two bodies of thoughts compete for a voice in this literature. One, responding to Hardin’s tragedy of the commons, is primarily concerned with the problems of achieving collective action to conserve natural resources which are both depleted and unregulated. A second, influenced by notions of moral economy . . . and entitlement . . . deals with the problem of creating and sustaining resource access for poor and vulnerable groups in society . . . Whereas “collective action scholars” analyze the rules and sanctions that encourage individuals to conserve the commons, “entitlement scholars” emphasize the historical struggles that determine resource access and entitlement, and the ways in which formal and informal rules create and reinforce unequal access to the commons. (408–409)

This distinction draws attention to an aspect of commons scholarship that is only occasionally emphasized. Most writers about the commons are primarily concerned with the effects of property rights and social institutions on the conservation of natural resources. These writers (Johnson’s “collective action scholars”) are also sometimes interested in the extent to which members of a community have equitable access to common resources. They rarely, however, consider in detail the effects of property rights and social institutions on those denied access to common resources.

Is There a Tragedy of the Commons? 151

Such “unentitled” potential resource users are implicitly or explicitly denied community membership. Despite Hardin’s emphases on population growth and state intervention, most analyses of commons situations have focused on how local institutions, rules, and regulations affect resource use. Arun Agrawal persuasively argues that this concentration on the small scale has prevented the emergence of a better understanding of how population, market demand, technology, and state policies interact with local institutional arrangements and resource systems. He points out that what is local is often created by connections with the external and nonlocal environment. Agrawal says an “almost exclusive focus upon the local” has made “[much] work on common property vulnerable to the same criticisms that apply to the work of anthropologists who see their field sites as miniature worlds in themselves, changing only in response to political and economic influences from the outside” (2001:1657). There is disagreement about the extent to which post-Hardin analysts of commons situations accept rational choice models. Craig Johnson asserts that a central assumption of collective action theorists is that “social outcomes can be explained in terms of calculations that individuals make about the perceived costs and benefits of future actions (methodological individualism)” (2004:411). Yet Elinor Ostrom, the best-known collective action scholar, has explicitly argued against Hardin’s idea that resource users are “norm-free maximizers of immediate gains, who will not cooperate to overcome the perverse incentives of dilemma situations in order to increase their own and others’ long-term benefits unless coerced by external authorities” (1999:496). She adopts instead a view of human decision making consistent with the findings of both experiments in cognitive psychology and cultural evolutionary theory: Given that predictions based on the model of a norm-free, myopic, and maximizing individual are not supported, I . . . [argue for] a closely related but alternative conception of human behavior—applicable to resource users and government officials alike. Humans are viewed as fallible, boundedly rational, and norm-using. In complex settings, no one is able to do a complete analysis before actions are taken, but individuals learn from mistakes and are able to craft tools—including rules—to improve the structure of the repetitive situations they face. (Ibid.)

These tools presumably include the creation of social institutions regulating access to common property.

152 Anthropology, Economics, and Choice

Anthropological Approaches to Common Property Some time ago James Acheson was asked to write a chapter on “management of common-property resources” for a textbook in economic anthropology (Plattner 1989). Acheson was an obvious choice for this task. He had published widely on commons situations (Acheson 1975, McCay and Acheson 1987, for example) and was well known for his lucid, intriguing accounts of ways in which Maine lobstermen controlled access to fishing areas. The chapter by Acheson (1989) summarizes the prevailing anthropological take on Hardin’s ideas. Twenty years later, this summary remains the clearest and most comprehensive discussion of anthropological approaches to common property. Near the beginning of his chapter Acheson presents four assumptions that he says are shared by Hardin and the many economists using his theoretical framework: 1. The users of common-property resources are individualistic profit maximizers driven by economic goals to overexploit the resources on which their livelihood depends despite the best interests of society as a whole. 2. The users of these resources have the technical capacity to exceed the biological maximum rate of renewal of the resource . . . 3. Those using common-property resources and the local communities they live in cannot or will not erect effective institutions to protect the resources they live on. 4. The exploitation of collectively owned resources can be halted only by either instituting private property or the government taking action. (357–358)

Acheson then asserts (358) that the primary contribution of anthropologists to theories about common-property resources has been to show that these assumptions either hold true only under certain circumstances or are just plain wrong. Most of the rest of the chapter consists of details about the empirical studies that led Acheson to reach this conclusion. His principal point is that there is not a strong relationship between types of property rights and the likelihood of resource degradation. Acheson does not confine his argument to noting the many ways that different societies alleviate environmental destruction by limiting access to commonly held property.

Is There a Tragedy of the Commons? 153

He also observes (364) that there is much disagreement and little hard evidence about the effectiveness of private property rights in conserving resources. Acheson points out that under certain conditions it makes sense for owners of resources to use them up rapidly, disregarding the future effects on conservation. He gives two examples from his state of Maine. Potato farmers rarely use contour plowing or other strategies to control soil erosion; the forestry practices of big paper companies are influenced more by their need to make a profit in the short term than by the long-term conservation of the forests they use. Because other cases of destruction of privately held resources are frequent and glaringly obvious in industrial societies, it is surprising that they are only occasionally mentioned by Hardin and other proponents of common-property theory. It is as if they had never thought seriously about the activities of farmers, oil companies, and owners of coal mines. Anticipating Agrawal (2001) by more than a decade, Acheson notes (1989:372) that many anthropologists and other social scientists argue that explaining the overexploitation of natural resources solely in terms of property rights is an oversimplification that ignores other aspect of socioeconomic systems. He briefly discusses the effects of population growth (as would Hardin), industrialization, and the expansion of capitalist systems and markets. Acheson concludes the chapter by saying that the work of anthropologists shows that Hardin’s theory of common-property resources needs considerable modification: In virtually all societies, there are controls on access to resources and various kinds of rules and arrangements to limit exploitative activities. Individuals are not allowed to seek their short-term goals at the expense of society. In many industrialized or overpopulated societies, tragedies of the commons do exist, but this is not due to the fact that societies generally abandon their resources to anyone who wants to exploit them. (375–376)

Although Acheson agrees with Hardin that resources can be conserved by privatization and government action, he says that a wide variety of locallevel institutions can also be effective (377). None of these solutions, however, is effective in every case.

154 Anthropology, Economics, and Choice

Maine Lobsters and the Tragedy of the Commons Ocean fishing would appear to be a paradigmatic example of the tragedy of the commons. Because there are no individual rights to pieces of ocean, fishers have no incentive to limit their scale of operation. What one fisher fails to catch often just goes to another fisher. When technology becomes sufficiently efficient that fish stocks can be depleted, a tragedy of the commons may develop. Despite Acheson’s later critiques of common-property theory, he is perhaps best known for earlier research on fishing territories in Maine that supported some of Hardin’s ideas. Acheson conducted detailed studies of the informal arrangements that Maine lobstermen (women rarely fished for lobster) had developed to limit access to particular parts of the ocean. He showed that more restrictive arrangements better preserved the stock of lobsters. In Maine, only those lobstermen recognized as local community members are permitted to fish in parts of the ocean near where they live. Interlopers are initially warned to leave the area, often by having their traps opened. If an outsider persists in fishing, his traps are pulled in and pushed into deep water.3 Because trap cutting is illegal, most enforcement of these social norms is done quietly. State officials are likely to hear only rumors; even the victims usually keep quiet. Acheson made a distinction between “nucleated” and “perimeterdefended” areas of ocean. Nucleated areas, typically found near coastal communities, were not clearly delineated. Although the areas near towns were sharply defended, the “ownership” status of parts of ocean farther away from population centers was hazier. Often there were overlapping zones exploited by lobstermen from two communities. Perimeterdefended areas, in contrast, had well- defined boundaries. All lobsterfishing territories were perimeter-defended before 1920. Along the coast of Maine, later technological changes allowed lobstermen to exploit larger areas than previously, leading them to be less willing and able to defend every part of the ocean they used. The result has been a transition from perimeter-defended to nucleated zones. Nowadays, almost all perimeterdefended zones are near islands. Because of the more rigorous enforcement of boundaries in perimeterdefended zones, Acheson found that they were fished less intensively than nucleated zones. He also found that lobstermen fishing in perimeterdefended areas sometimes voluntarily agreed to take measures to conserve resources. These included limiting the number of traps they used

Is There a Tragedy of the Commons? 155

and imposing a closed season when they did not fish at all. Such voluntary conservation measures were much less common in nucleated areas that included overlapping territories exploited by lobstermen from more than one community. Acheson’s research showed that the reduced fishing effort in perimeterdefended areas had both biological and economic benefits. In perimeterdefended zones, the stock density and average lobster size were greater than in nucleated areas. Men in perimeter-defended areas caught more lobsters per hour. As a result of better stock and easier fishing, lobstermen in perimeter-defended areas did better economically than those in nucleated zones. In his later publications, Acheson emphasizes how the informal sanctions against boundary violations contradicted Hardin’s theory. But when Acheson first wrote about Maine lobstermen (in 1975), he focused on aspects of his findings that were congruent with Hardin’s ideas. Clearly technological improvements had led to the potential for a tragedy of the commons in lobster-fishing areas. Those communities that took strong action to restrict access to these areas did better, as Hardin would have expected, than those that took weaker measures. In his discussions of territorial boundaries in Maine lobster-fishing areas, Acheson downplays two aspects of the situation that are only loosely related to the points that he wants to make. As entitlement scholars would point out, Acheson says little about the ways in which the arrangements he describes harm those who might want to become lobstermen but are not members of any fishing community. Unless such people marry into a community or settle in for a number of years, they are effectively shut out of the lobster economy. More importantly, the informal sanctions against boundary violations do not seem to be preventing the depletion of lobster. As Acheson observes, the technological developments that led to overfishing of lobster also resulted in a transition in most places from perimeter-defended to nucleated zones. Because lobster stocks are declining in nucleated areas, overfishing seems inevitable unless there is more extensive state intervention of the type recommended by Hardin.

Comparative Studies Anthropologists were not the only social scientists carrying out detailed studies of the allocations of commons in the 1970s and 1980s. A bibliography compiled two decades ago (Martin 1989) includes almost 5,000

156 Anthropology, Economics, and Choice

case studies of the use of common property. The disciplines represented included rural sociology, anthropology, history, economics, political science, forestry, and human ecology. Despite this vast amount of information about the use of common property, there had not yet been serious attempts at synthesis (Ostrom 1990:xv). Researchers in anthropology and other disciplines had convincingly demonstrated that Hardin’s model was based on assumptions that did not apply in many places and times. They had, however, rarely attempted to make generalizations about how the use of common resources varied in different circumstances. In order for such generalizations to be made, it would be necessary for someone to make systematic comparisons of the case studies that had already been carried out. Elinor Ostrom’s book Governing the Commons: The Evolution of Institutions for Collective Action (1990) is a thorough, cogent attempt to make such comparisons. Ostrom begins by observing that the tragedy of the commons can best be understood as an example of a recurring issue in social science, the conflict between individual self-interest and the public good (2–7). She discusses three closely related models of such conflicts, noting that they raise free-rider problems of the type discussed by scholars interested in the evolution of altruism:4 [These models] have defined the accepted way of viewing many problems that individuals face when attempting to achieve collective benefits. At the heart of each of these problems is the free-rider problem. Whenever one person cannot be excluded from the benefits that others provide, each person is motivated not to contribute to the joint effort, but to free-ride on the efforts of others. If all participants choose to free-ride, the collective benefit will not be produced. The temptation to free-ride, however, may dominate the decision process, and this all will end up where no one wanted to be. (6)

Ostrom says that such models are interesting and powerful because they capture important aspects of many different problems in diverse settings around the world. But she says that these models are often wrong empirically and lead to bad public policy: Instead of presuming that the individuals sharing a commons are inevitably caught in a trap from which they cannot escape, I argue that the capacity of individuals to extricate themselves from various types of dilemma situations varies from situation to situation. The cases to be discussed in this book illustrate both successful and unsuccessful efforts to

Is There a Tragedy of the Commons? 157

escape tragic outcomes. Instead of basing policy on the presumption that the individuals involved are helpless, I wish to learn more from the experiences of individuals in field settings. Why have some efforts to solve commons problems failed, while others have succeeded? What can we learn from experience that will help stimulate the development and use of a better theory of collective action—one that will identify the key variables that can enhance or detract from the capabilities of individuals to solve problems? (14)

Ostrom is not content to simply say, as many anthropologists have, that the world is more complex than is suggested by models such as the tragedy of the commons. Such observations, she asserts, are not in themselves useful (183). But Ostrom is able in her book to make only some difficultto-test, loosely stated generalizations about when commons management is likely to be successful. She restricts her conclusions to CPRs (commonproperty resources) in remote locations under political regimes that are indifferent as to how these resources are managed. By so doing, Ostrom ignores most of the cases that inspired Hardin and others to develop their theories. She concludes that under the conditions she describes the likelihood of resource users adopting changes in management rules to improve joint welfare will be positively related to the follow characteristics (listed in order of importance): 1. Resource users agree that they will be harmed if they do not adopt changes. 2. Most users will be affected in similar ways by the proposed rule changes. 3. Most users are concerned about the long-term consequences of their actions. 4. Users face relatively low information, transformation, and enforcement costs. 5. Users share generalized norms of reciprocity and trust. 6. The user group is relatively small and stable. (211)

In later publications Ostrom and others (for example, Dietz, Ostrom, and Stern 2003; Ostrom 1999) have identified numerous other variables that they regard as relevant to the use of commons. According to Agrawal (2001:1651), as many as thirty-five factors have been suggested as being critical to the organization, adaptability, and sustainability of common property. Nonetheless, most scholars agree with Agrawal that there is no

158 Anthropology, Economics, and Choice

widely accepted theory about what leads to the effective management of common-property resources (ibid.). Furthermore, few researchers have attempted tests of the relative importance of factors said to be crucial to the sustainability, efficiency, and equity of commons allocations at different places and times. There have been two quite different reactions to the difficulties that social scientists have encountered in their attempts to make generalizations about commons management. Some writers including Agrawal contend that attempts to make generalizations have not been sufficiently systematic in their efforts to identify relevant variables and describe their relationships. Agrawal points out (2001:1653) four general types of variables that affect the success or failure of commons management—characteristics of resources, characteristics of groups depending on resources (emphasized by Ostrom in Governing the Commons), characteristics of institutions managing commons, and characteristics of relationships between resource users and external forces and authorities such as markets, states, and technology. He urges a two-pronged approach to improve our understanding of commons management (1662). First, there need to be theoretically motivated comparative case analyses to identify the most important causal mechanisms and to narrow the range of relevant variables and their interactions. Second, there need to be more large-scale statistical studies to specify the strength of causal relations. Agrawal obviously thinks that Ostrom and her colleagues have been insufficiently rigorous in their search for generalities and have overemphasized the complexity of particular cases. Whatever the merits of Agrawal’s argument, we are no closer today to a general theory of commons management than we were when he wrote his article. Many anthropologists and historians agree with the political scientist Craig Johnson (2004) that searches for generalizations about commons management are likely to have only limited success. Johnson favors the approach of entitlement scholars who understand the degradation of commons resources primarily in terms of historical processes associated with the privatization and commercialization of local resource systems. He contrasts this approach with what he calls the “tragedy school” (420) that explains commons dilemmas and their resolution in terms of individually calculated responses to structural incentives. This strikes me as at best an oversimplification and at worst a misrepresentation of the ideas of Ostrom and other collective action scholars. Nevertheless, Johnson’s emphasis on the intersection of large-scale historical processes and the par-

Is There a Tragedy of the Commons? 159

ticularities of local circumstances is shared by the many scholars whose analyses of commons focus on political economy and political ecology.

The Tragedy of the Commons—Model, Theory, or Metaphor? Some years ago Ottar Brox wrote a provocative, idiosyncratic article (1990) in which he argues that explanatory traditions in different disciplines lead to competing interpretations of theories about common property. Brox is a Norwegian social scientist whose interest in the commons focuses on the causes of overfishing in oceans. His examples are almost entirely restricted to the allocation and use of marine resources. Furthermore, he ignores much relevant work on the commons, notably the research of Ostrom and her colleagues.5 Despite these limitations, Brox’s essay addresses some fundamental issues associated with the use of the concept of the tragedy of the commons. Although Brox recognizes that researchers from many disciplines have worked on commons problems, he frames his article primarily in terms of a contrast between anthropologists and economists. According to Brox, anthropologists conducting studies of fishing regard the tragedy of the commons as a statement about empirical conditions in the world. He notes that when anthropologists such as Acheson present cases that do not fit Hardin’s theory, they think they are falsifying a hypothesis. Economists, however, regard the tragedy of the commons as a model of human behavior that shows what the consequences are if certain conditions hold. For economists, the empirical studies cited by Acheson and others do not falsify Hardin’s theory; instead they are just cases in which the assumptions of the model are not satisfied. Brox attributes this disagreement between anthropologists and economists about the usefulness of the concept of the tragedy of the commons to different scientific traditions in their disciplines. He says that “if natural history is the grandfather of modern social anthropology, economics is the offspring of mathematics and a priori logic” (1990:229). Brox argues that commons models have had analytic utility because they lead researchers to pay attention to aspects of resource use that they have previously neglected to consider. What makes the CPT [common-property theory] so useful in the analyses of specific cases is that it forces us to keep aggregate consequences

160 Anthropology, Economics, and Choice

of individual adaptations and careers constantly in mind. It makes us see, very clearly, the possible tragic end result of natural resources . . . being available free of charge to the exploiter. (234)

Another way of making Brox’s point would be to say that the tragedy of the commons is a metaphor for a diversity of situations. This metaphor has inspired researchers to carefully examine similarities and differences in the use of common resources in diverse times and places. In some circumstances, Hardin’s analogy to a pasture works well; in most cases it is incomplete or misleading. Whether or not the metaphor works in a particular place, taking it seriously forces researchers to examine thoughtfully the environmental effects of the ways in which commons are allocated and used. Brox is clearly right that Hardin’s ideas have inspired much useful research. However, he downplays the damage caused by the popularity of the idea of the tragedy of the commons. Economists are much more influential than anthropologists in the formulation of public policy. Policy makers influenced by the idea of the tragedy of the commons are likely to underestimate the extent to which local institutions can alleviate overexploitation of forests, oceans, and lakes. Furthermore, a focus on individual self-interest as a major cause of the tragedy of the commons can lead to neglect of wider economic, political, and technological changes that result in environmental damage. The consequences in some places have been public policies that emphasize either privatization of resources or the development of structural incentives intended to change the costs and benefits to individuals of environmentally destructive actions. While such policies can be beneficial, they largely ignore inequities in access to resources. They also sometimes result in an unjustified trust in the effectiveness of privatization in alleviating environmental problems.

Politics, Economics, and the Commons The diverse reactions to Ostrom’s 2009 Nobel Prize illustrate the unusual position of commons research along the political spectrum. The award to Ostrom was applauded by both conservatives and liberals. Peter Boettke, a free-market enthusiast who admires Gary Becker and Milton Friedman, has said that Ostrom’s research shows the “wisdom of decentralized government” and demonstrates “the extraordinary capacity of indi-

Is There a Tragedy of the Commons? 161

viduals to solve problems without resorting to top down plans” (2009). Similar comments have been made by members of two conservative think tanks, Michael Spence of the Hoover Institution and Arnold Kling of the Cato Institute (in Schneiderman 2009). Randall Amster, a self-described “peace educator, author, and activist,” has interpreted Ostrom’s research quite differently. Amster blogs on the Huffington Post that “Ostrom’s body of work is inherently radical, demonstrably anti-corporate, and implicitly socialist” (2009). He goes on to say that “in choosing to honor her, the Nobel selection committee has provided an intriguing buttress against the self-referential ‘only money matters’ work of people like Milton Friedman, and has extended its influences into a new generation of economics premised on sustainability and community-based management.” Such remarks may tell us as much about the commentators as they do about Ostrom’s work. The embracing of Ostrom’s ideas by writers of such divergent political perspectives suggests, however, that much commons research cannot be easily fit into conservative (free market) or liberal (public good, collectivist) economic perspectives. Commons research also occupies an ambivalent position within the field of economics. Many economists seemed bemused by Ostrom’s award. Both Paul Krugman, New York Times columnist and 2008 Nobel Laureate, and Steven Levitt, co-author of Freakonomics (Levitt and Dubner 2005) and Superfreakonomics (Levitt and Dubner 2009), confessed that they had never heard of Ostrom. Krugman was one of numerous commentators who saw the prize as an award for the subfield of institutional economics: Neoclassical economics basically assumes that the units of economic decision making are a given, and focuses on how they interact in markets. It’s not much good at explaining the creation of these units—at explaining, in particular, why some activities are carried out by large corporations, while others aren’t . . . a quick scan [of Ostrom’s work] shows why she shared the prize; if the goal is to understand the creation of economic institutions, it’s crucial to be aware that there is more variety in institutions, a wider range of strategies that work, than simply the binary divide between individuals and firms. (2009)

This perspective seems narrower than that of most anthropologists, political scientists, and historians who work on commons questions. Levitt’s comments are perhaps more interesting. He mostly ignores the substance of Ostrom’s work, saying instead that

162 Anthropology, Economics, and Choice

the economics profession is going to hate the prize going to Ostrom even more than Republicans hated the Peace Prize going to Obama. Economists want this to be an economists’ prize (after all, economists are selfinterested). This award demonstrates, in a way that no previous prize has, that the prize is moving toward a Nobel in Social Science, not a Nobel in Economics. (2009)

This comment must be considered in the context of previous Nobels in economics given to mathematicians (Robert Aumann, Leonid Kantorovich, John Nash), psychologists (Daniel Kahneman), and lawyers (Fredrich Hayek, Leonid Hurwicz). Clearly for Levitt and many other economists, Ostrom’s research seemed further from the mainstream than that of these other scholars. There are several reasons Ostrom’s work is unfamiliar and a bit alien to many mainstream economists. Ostrom makes relatively little use of formal models and quantitative methods. Even though Ostrom has conducted numerous economic experiments, she is best known for her detailed, descriptive accounts of particular commons arrangements. While many mainstream economists carry out empirical studies, their research is ordinarily more quantitative, more model-driven, and much less ethnographic than the work of Ostrom. Furthermore, the informal social arrangements and institutions studied by Ostrom, often in nonwestern settings, differ greatly from the firms, markets, consumers, investors, and entrepreneurs in western countries that comprise most of the subject matter of contemporary economics. Ostrom’s Nobel Prize was celebrated by many members of the listserv of the Society for Economic Anthropology (SEA). The reaction of Carolyn Lesorogol of Washington University, St. Louis, is typical: I think economic anthropologists should be very pleased with this award and its implicit challenge to standard economic approaches . . . Lin Ostrom’s work on common property, cooperation, and trust is a fundamental contribution to studies of institutions and is highly relevant for anthropologists similarly interested in how ordinary people organize themselves to manage their shared resources. (2009)

Other SEA members also seemed particularly pleased by what they saw (perhaps mistakenly) as a rebuke to the model building of many economists.

Is There a Tragedy of the Commons? 163

Conclusions Elinor Ostrom is only one of the many scholars who have carried on research of commons issues in the four decades since Hardin’s article appeared in Science. Her work, however, exemplifies much commons research in its attention to historical and ethnographic details. Hardin’s original formulation of the tragedy of the commons could have resulted in the development of complex mathematical models of the sort found in economists’ analyses of risk, rational choice, and household decision making. His ideas might also have led to extensive testing of hypotheses in laboratory experiments such as those common in studies of subjective probabilities and cooperation. Although some researchers examining commons questions have used formal models and experiments, most have emphasized the findings of empirical studies in field settings. Much of this research has been carried out by noneconomists, including many anthropologists. In most of the earlier chapters of this book, the dominant, model-driven approaches by economists to various issues related to decision making are contrasted with less influential, heterodox ethnographic approaches. The situation is reversed in commons research, which is dominated by ethnographic studies.

Conclusion

Social scientists examining choice vary in their positions along what might be roughly labeled a science-humanities continuum. Economists and experimental psychologists are at the science end of this continuum. Their attempts to construct theories about choice involve models that specify how a few key variables are related to one another. Although economists usually stress the testability of the models they create, in practice their work often concentrates on exploring the mathematical implications of different assumptions about relationships among variables. Experimental psychologists are less concerned about the mathematical complexities of their models; their focus is on learning how people make choices in controlled laboratory conditions. Both economists and experimental psychologists know that their models ignore many factors that affect both the possible choices available to decision makers and the ways in which particular options are selected. Furthermore, the decision-making experiments of psychologists and behavioral economists take place in laboratory settings that usually differ in significant ways from real-world conditions. Nonetheless, these social scientists argue that the careful construction and testing of theories requires models and experiments that are intentional simplifications of a complex world. Many anthropologists and historians are at the humanities end of a science-humanities continuum; some would reject the term “social science” as a label for the kind of research they carry out. When these scholars examine decisions, their focus is often on the historical changes, cultural norms, and socioeconomic institutions that constrain the choices possible for different groups of people at particular places and times. Most geographers, political scientists, and sociologists are somewhere between

Conclusion 165

economists and historians on a science-humanities continuum. Scholars analyzing choice in these fields usually attempt to be “scientific” but reject the radically simplified, nondescriptive approaches of many economists and experimental psychologists. These broad characterizations of entire fields of study ignore significant intradisciplinary variation. Feminist, Marxist, and other heterodox economists often eschew mathematical models and focus instead on history, social institutions, and culture. Freudian psychologists neither construct testable hypotheses nor run experiments in laboratories. Formalist economic anthropologists create elaborate models of choice; anthropologists influenced by evolutionary psychology run experimental games in nonwestern settings in which they examine the effects of a few carefully defined variables on decision making. Although I have examined here ideas about decision making from diverse natural and social sciences, I have concentrated on contrasts between the approaches of economists and anthropologists. The usual approaches to decision making in the two fields are strikingly different. Economists often scorn the ways that anthropologists analyze choice as meandering, unscientific, anecdotal, and untestable. Anthropologists can be equally contemptuous of economists, thinking that their models ignore history, culture, ethnography, and the real world. To summarize the differences between the two fields in their approaches to decision making, it is useful to list some of the continuums along which they vary, for each of which economists are ordinarily on the “scientific” end and anthropologists on the “humanistic” end: • Mathematical models important, little emphasis on ethnography; ethnography important, little emphasis on mathematical models • Theoretical parsimony and few variables; holistic descriptions and many variables • Decisions analyzed as being made by indivisible, autonomous units with specified goals; decisions analyzed as the end result of negotiations among individuals and groups with differing goals • “Economy” and “society” treated as separate spheres of analysis; economy treated as being embedded in society and culture • History and evolution of economic institutions irrelevant to analysis; history and evolution of economic institutions part of analysis • Laboratory experiments regarded as informative about behavior in the wider world; skepticism about usefulness of laboratory experiments

166 Anthropology, Economics, and Choice

Such abstract comparisons of the mathematical models of economists and the ethnographic descriptions of anthropologists do not tell us anything about the relative usefulness of ideas from these fields in assessing diverse types of decisions. That is why this book examines in detail the approaches of economists and anthropologists to five classic issues in decision making. The methods and assumptions of mainstream economics, exemplified by rational choice theory, seem more useful for the analysis of some of these problems than they are for others. For one of the issues I look at, the value of unpaid work, the approaches of conventional economists are dominant; critiques from scholars in other disciplines have had little impact. For the other problems examined here, critiques of mainstream economics by anthropologists, cognitive psychologists, political scientists, and sociologists have challenged important aspects of rational choice approaches.

Choices between Paid and Unpaid Work People in market societies often must choose between paid and unpaid work. Because of difficulties associated with assigning monetary costs and benefits to unpaid work, however, the statistics and analyses of economists often ignore subsistence production, household labor, and community service. Heterodox economists, anthropologists, and feminist scholars in diverse disciplines have therefore cogently criticized conventional economic analyses for downplaying the contributions of unpaid labor to society. Although there have been numerous calculations of the monetary “values” of subsistence production and unpaid labor in market societies, these can be lumped fairly easily into two principal methods. Both are consistent with the ideas of mainstream economics and rational choice theory. The first is to calculate an opportunity cost, the monetary returns from alternative paid work. The second is to calculate a replacement value, the cost of hiring someone to carry out an unpaid activity. These methods can result in very different estimates of the monetary value of subsistence crops and unpaid work. Moreover, there are certain empirical and logical difficulties associated with calculations of both opportunity costs and replacement values. Neither method, for example, pays much attention to what the decision makers themselves see as the costs and benefits of different types of work. Nonetheless, few people who have

Conclusion 167

thought seriously about this issue have come up with plausible alternative ways of measuring the value of subsistence production and unpaid work.

Risk, Uncertainty, and Decision Making The analysis of decision making in risky situations is a basic part of rational choice theory. Economists have developed elaborate mathematical models of risk that have been used as guidelines for stock market investments and other resource allocation decisions. These models, however, often make unrealistic assumptions about the availability of information and the ways in which most humans make choices. Although the shortcomings of rational choice approaches to risk have long been evident to cognitive psychologists and anthropologists, such approaches have continued to be taught in textbooks in business and economics and used as guidelines for decision making. The dramatic failures of arcane risk-related investment tools such as derivatives in recent years have led increasing numbers of scholars to question conventional economic approaches to decision making in risky situations. Inspired by the research of cognitive psychologists such as Kahneman and Tversky, behavioral economists including Shiller (2000) have urged that more attention be given to how people actually behave when confronted with risky decisions. Economists and psychologists have run countless experiments in laboratories aimed at determining how people make choices when they cannot be certain of the outcomes of their decisions. The fundamental dilemma confronting these economists and psychologists is the lack of a mathematically tractable conceptual framework for dealing with uncertain situations in which decision makers have little idea about how to estimate the probabilities of alternative outcomes. Both economists constructing models of choice and cognitive psychologists running experiments therefore often act as if the distinction between risk and uncertainty is of relatively little importance. The shortcomings of this approach are obvious during times such as the present, when no one can be confident about the future performance of many components of the global economy. Although some famous economists wrote long ago (Keynes 1937, Knight 1921) about decision making under uncertainty, as a group anthropologists have been more willing to emphasize limitations of con-

168 Anthropology, Economics, and Choice

ventional assumptions about choices in risky situations. Frank Cancian (1979) and other anthropologists have created models of decision making incorporating the risk/uncertainty distinction. Most, however, have confined their research to descriptions of the strategies that individuals and groups have developed in different times and places to cope with multiple risks and uncertainties. These ethnographic descriptions are an important complement to the models of economists and the experiments of psychologists.

Choices about Cooperation The fundamental assumption of rational choice theory is that individuals and groups making decisions act in ways to further their self-interest. A difficult theoretical question confronting economists therefore is how to analyze the many situations in which the self-interest of individuals differs from the interests of the groups and societies they belong to. Sociocultural anthropologists, unlike economists, do not ordinarily regard differences between individual and group goals as a challenge to their underlying theoretical assumptions. Anthropologists (along with sociologists and political scientists) have long emphasized how societies are organized to prevent individuals from behaving in ways—including the excessive pursuit of self-interest—that are harmful to their groups. In introductory courses, anthropologists teach how mores against selfishness in unstratified societies are enforced via ridicule, gossip, and the threat of banishment; they also point out how in state societies institutions such as laws, courts, police, and jails prevent the untrammeled pursuit of self-interest. Such legal and social sanctions might with difficulty be fit into rational choice theories by arguing that individuals pursue their self-interest, given the constraints of the particular society in which they live. Such an ahistorical approach, however, begs the question of how and why these legal and social sanctions evolved. Furthermore, much of culture acts to make individuals want to behave much of the time in ways that are useful to society—that is, to behave morally—even if this involves deferred gratification and sacrifice. This cultural morality transmitted across generations often involves sharing resources and treating others fairly. Much has been written in diverse social and natural sciences about the advantages and disadvantages of cooperation in particular situations among humans and other species. Where cooperation is mutually beneficial for two or more parties, the rational choice models of economists and

Conclusion 169

the evolutionary theories of biologists can be useful in understanding why certain choices are made. Much of anthropological theory and ethnography also concerns the reasons for and the nature of cooperation in such circumstances. Evolutionary biologists interested in altruism, economists using game theory, and anthropologists examining decision making, however, are less certain why members of human and other species cooperate with one another when this incurs costs for some parties. These scholars are especially interested in situations in which individuals demonstrate their willingness to cooperate with others by altruistically incurring costs in order to punish selfish behavior that is harmful for their group. Behavioral economists and experimental psychologists have attempted to examine altruism by running experiments, most notably the Ultimatum Game, in controlled situations in laboratories. Most of these experiments have been conducted in western industrial countries. A few heterodox anthropologists have attempted to examine their cross-cultural validity by running similar experiments in nonwestern, nonindustrial settings. The anthropologists conducting these experimental games all have carried out detailed ethnographic studies among the participating groups. They have the ambitious goal of combining the rigorous model building and experimentation of evolutionary biology, economics, and psychology with the cross-cultural perspective and ethnography of anthropology. The reception to these cross-cultural experiments illustrates well how unenthusiastic most anthropologists are about this kind of research. The results of the experiments have been widely discussed among economists and psychologists and given extensive publicity in the popular media. There has been, however, much less interest in these experiments among economic anthropologists, whose ethnographic research is often directly concerned with relevant issues in decision making. The reason for this lack of enthusiasm, I think, is that the experiments—notwithstanding the researchers’ cultural knowledge—seem “unanthropological” in their assumption that research in controlled quasi-laboratory settings can shed light on decision making in the wider world.

Problems with Methodological Individualism The methodological individualism characteristic of economics causes certain difficulties in analyses of the value of unpaid work, the effects of risk and uncertainty on decision making, and choices about whether to cooperate with others. Rational choice theory requires that a unit of some

170 Anthropology, Economics, and Choice

sort (such as an individual, household, or business firm) attempts to maximize its “utility” when making decisions. Economists only occasionally worry about whether this assumption of a unitary decision maker makes sense. Yet choices such as growing new cash crops, undertaking dangerous border crossings in search of higher wages, and working with one’s neighbors in a cooperative cannot be neatly pigeonholed as either individual or group decisions. They are often the outcomes of discussions and negotiations. In some cases, individuals make decisions that are disapproved by most other members of their group. In other situations, group norms and social institutions constrain the types of decisions that individuals are able to make. Methodological individualism is an especially serious impediment to the understanding of household decision making. Economists typically regard households as unitary decision makers that may make important choices about consumption and production. But households usually consist of several members, whose interests and goals can differ. Furthermore, in many places households can be difficult to delineate because individuals in different dwellings cooperate in some but not all economic activities. Even when economists like Gary Becker recognize these difficulties, they ordinarily resolve them by assuming that one member of a (supposedly unproblematically delineated) household has the power to compel other members to act in the interest of all residents. This lack of realism has often led anthropologists and sociologists to be skeptical of economic analyses of household decision making. The question of units of analysis is also key to the understanding of tragedies of the commons in which individuals have unrestricted access to publicly held resources. In influential publications starting in the late 1960s, Garrett Hardin, a biologist borrowing methods from economics, pointed out that in these situations decision-making units such as individuals, households, and businesses pursuing their self-interests often destroy common resources. His solutions to such dilemmas, all consistent with rational choice theory, include privatizing resources where possible and restricting individual access to common property via laws and incentives. Scholars from diverse disciplines, including anthropology, have shown that the assumptions about unitary decision makers lead to mistaken depictions of how societies around the world have dealt with commons issues. In very few circumstances do individuals actually have untrammeled access to public resources. Instead, in most places supra-individual social groups exert substantial but not complete control over such access.

Conclusion 171

An analysis of commons situations therefore requires ethnographic descriptions of both the extent to which individuals have access to common resources and the ways in which customs, social institutions, and laws constrain such access. Perhaps in part because of the limitations of methodological individualism, economists have not been particularly prominent in studies of the use of common resources. Even when the Nobel Prize in Economics was given in 2009 to a scholar conducting research on commons, the winner was a political scientist, Elinor Ostrom, whose ethnographically oriented work explicitly challenged some of the assumptions of mainstream economics about decision-making units.

Final Thoughts My principal aim in this book has been to contrast economic and anthropological approaches to the analysis of decision making. Both approaches provide useful insights about decision making. Economists are able to isolate important variables relevant to choice and to state clearly how these variables might be related. Anthropologists provide cross- cultural data that can be used to test economic theories; they also describe well the complexities of many decision-making situations. There are also shortcomings to both approaches. Economists too often oversimplify, make unrealistic assumptions, ignore history and culture, and downplay factors influencing choice that are difficult to quantify. Anthropologists too often express ideas imprecisely, mindlessly oppose quantification, and refuse to generalize. Much of this book has consisted of critiques of economic approaches to choice. There are several reasons I have focused more on the shortcomings of economists’ analyses than on those of anthropologists. Economics is a much more influential discipline. Economists are important policy makers in national, regional, and local governments; anthropologists rarely hold such positions. Newspapers, bookstores, and libraries are filled with economists’ explanations of the causes of assorted financial crises; anthropologists’ views on such matters receive little publicity. Furthermore, many economists openly disdain or intentionally ignore anthropological ways of looking at decision making. Moreover, economists’ willingness to make explicit assumptions makes it easy to see exactly what they are leaving out of their analyses. I am not suggesting that anthropology can or should replace economics as the principal method for thinking about decision making. Any anthro-

172 Anthropology, Economics, and Choice

pologist reading this book will notice that I totally ignore or mention only in passing whole schools of thought in our discipline that are devoted in part to a consideration of economic institutions. Much of the writing by scholars adopting these theoretical frameworks is, in my view, dense, opaque, and of little practical use. In the language of some such scholars, I have not made the effort here to “unpack” or “interrogate” these approaches. The ideas of these anthropologists have had practically no impact outside of academia. My advocacy of anthropological approaches is not an argument in favor of our grand theoretical proclamations; instead it is a plea for more attention to ethnography. The examination here of five important issues in the analysis of decision making suggests that the methods of economics are of limited use in the understanding of key aspects of decision making. These methods must be complemented with more descriptive approaches that consider the context within which choices are made. If economics and anthropology were equally influential disciplines, this would be a banal conclusion emptily asserting the advantages of diverse perspectives. But in the world as it is, the virtues of ethnography seem worth mentioning.

Notes

Introduction 1. University of Chicago, “The Sveriges Riksbank (Bank of Sweden) Prize in Economic Sciences in Memory of Alfred Nobel for 1992,” press release, October 13, 1992. http://home.uchicago.edu/~gbecker/Nobel/nobel.html. 2. Scott Cook argues (2004:87–120) that the formalist-substantivist debate remains important. 3. As will be seen in later chapters, behavioral economists and experimental psychologists have devised various methods in their attempts to make such comparisons. The usual method is to ask a series of hypothetical questions, for example, whether people would prefer a certain amount of money but no championship or a championship and no money. 4. All names of people discussed in this example are pseudonyms. 5. A good example of such work is an essay by Michael Hudson (2010) entitled “The Use and Abuse of Mathematical Economics.” Hudson’s scathing critiques of his professional colleagues echo much of what I say more gently in this book. For example, Hudson argues (ibid:14) that “lacking empirical testing and measurement, economics narrows into a mock-science of abstract assumptions without much regard as to whether its axioms are historically grounded.” He later says (ibid:17): The word “exogenous” is heard so often these days . . . that one wonders what is left in economics proper. At issue for a more relevant empirical economics are the dynamics of social history, political institutions and the environment, not just the mechanics of supply and demand.

Chapter 1 1. The ways I picked these textbooks might be of minor interest to students of decision making. Samuelson’s text (formerly sole-authored) is doubtlessly the

174 Notes to pages 37–82

most famous in the field. Mankiw’s blog was familiar to me. I had not previously known about the McConnell and Brue book but learned that it was the current best-seller. Because the most recent editions of these textbooks were very expensive (more than $100 apiece) and unavailable at my university library, I decided to buy earlier versions via the Internet, thinking that the sections I was interested in had probably not changed much. Rational choice theory can deal with this latter choice easily. The reasons beyond my selection of these particular three textbooks, however, can only with difficulty be understood by mainstream economic theory. 2. This argument is made at length in one of my first articles, Chibnik 1981. Chapter 2 1. More precisely, the economic idea of “rationality” assumes that individuals of groups attempt to either maximize or minimize one particular goal such as money or time. These maximization or minimization assumptions often include certain constraints. For example, an economist might try to determine the best way to maximize income from sales with no more than X number of employees working no more than Y hours a week. 2. British Honduras was not on the metric system at the time of my research. Nowadays, most Belizeans continue to use “English” units of measurement. 3. This distinction has been made by diverse scholars including Aristotle, Adam Smith, and Marx. See Hornborg 2001:89–90, Sahlins 1972:83–85. 4. Chayanov’s advocacy of peasant farming and opposition to the collectivization of agriculture eventually resulted in his death in 1937 in one of Stalin’s purges. 5. Contemporary scholars generally agree that many of Chayanov’s ideas can be phrased in the language of classical economics. 6. Delang asserts (without giving reasons) that different techniques of evaluation would be needed to estimate the economic value of other types of NTFPs (2006:66). Chapter 3 1. This credit program ended in the early 1990s. 2. If the chances of a farmer defaulting in any one year were 1 in 10, the probability that a farmer borrowing in 10 different years defaults at least once is 1 − (9/10)10 = 65 percent. The actual probability of default was greater than 1 in 10 in most years and in some years was as high as 1 in 3. Making the unrealistic assumption that it is exactly 1 in 5 every year, the probability that a farmer borrowing in 10 different years would default is 1 − (4/5)10 = 89 percent. 3. In The Black Swan: The Importance of the Highly Improbable (2007), Nassim Nicholas Taleb argues that investors often suffer disastrous losses by ignoring the possibility of unlikely events. This cranky, eccentric, insightful book was widely cited during the stock market crash of 2008–2009. 4. I do not know how important women were in Paez agriculture at the time

Notes to pages 85–126 175

Ortiz did her research. Certainly many of the decisions she describes were made by family groups in which women had some influence. Ortiz’s use of the pronoun “he” to describe a farmer was, of course, conventional in the 1970s. 5. I have argued (in Chibnik 1981) that much of cultural evolution consists of information gathering and small-scale experimentation resulting in the development of such heuristics (“cultural rules”). Chapter 4 1. This section about the ideas of evolutionary biologists is largely based on the lucid discussions in Dugatkin 2006. 2. Hamilton at one point pursued the possibility of studying social anthropology at Cambridge with Edmund Leach along with genetics. Hamilton’s approach to altruism was unpalatable to Leach, who discouraged the prospective student; the geneticists were also unenthusiastic about the idea. Dugatkin 2006:90–91. 3. Wilson later came to have doubts about the explanatory power of kin selection. Wilson and Hölldobler 2005. 4. In public goods games, each participant starts with a certain amount of money. Players may contribute, if they wish, some or all of their money to a group project. For each unit of money X contributed to the group project, every group member (including those contributing nothing) earns M (where M is greater than X) units of money. For example, each participant might start off with $10. For every dollar contributed to the group project, each member might receive $2. Let us say there are three participants—A, B, and C. A contributes $4, B $2, and C nothing. A gets $10 − $4 + (6 × $2) = $18; B gets $10 − $2 + (6 × $2) = $20, and C gets $10 + (6 × $2) = $22. The least generous participant ends up with the most money. 5. Note that this case is an exception to the project’s generalization that members of foraging societies tend not to be generous in the Ultimatum Game. Chapter 5 1. Although many of these “spouses” were not legally married, they were involved in long-term relationships and were regarded by the community as being married couples. 2. I ended up surveying 82 of the 84 farming households in the three communities. 3. The economic situations in Belize and the Peruvian Amazon have obviously changed since I conducted my fieldwork. The standard of living in the part of Belize where I conducted research in the 1970s is comparable nowadays to that in contemporary Oaxacan wood-carving communities. The area of the Peruvian Amazon where I did fieldwork in the 1980s, however, is still much poorer (by any measure of well-being) than contemporary Oaxaca and Belize. 4. Migration rates from Belize to the United States nowadays are comparable

176 Notes to pages 142–159

to those in Oaxaca. There is still relatively little migration from the Peruvian Amazon to other countries; most migration from the rural villages where I did research is to the city of Iquitos. Chapter 6 1. Hardin’s metaphor of a pasture was taken from lectures by William Forster Lloyd at Oxford University in the 1830s. Lloyd 1977. 2. This is obviously related to the discussion in the previous chapter about the extent to which households can be regarded as economic units. 3. About once a decade, small incidents escalate into “lobster wars” in which dozens of men attack one another’s equipment. Acheson 1979:271. 4. The three models discussed by Ostrom are the tragedy of the commons, the prisoner’s dilemma, and the logic of collective action. The prisoner’s dilemma is a situation when two people have been arrested and accused of committing a crime together. If neither confesses, both receive short sentences. If both confess, each receives a medium-length sentence. If only one confesses, the betrayer goes free and the mute accomplice receives a long sentence. Under these circumstances, the best game theory strategy for each individual is to confess, leading to an outcome that is worse than if neither confessed. The prisoner’s dilemma has often been used as a metaphor for situations in which individuals pursuing their own selfinterests cause problems for larger groups. The logic of collective action refers to the ideas of Mancur Olson (1965) challenging the viewpoint that individuals with common interests would voluntarily act to further these interests. Olson anticipated Hardin in arguing that someone who cannot be excluded from obtaining the benefits of a collective good has little incentive to contribute voluntarily to the production of that good. 5. Even though Ostrom’s best-known publications appeared after Brox wrote his article, her work was already influential among commons researchers during the 1980s.

References

Acheson, James

1975 The Lobster Fiefs: Economic and Ecological Effects of Territoriality in the Maine Lobster Industry. Human Ecology 3 (3): 183–207. 1979 Variations in Traditional Inshore Fishing Rights in Maine Lobstering Communities. In North Atlantic Maritime Cultures, ed. Raoul Andersen, pp. 253–276. The Hague: Mouton. 1987 The Lobster Fiefs Revisited. In The Question of the Commons: The Culture and Ecology of Communal Resources, ed. Bonnie McCay and James Acheson, pp. 37–65. Tucson: University of Arizona Press. 1989 Management of Common-Property Resources. In Economic Anthropology, ed. Stuart Plattner, pp. 351–378. Stanford, California: Stanford University Press. Adamowicz, W., T. Beckley, D. H. MacDonald, L. Just, M. Luckert, E. Murray, and W. Phillips

1998 In Search of Forest Resource Value of Indigenous Peoples: Are Nonmarket Valuation Techniques Applicable? Society and Natural Resources 11 (1): 51–66. Agrawal, Arun

2001 Common Property Institutions and Sustainable Governance of Resources. World Development 29 (1): 1649–1672. Akerlof, George

1970 The Market for “Lemons”: Quality, Uncertainty, and the Market Mechanism. Quarterly Journal of Economics 84 (3): 488–500. Alvard, Michael

2004 The Ultimatum Game, Fairness, and Cooperation among Big Game Hunters. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 413–435. Amster, Richard

2009 Why Ostrom’s Nobel Is Even More Shocking than Obama’s. Huffington Post, October 14. http://www.huffingtonpost.com/randall-amster/whyostroms-nobel-is-even_b_320172.html, downloaded November 10, 2009.

178 Anthropology, Economics, and Choice Ashcraft, Norman

1968 Land-Use and Trade: The Processes of Economic Change in British Honduras. Ph.D. dissertation, Brandeis University. 1973 Colonialism and Underdevelopment: Processes of Political Economic Change in British Honduras. New York: Teachers College Press. Axelrod, Robert

1984 The Evolution of Cooperation. New York: Basic Books. Axelrod, Robert, and William Hamilton

1981 The Evolution of Cooperation. Science 211: 1390–1396. Baksh, Michael, and Allen Johnson

1990 Insurance Policies among the Machiguenga: An Ethnographic Analysis of Risk Management in a Non-Western Society. In Risk and Uncertainty in Tribal and Peasant Economies, ed. Cashdan, pp. 193–227. Barbash, Shephard

1993 Oaxacan Wood Carving: The Magic in the Trees. San Francisco: Chronicle Books. Barlett, Peggy, ed.

1980 Agricultural Decision-Making: Anthropological Approaches to Rural Development. New York: Academic Press. Barnes, John

1972 Social Networks. Reading, Massachusetts: Addison-Wesley. Baron, Jonathan

2008 Thinking and Deciding. Fourth edition. New York: Cambridge University Press. Barth, Fredrik

1967 On the Study of Social Change. American Anthropologist 69 (6): 661–669. Bator, Francis

1958 The Anatomy of Market Failure. Quarterly Journal of Economics 72 (3): 351–379. Becker, Gary

1974 A Theory of Social Interactions. Journal of Political Economy 82 (6): 1063–1093. 1976a Altruism, Egoism, and Genetic Fitness: Economics and Sociobiology. Journal of Economic Literature 14 (3): 817–826. 1976b The Economic Approach to Human Behavior. Chicago: University of Chicago Press. 1981 A Treatise on the Family. Cambridge: Harvard University Press. Becker, Gary, and Guity Nashat Becker

1997 The Economics of Life. New York: McGraw-Hill. Beckert, Jens

1996 What Is Sociological about Economic Sociology? Uncertainty and the Embeddedness of Economic Action. Theory and Society 25 (6): 803–840.

References 179 Bender, Donald

1967 A Refinement of the Concept of Household: Families, Co-Residence, and Domestic Functions. American Anthropologist 69 (5): 493–504. Benedict, Ruth

1934 Patterns of Culture. New York: Houghton Mifflin. Bergstrom, Ted

1989 A Fresh Look at the Rotten Kid Theorem and Other Household Mysteries. Journal of Political Economy 97 (5): 1138–1159. Bernheim, Douglas, and Oded Stark

1988 Altruism within the Household Reconsidered: Do Nice Guys Finish Last? American Economic Review 78 (5): 1034–1045. Blanchard, Olivier

2009 (Nearly) Nothing to Fear but Fear Itself. The Economist, January 31– February 6, p. 84. Bliege Bird, Rebecca, and Eric Smith

2005 Signaling Theory, Strategic Interaction, and Symbolic Capital. Current Anthropology 46 (2): 221–248. Boardman, Anthony, David Greenberg, Aidan Vining, and David Weimer

2006 Cost Benefit Analysis: Concepts and Practice. Upper Saddle River, New Jersey: Prentice Hall. Boettke, Peter

2009 Lin Ostrom—Political Economist—Wins 2009 Nobel. Austrian Economists, October 12. http://austrianeconomists.typepad.com/weblog/2009/ 10/lin-ostrom-political-economist-wins-2009-nobel.html, downloaded November 10, 2009. Bohannan, Paul

1963 Social Anthropology. New York: Holt, Reinhart, and Winston. Boserup, Ester

1965 The Conditions of Agricultural Growth. Chicago: Aldine. Boudon, Raymond

1998 Limitations of Rational Choice Theory. American Journal of Sociology 104 (3): 817–828. Bourdieu, Pierre

1977 Outline of Theory of Practice. Stanford, California: Stanford University Press. Bower, Bruce

2002 A Fair Share of the Pie. Science News 161 (7): 104–106. Bowles, Samuel, Robert Boyd, Ernst Fehr, and Herbert Gintis

1997 A Research Initiative on the Origins, Dimensions, and Policy Implications of Reciprocal Fairness. Unpublished manuscript. Boyd, Robert, and Peter Richerson

1985 Culture and the Evolutionary Process. Chicago: University of Chicago Press.

180 Anthropology, Economics, and Choice Briscoe, Simon, and Hugh Aldersey- Williams

2009 Panicology: Two Statisticians Explain What’s Worth Worrying About (and What’s Not) in the 21st Century. New York: Skyhorse. Britan, Gerald, and Michael Chibnik

1980 Bureaucracy and Innovation: An American Case. In Hierarchy and Society: Anthropological Perspectives on Bureaucracy, ed. Gerald Britan and Ronald Cohen, pp. 61–71. Philadelphia: Institute for the Study of Human Issues. Brian, Gerald, and Bette Denich

1976 Environment and Choice in Rapid Social Change. American Ethnologist 3 (1): 55–72. Brox, Ottar

1990 The Common Property Theory: Epistemological Status and Analytic Utility. Human Organization 49 (3): 227–235. Bulte, E. H., and G. C. Van Kooten

1999 Marginal Value of Charismatic Species: Implications for Conservation. Environmental and Resource Economics 14 (1): 119–130. Burling, Robbins

1962 Maximization Theories and the Study of Economic Anthropology. American Anthropologist 64 (4): 802–821. Cancian, Frank

1979 The Innovator’s Situation: Upper-Middle-Class Conservativism in Agricultural Communities. Stanford, California: Stanford University Press. Cashdan, Elizabeth, ed.

1990 Risk and Uncertainty in Tribal and Peasant Economies. Boulder, Colorado: Westview Press. Cassidy, John

2009 How Markets Fail: The Logic of Economic Calamities. New York: Farrar, Strauss and Giroux. Chamberlin, Edward

1948 An Experimental Imperfect Market. Journal of Political Economy 56 (2): 95–108. Chaudhuri, Ananish

2009 Experiments in Economics: Playing Fair with Money. London: Routledge. Chayanov, Alexander

1966 The Theory of Peasant Economy. (Translated from Russian and German. Writings originally published in the 1920s.) Homewood, Illinois: R. Irwin. Chen, Martha Alter, and Elizabeth Dunn

1996 Household Economic Portfolios. Washington, D.C.: Management Systems International. Chibnik, Michael

1975 Economic Strategies of Small Farmers in Stann Creek District, British Honduras. Ph.D. dissertation, Columbia University.

References 181

1978 The Value of Subsistence Production. Journal of Anthropological Research 34 (4): 561–576. 1980 Working Out or Working In: The Choice Between Wage Labor and Cash Cropping in Rural Belize. American Ethnologist 7 (1): 86–105. 1981 The Evolution of Cultural Rules. Journal of Anthropological Research 37 (3): 256–268. 1984 A Cross- Cultural Examination of Chayanov’s Theory. Current Anthropology 25 (3): 335–340. 1987 Saving Soil by Abandoning the Plow: Experimentation with No-Till Farming in an Iowa County. In Farm Work and Fieldwork: American Agriculture in Anthropological Perspective, ed. Michael Chibnik, pp. 90–117. Ithaca, New York: Cornell University Press. 1990 Double-Edged Risks and Uncertainties: Choices about Rice Loans in the Peruvian Amazon. In In Risk and Uncertainty in Tribal and Peasant Economies, ed. Cashdan, pp. 279–302. 1994 Risky Rivers: The Economics and Politics of Floodplain Farming in Amazonia. Tucson: University of Arizona Press. 2003 Crafting Tradition: The Making and Marketing of Oaxacan Wood Carvings. Austin: University of Texas Press. Choi, Hyup, and C. Milton Coughenour

1979 Socioeconomic Aspects of No-Tillage Agriculture: A Case Study of Farmers in Christian County, Kentucky. Lexington, Kentucky: Agricultural Experimentation Station. Clark, Gracia

1989 Separation between Trading and Home for Asante Women in Kumasi Central Market, Ghana. In The Household Economy: Reconsidering the Domestic Mode of Production, ed. Wilk, pp. 91–118. Clarke, Colin

2000 Class, Ethnicity, and Community in Southern Mexico. New York: Oxford University Press. Cohen, Jeffrey

1999 Cooperation and Community: Economy and Society in Oaxaca. Austin: University of Texas Press. 2004 The Culture of Migration in Southern Mexico. Austin: University of Texas Press. Comaroff, Jean

1985 Body of Power: Spirit of Resistance. Chicago: University of Chicago Press. Combs, Barbara, and Paul Slovic

1979 Newspaper Coverage of Causes of Death. Journalism Quarterly 56 (4): 837–849. Cook, Scott

1966 The Obsolete “Anti-Market” Mentality: A Critique of the Substantivist Approach to Economic Anthropology. American Anthropologist 68 (2): 323–345.

182 Anthropology, Economics, and Choice Cook, Scott, and Leigh Binford

1990 Obliging Need: Rural Petty Industry in Mexican Capitalism. Austin: University of Texas Press. Cosper, Harold

1983 Soil Suitability for Conservation Tillage. Journal of Soil and Water Conservation 38 (3): 152–155. Dalton, George

1961 Economic Theory and Primitive Society. American Anthropologist 63 (1): 1–25. Dawkins, Richard

1976 The Selfish Gene. New York: Oxford University Press. Delang, Claudio

2006 Not Just Minor Forest Products: The Economic Rationale for the Consumption of Wild Food Plants by Subsistence Farmers. Ecological Economics 59 (1): 64–73. Dharmaratne, G. S., F. Y. Sang, and L. J. Walling

2000 Tourist Potential for Financing Protected Areas. Annals of Tourism Research 27 (3): 590–610. Dietz, Thomas, Elinor Ostrom, and Paul Stern

2003 The Struggle to Govern the Commons. Science 302 (December 12): 1907–1910. Douglas, Mary

1985 Risk Acceptability According to the Social Sciences. New York: Russell Sage Foundation. 1990 Risk as a Forensic Resource. Daedalus 119 (4): 1–16. 1992 Risk and Blame: Essays in Cultural Theory. London: Routledge. Douglas, Mary, and Aaron Wildavsky

1982 Risk and Culture: An Essay on the Selection of Technological and Environmental Dangers. Berkeley: University of California Press. Duesenberry, James

1960 Comment on Becker’s “An Economic Analysis of Fertility.” In Demographic and Economic Change in Developing Countries, conference of Universities–National Bureau Committee for Economic Research, pp. 231–234. Dugatkin, Lee

2006 The Altruism Equation: Seven Scientists Search for the Origins of Goodness. Princeton: Princeton University Press. Durham, William

1991 Coevolution: Genes, Culture, and Human Diversity. Stanford: Stanford University Press. Durrenberger, E. Paul, ed.

1984 Chayanov, Peasants, and Economic Anthropology. New York: Academic Press.

References 183 Ellsberg, Daniel

1961 Risk, Ambiguity, and the Savage Axioms. Quarterly Journal of Economics 75 (4): 643–699. Elster, Jon

1989 Solomonic Judgements: Studies in the Limitations of Rationality. Cambridge: Cambridge University Press. Ensminger, Jean

2002 Experimental Economics: A Powerful New Method for Theory Testing in Anthropology. In Theory in Economic Anthropology, ed. Jean Ensminger, pp.59–78. Walnut Creek, California: AltaMira Press. 2004 Market Integration and Fairness: Evidence from Ultimatum, Dictator, and Public Goods Experiments in East Africa. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 356–381. Fehr, Ernst, and Simon Gachter

2002 Altruistic Punishment in Humans. Nature 415 (January 10): 137–140. Faulkner, Edward

1943 Plowman’s Folly. Norman, Oklahoma: University of Oklahoma Press. Fink, Deborah

1987 Farming in Open Country, Iowa. Women and the Changing Farm Economy. In Farm Work and Fieldwork, ed. Chibnik, pp. 121–144. Firth, Raymond

1951 Elements of Social Organization. London: Watts. Fox, Justin

2009 The Myth of the Rational Market: A History of Risk, Reward, and Delusion on Wall Street. New York: HarperCollins. Friedman, Milton

1953 Essays in Positive Economics. Chicago: University of Chicago Press. Folbre, Nancy

1986 Hearts and Spades: Paradigms of Household Economics. World Development 14 (2): 245–255. 1994 Who Pays for the Kids? Gender and the Structure of Constraint. New York: Routledge. Gigerenzer, Gerd

2008 Rationality for Mortals: How People Cope with Uncertainty. New York: Oxford University Press. Gladwell, Malcolm

2005 Blink: The Power of Thinking without Thinking. New York: Little, Brown. Godoy, Ricardo, and Ruben Lubowski

1992 Guidelines for the Economic Valuation of Nontimber Tropical-Forest Products. Current Anthropology 33 (4): 423–433. Godoy, Ricardo, Ruben Lubowski, and Anil Markandya

1993 A Method for the Economic Valuation of Non-Timber Tropical Forest Products. Economic Botany 47 (3): 220–233.

184 Anthropology, Economics, and Choice Goland, Carol

1993 Field Scattering as Agricultural Risk Management: A Case Study from Cuyo Cuyo, Department of Puno, Peru. Mountain Research and Development 13 (4): 317–338. Granovetter, Mark

1985 Economic Action and Social Structure: The Problem of Embeddedness. American Journal of Sociology 31 (3): 481–510. Groopman, Jerome

2007 How Doctors Think. Boston: Houghton Mifflin. Grossbard- Shechtman, Shoshana

2001 The New Home Economics at Columbia and Chicago. Feminist Economics 7 (3): 103–130. Gunder Frank, Andre

1969 Latin America: Underdevelopment or Revolution. New York: Monthly Review Press. Gurven, Michael

2004 Does Market Exposure Affect Economic Game Behavior? The Ultimatum Game and the Public Goods Game among the Tsimane’ of Bolivia. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 194–231. Gurven, Michael, and Jeffrey Winking

2008 Collective Action in Action: Prosocial Behavior in and out of the Laboratory. American Anthropologist 110 (2): 179–190. Güth, Werner, Rolf Schmittberger, and Bernd Schwarze

1982 An Experimental Analysis of Ultimatum Bargaining. Journal of Economic Behavior and Organizations 3: 367–388. Guyer, Jane

1981 Household and Community in African Studies. African Studies Review 24 (2/3): 87–137. 1988 The Multiplication of Labor: Historical Methods in the Study of Gender and Agricultural Change in Modern Africa. Current Anthropology 29 (2): 247–272. Hamilton, William

1963 The Evolution of Altruistic Behavior. American Naturalist 97: 354–356. Hammel, Eugene

1984 On the *** of Studying Household Form and Function. In Households: Comparative and Historical Studies of the Domestic Group, ed. Robert Netting, Richard Wilk, and Eric Arnould, pp. 29–43. Hardin, Garrett

1968 The Tragedy of the Commons. Science 163: 1243–1248. Harford, Tim

2005 The Undercover Economist. New York: Random House. 2008 The Logic of Life: The Rational Economics of an Irrational World. New York: Random House.

References 185 Harris, Marvin

1977 Cannibals and Kings: The Origins of Cultures. New York: Random House. Hart, Gillian

1992 Imagined Unities: Constructions of “the Household” in Economic Theory. In Understanding Economic Process, ed. Sutti Ortiz and Susan Lees, pp. 112–129. Lanham, Maryland: University Press of America. Hartmann, Heidi

1979 The Unhappy Marriage of Marxism and Feminism: Towards a More Progressive Union. Capital and Class 8 (Summer): 1–33. Hegmon, Michelle

1989 Risk Reduction and Variation in Agricultural Economies: A Computer Simulation of Hopi Agriculture. Research in Economic Anthropology 11: 89–121. Henrich, Joseph, Robert Boyd, Samuel Bowles, Colin Camerer, Ernst Fehr, and Herbert Gintis, eds.

2004 Foundations of Human Sociality: Economic Experiments and Ethnographic Evidence from Fifteen Small-Scale Societies. New York: Oxford University Press. Henrich, Joseph, Robert Boyd, Samuel Bowles, Colin Camerer, Ernst Fehr, Herbert Gintis, and Richard McElreath

2004 Overview and Synthesis. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 8–54. Henrich, Joseph, and Richard McElreath

2002 Are Peasants Risk-Averse Decision Makers? Current Anthropology 43 (1): 172–181. Henrich, Joseph, and Natalie Smith

2004 Comparative Experimental Evidence from Machiguenga, Mapuche, Huinca, and American Populations. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 125–167. Henrich, Natalie, and Joseph Henrich

2007 Why Humans Cooperate: A Cultural and Evolutionary Explanation. New York: Oxford University Press. Hill, Kim, and Michael Gurven

2004 Economic Experiments to Examine Fairness and Cooperation among the Ache Indians of Paraguay. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 382–412. Himmelweit, Susan, and Simon Mohun

1977 Domestic Labour and Capital. Cambridge Journal of Economics 1 (1): 15–31. Hirschmann, Albert

1982 Rival Interpretations of Market Society: Civilizing, Destructive, or Feeble. Journal of Economic Literature 20 (4): 1463–1484.

186 Anthropology, Economics, and Choice Hoddinott, John, and Lawrence Haddad

1994 Women’s Income and Boy-Girl Nutrition Outcomes in the Côte d’Ivoire. World Development 22 (4): 543–553. Hornborg, Alf

2001 The Power of the Machine: Global Inequalities of Economy, Technology, and Environment. Walnut Creek, California: AltaMira Press. Howes, David, ed.

1996 Cross-Cultural Consumption. New York: Routledge. Hudson, Michael

2010 The Use and Abuse of Mathematical Economics. Real-World Economics Review 55(17): 2–22. Huxley, Aldous

1934 Beyond the Mexique Bay. New York: Harper and Brothers. Ironmonger, Duncan

2001 Household Production. International Encyclopedia of the Social and Behavioral Sciences, volume 10, pp. 6934–6939. New York: Elsevier Science. Johnson, Allen

1978 Quantification in Cultural Anthropology. Stanford, California: Stanford University Press. Johnson, Craig

2004 Uncommon Ground: The “Poverty of History” in Common Property Discourse. Development and Change 35 (3): 407–433. Kahneman, Daniel, and Amos Tversky

1982a Judgment under Uncertainty: Heuristics and Biases. In Judgment under Uncertainty: Heuristics and Biases, ed. Daniel Kahneman, Paul Slovic, and Amos Tversky, pp. 3–22. New York: Cambridge University Press. 1982b On the Psychology of Prediction. In Judgment under Uncertainty: Heuristics and Biases, ed. Daniel Kahneman, Paul Slovic, and Amos Tversky, pp. 48–68. New York: Cambridge University Press. 1984 Choices, Values, and Frames. American Psychologist 39 (4): 341–350. Kaprow, Miriam

1985 Manufacturing Danger: Fear and Pollution in Industrial Society. American Anthropologist 87 (2): 342–356. Keane, Webb

2003 Self-Interpretation, Agency, and the Objects of Anthropology: Reflections on a Genealogy. Comparative Studies in Society and History 45 (2): 222–248. Keesing, Felix

1958 Cultural Anthropology: The Science of Custom. New York: Holt, Rinehart, and Winston. Keynes, John Maynard

1936 The General Theory of Employment, Interest, and Money. New York: Harcourt, Brace, and World.

References 187

1937 The General Theory of Unemployment. Quarterly Journal of Economics 51 (2): 209–223. Knight, Frank

1921 Risk, Uncertainty, and Profit. Boston: Houghton Mifflin. Kohler, Timothy, and Carla Van West

1996 The Calculus of Self-Interest in the Development of Cooperation. Sociopolitical Development and Risk among the Northern Anasazi. In Evolving Complexity and Environmental Risk in the Prehistoric Southwest, ed. Joseph Tainter and Bonnie Bagley Tainter, pp. 169–196. Reading, Massachusetts: Addison-Wesley. Krippner, Greta, and Anthony Alvarez

2007 Embeddedness and the Intellectual Projects of Economic Sociology. Annual Review of Sociology 33: 219–240. Krippner, Greta, Mark Granovetter, Fred Block, Nicole Biggart, Tom Beamish, Youtien Hsing, Gillian Hart, Giovanni Arrighi, Margie Mendell, John Hall, Michael Burawoy, Steve Vogel, and Sean O’Riain

2004 Polanyi Symposium: A Conversation on Embeddedness. Socio-Economic Review 2: 109–135. Krishna, Raj

1969 Models of the Family Farm. In Subsistence Agriculture and Economic Development, ed. Clifton Wharton, pp. 185–189. Chicago: Aldine. Krugman, Paul

2009 An Institutional Economics Prize. New York Times, October 12. http:// krugman.blogs.nytimes.com/2009/10/12/an-institutional-economicsprize, downloaded November 10, 2009. Kunstadter, Peter

1984 Cultural Ideals, Socioeconomic Change, and Household Composition: Karen, Lua’, Hmong, and Thai in Northwestern Thailand. In Households: Comparative and Historical Studies of the Domestic Group, ed. Netting, Wilk, and Arnould, pp. 299–329. Kuznar, Lawrence

2001 Risk Sensitivity and Value among Andean Pastoralists: Measures, Models, and Empirical Tests. Current Anthropology 42 (3): 432–440. Kyrk, Hazel

1933 Economic Problems of the Family. New York: Harper. LeClair, Edward

1962 Economic Theory and Economic Anthropology. American Anthropologist 64 (6): 1179–1203. Lee, Richard

1979 The !Kung San: Men, Women, and Work in a Foraging Society. Cambridge: Cambridge University Press. Lehrer, Jonah

2009 How We Decide. New York: Houghton Mifflin Harcourt.

188 Anthropology, Economics, and Choice Lesorogol, Carolyn

2009 Comment on list-serve, Society for Economic Anthropology, October 12. Levitt, Steven

2009 What This Year’s Nobel Prize in Economics Says about the Nobel Prize in Economics. New York Times, October 12. http://freakonomics.blogs .nytimes.com/2009/10/12/what-this-years-nobel-prize-in-economicssays-about-the-nobel-prize-in-economics, downloaded November 10, 2009. Levitt, Steven, and Stephen Dubner

2005 Freakonomics: A Rogue Economist Explores the Hidden Side of Everything. New York: William Morrow. 2009 Superfreakonomics. New York: William Morrow. Levitt, Steven, and John List

2007 What Do Laboratory Experiments Measuring Social Preferences Reveal about the Real World? Journal of Economic Perspectives 21 (2): 153–174. Lloyd, William Forster

1977 [1833] On the Checks to Population. In Managing the Commons, ed. Garret Hardin and John Baden, pp. 8–15. San Francisco: W. H. Freeman. Lockwood, Victoria

1989 Tubuai Women Potato Planters and the Political Economy of IntraHousehold Gender Relations. In The Household Economy: Reconsidering the Domestic Mode of Production, ed. Wilk, pp. 197–220. Lowie, Robert

1920 Primitive Society. New York: Liveright. Lucas, Robert

1987 Models of Business Cycles. Oxford: Blackwell. Lundberg, Shelly, Robert Pollak, and Terence Wales

1997 Do Husbands and Wives Pool Their Resources? Evidence from the United Kingdom Child Benefit. Journal of Human Resources 32 (3): 463–480. Lupton, Deborah

1999 Risk. London: Routledge. Mace, Ruth, and Alasdair Houston

1989 Pastoral Strategies for Survival in Unpredictable Environments: A Model of Herd Composition that Maximizes Household Viability. Agricultural Systems 31 (2): 185–204. Maclachlan, Morgan, ed.

1987 Household Economies and Their Transformations. Lanham, Maryland: University Press of America. Malinowski, Bronislaw

1922 Argonauts of the Western Pacific. London: Routledge and Kegan Paul. Mankiw, Gregory

2005 Principles of Microeconomics. Fourth edition. Mason, Ohio: Thomson.

References 189 Manser, Marilyn, and Murray Brown

1980 Marriage and Household Decision-Making: A Bargaining Analysis. International Economic Review 21 (2): 31–44. Marlowe, Frank

2004 Dictators and Ultimatums in an Egalitarian Society of HunterGatherers: The Hazda of Tanzania. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 168–193. Marra, Michele, David Pannell, and Amir Abadi Ghadim

2003 The Economics of Risk, Uncertainty, and Learning in the Adoption of New Agricultural Technologies: Where Are We on the Learning Curve? Agricultural Systems 75 (2–3): 215–234. Martin, Fenton

1989 Common Pool Resources and Collective Action: A Bibliography. Bloomington: Indiana University Workshop in Political Theory and Policy Analysis. Massey, Douglas, Luin Goldring, and Jorge Durand

1994 Continuities in Transnational Migration: An Analysis of Nineteen Mexican Communities. American Journal of Sociology 99 (6): 1492–1533. McCay, Bonnie, and James Acheson, eds.

1987 The Question of the Commons. Tucson: University of Arizona Press. McCloskey, Deirdre

1998 The Rhetoric of Economics. Second edition. Madison: University of Wisconsin Press. McConnell, Campbell, and Stanley Brue

2005 Economics: Principles, Problems, and Policies. Sixteenth edition. Boston: McGraw-Hill. McElreath, Richard

2004 Community Structure, Mobility, and the Strength of Norms in an African Society. In Foundations of Human Society, ed. Joseph Henrich et al., pp. 335–355. McElroy, Marjorie, and Mary Jean Horney

1981 Nash Bargained Household Decisions: Toward a Generalization of the Theory of Demand. International Economic Review 22 (2): 333–349. Mellor, John

1966 The Economics of Agricultural Development. Ithaca, New York: Cornell University Press. Meyer, Donald, and Jack Meyer

2006 Measuring Risk Aversion. Boston: Now. Miller, Daniel

1998 A Theory of Shopping. Ithaca, New York: Cornell University Press. Mincer, Jacob

1962 Labor Force Participation of Married Women: A Study of Labor Supply. In Aspects of Labor Economics, ed. H. Gregg Lewis, pp. 1–35. Princeton: Princeton University Press.

190 Anthropology, Economics, and Choice Mintz, Sidney

1985 Sweetness and Power. New York: Viking Penguin. Mitchell, J. Clyde

1974 Social Networks. Annual Review of Anthropology 3: 279–299. Moberg, Mark

1992 Citrus, Strategy, and Class: The Politics of Development in Southern Belize. Iowa City: University of Iowa Press. Moerman, Michael

1968 Agricultural Change and Peasant Choice in a Thai Village. Berkeley: University of California Press. Molyneux, Maxine

1979 Beyond the Domestic Labour Debate. New Left Review 116 ( July– August): 3–27. Murdock, George Peter

1949 Social Structure. New York: Macmillan. Murphy, Arthur, and Alex Stepick

1991 Social Inequality in Oaxaca: A History of Resistance and Change. Philadelphia: Temple University Press. Muth, John

1961 Rational Expectations and the Theory of Price Movement. Econometrica 29 (3): 315–335. Nag, Moni, Benjamin White, and R. Creighton Peet

1978 An Anthropological Approach to the Study of Economic Value of Children in Java and Nepal. Current Anthropology 19 (2): 293–306. Nas, Tevfik

1996 Cost-Benefit Analysis: Theory and Application. Thousand Oaks, California: Sage. Netting, Robert

1993 Smallholders, Householders: Farm Families and the Ecology of Intensive, Sustainable Agriculture. Stanford, California: Stanford University Press. Netting, Robert, Richard Wilk, and Eric Arnould, eds.

1984 Households: Comparative and Historical Studies of the Domestic Group. Berkeley: University of California Press. New York Times

1989 Rain Forest Worth More If Uncut, Study Says. July 4, science section. Nowak, Martin, Karen Page, and Karl Sigmund

2000 Fairness versus Reason in the Ultimatum Game. Science 289: 1773–1775. Olson, Mancur

1965 The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, Massachusetts: Harvard University Press. Oosterbeek, Hessel, Randolph Sloof, and Gus van de Kuilen

2004 Cultural Differences in Ultimatum Game Experiments: Evidence from a Meta-Analysis. Experimental Economics 7: 171–188.

References 191 Orlove, Benjamin, ed.

1997 The Allure of the Foreign: Imported Foreign Goods in Postcolonial Latin America. Ann Arbor: University of Michigan Press. Ortiz, Sutti

1973 Uncertainties in Peasant Farming: A Colombian Case. New York: Humanities Press. 1980 Forecasts, Decisions, and the Farmer’s Response to Uncertain Environments. In Agricultural Decision Making, ed. Barlett, pp. 177–202. Ortner, Sherry

1984 Theory in Anthropology Since the Sixties. Comparative Studies in Society and History 26 (1): 126–166. 2006 Anthropology and Social Theory. Durham, North Carolina: Duke University Press. Ostrom, Elinor

1990 Governing the Commons: The Evolution of Institutions for Collective Action. New York: Cambridge University Press. 1999 Coping with Tragedies of the Commons. Annual Review of Political Science 2: 493–535. Overbye, Einar

1995 Making a Case for the Rational, Self-Regarding, “Ethical” Voter . . . and Solving the “Paradox of Not Voting” in the Process. European Journal of Political Research 27 (3): 369–396. Parsons, Stephen

2005 Rational Choice and Politics. London: Continuum. Peters, Charles, Alwyn Gentry, and Robert Mendelsohn

1989 Valuation of an Amazonian Rainforest. Nature 339 (June 29): 655–656. Plattner, Stuart, ed.

1989 Economic Anthropology. Stanford, California: Stanford University Press. Polanyi, Karl

1944 The Great Transformation. New York: Rinehart. 1957 The Economy as Instituted Process. In Trade and Market in Early Empires, ed. Karl Polanyi, Conrad Arensberg, and Harry Pearson, pp. 243– 270. New York: Free Press. Pollak, Robert

2003 Gary Becker’s Contributions to Family and Household Economics. Review of Economics of the Household 1 (1–2): 111–141. Quinn, Naomi

1978 Do Mfantse Fish Sellers Estimate Probabilities in Their Heads? American Ethnologist 5 (2): 206–226. 1982 “Commitment” in American Marriage: A Cultural Analysis. American Ethnologist 9 (4): 755–798. 1996 Culture and Contradiction: The Case of Americans Reasoning About Marriage. Ethos 24 (3): 391–425.

192 Anthropology, Economics, and Choice Reddy, Sanjay

1996 Claims to Expert Knowledge and the Subversion of Democracy: The Triumph of Risk over Uncertainty. Economy and Society 25 (2): 222–254. Redfield, Robert

1947 The Folk Society. American Journal of Sociology 52 (4): 293–308. Reid, Margaret

1934 The Economics of Household Production. New York: Wiley. Richards, Robert

1987 Darwin and the Emergence of Evolutionary Theories of Mind and Behavior. Chicago: University of Chicago Press. Richerson, Peter, and Robert Boyd

2005 Not by Genes Alone: How Culture Transformed Human Evolution. Chicago: University of Chicago Press. Rogers, Everett

2003 Diffusion of Innovations. Fifth edition. New York: Free Press. Romney, D., A. Wright, R. Arbuckle, and V. Vial

1959 Land in British Honduras. Report of the British Honduras Land Survey Team. London: Her Majesty’s Stationery Office. Roseberry, William

1983 Coffee and Capitalism in the Venezuelan Andes. Austin: University of Texas Press. Roth, Alvin

1995 Introduction to Experimental Economics. In The Handbook of Experimental Economics, ed. John Kagel and Alvin Roth, pp. 3–109. Princeton: Princeton University Press. Roth, Alvin, Vesna Prasnikar, Mashiro Okuno- Fujiware, and Shmuel Zamir

1991 Bargaining and Market Behavior in Jerusalem, Ljubljana, Pittsburgh, and Tokyo: An Experimental Study. American Economic Review 81 (5): 1068–1095. Rutz, Henry, and Benjamin Orlove, eds.

1989 The Social Economy of Consumption. Lanham, Maryland: University Press of America. Sahlins, Marshall

1972 Stone Age Economics. Chicago: Aldine. 1976 Culture and Practical Reason. Chicago: University of Chicago Press. Samuelson, Paul

1956 Social Indifference Curves. Quarterly Journal of Economics 70 (1): 1–22. Samuelson, Paul, and William Nordhaus

2005 Economics. Eighteenth edition. Boston: McGraw-Hill. Scharfstein, David, and Jeremy Stein

1990 Herd Behavior and Investment. American Economic Review 80 (3): 465–479. Schneiderman, Ross

2009 Reactions to the Nobel in Economic Science. New York Times, Octo-

References 193

ber 12. http://economix.blogs.nytimes.com/2009/10/12/reactions-tothe-nobel-prize-in-economics, downloaded November 10, 2009. Schweers Cook, Karen, and Margaret Levi, eds.

1990 The Limits of Rationality. Chicago: University of Chicago Press. Scott, James

1976 The Moral Economy of the Peasant. New Haven: Yale University Press. 1990 Domination and the Arts of Resistance. New Haven: Yale University Press. Seccombe, Wally

1974 The Housewife and Her Labour under Capitalism. New Left Review 83 ( January–February): 3–24. Sen, Amartya

1997 Maximization and the Act of Choice. Econometrica 65 (4): 745–779. Settlement Report

1948 Report of the British Guiana and British Honduras Settlement Commission. London: Her Majesty’s Stationery Office. Shiller, Robert

2000 Irrational Exuberance. Princeton: Princeton University Press. Sigmund, Karl, Ernst Fehr, and Martin Nowak

2002 The Economics of Fair Play. Scientific American, January, pp. 82–87. Skidelsky, Robert

2008 The Remedist: Why John Maynard Keynes Is the Man of the Year. New York Times Magazine, December 14, pp. 21–22. Smith, Eric

1983 Anthropological Applications of Optimal Foraging Theory. Current Anthropology 24 (5): 625–651. Smith, Joyotee, and Abraham Mandac

1995 Subjective versus Objective Yield Distributions as Measures of Production Risk. American Journal of Agricultural Economics 77 (1): 152–162. Stephen, Lynn

2005 Zapotec Women: Gender, Class, and Ethnicity in Globalized Oaxaca. Second edition. Durham, North Carolina: Duke University Press. Steward, Julian

1955 Theory of Culture Change: The Methodology of Multilinear Evolution. Urbana, Illinois: University of Illinois Press. Surowiecki, James

2004 The Wisdom of Crowds. New York: Doubleday. Taleb, Nassim Nicholas

2007 The Black Swan: The Impact of the Highly Improbable. New York: Random House. Tax, Sol

1953 Penny Capitalism: A Guatemalan Indian Economy. Washington, D.C.: Smithsonian Institution.

194 Anthropology, Economics, and Choice Taylor, C. Robert

2003 The Role of Risk versus the Role of Uncertainty in Economic Systems. Agricultural Systems 75 (2–3): 251–264. Thaler, Richard

1988 Anomalies: The Winner’s Curse. Journal of Economic Perspectives 2 (1): 191–202. Thaler, Richard, and Cass Sunstein

2008 Nudge: Improving Decisions about Health, Wealth, and Happiness. New Haven: Yale University Press. Thomas, Duncan

1990 Intra-Household Resource Allocation: An Inferential Approach. Journal of Human Resources 25 (4): 635–664. Tracer, David

2004 Market Integration, Reciprocity, and Fairness in Rural Papua New Guinea: Results from a Two-Village Ultimatum Game Experiment. In Foundations of Human Sociality, ed. Joseph Henrich et al., pp. 232–259. VanWey, Leah, Catherine Tucker, and Eileen Diaz McConnell

2005 Community Organization, Migration, and Remittances in Oaxaca. Latin American Research Review 40 (1): 83–107. von Neumann, John, and Oskar Morgenstern

1944 Theory of Games and Economic Behavior. Princeton: Princeton University Press. Wachter, Kenneth, Eugene Hammel, and Peter Laslett

1978 Statistical Studies of Historical Social Structure. New York: Academic Press. Wallerstein, Immanuel

1974 The Modern World System: Capitalist Agriculture and the Origins of the European World Economy in the Sixteenth Century. New York: Academic Press. Waring, Marilyn

1996 Counting for Nothing: What Men Value and What Women Want. Second edition. Toronto: University of Toronto Press. Warner, Kenneth

1974 The Need for Some Innovative Concepts of Innovation: An Examination of Research on the Diffusion of Innovations. Policy Sciences 5 (4): 433–451. Wessel, David

2002 The Civilizing Effect of the Market. Wall Street Journal, January 24, A1. White, Benjamin

1976 Production and Reproduction in a Javanese Village. Ph.D. dissertation, Department of Anthropology, Columbia University. (Printed in Bogor, Indonesia, Agricultural Development Council.)

References 195 White, Douglas, and Thomas Schweizer

1998 Kinship, Networks, and Exchange. Cambridge: Cambridge University Press. Wiessner, Polly

2009 Experimental Games and Games of Life among the Ju/’hoan Bushmen. Current Anthropology 50 (1): 133–138. Wilk, Richard, ed.

1989 The Household Economy: Reconsidering the Domestic Mode of Production. Boulder, Colorado: Westview Press. Wilk, Richard, and Lisa Cliggett

2007 Economies and Culture: Foundations of Economic Anthropology. Second edition. Boulder, Colorado: Westview. Wilk, Richard, and Stephen Miller

1997 Some Methodological Issues in Counting Communities and Households. Human Organization 56 (1): 64–70. Williams, Wendy, and Robert Whitcomb

2007 Cape Wind: Money, Celebrity, Class, Politics, and the Battle for Our Energy Future. New York: Public Affairs. Wilson, Edward

1975 Sociobiology: The New Synthesis. Cambridge, Massachusetts: Harvard University Press. Wilson, Edward, and Bert Hölldobler

2005 Eusociality: Origins and Consequences. Proceedings of the National Academy of Sciences 102: 13367–13371. Winterhalder, Bruce

1987 The Analysis of Hunter- Gatherer Diet: Stalking an Optimal Foraging Model. In Food and Evolution, ed. Marvin Harris and Eric Ross, pp. 311–339. Philadelphia: Temple University Press. Winterhalder, Bruce, Flora Liu, and Bram Tucker

1999 Risk-Sensitive Adaptive Tactics: Models and Evidence from Subsistence Studies in Biology and Anthropology. Journal of Archaeological Research 7 (4): 301–348. Wirth, Louis

1938 Urbanism as a Way of Life. American Journal of Sociology 44 (1): 1–24. Yanagisako, Sylvia Junko

1979 Family and Household: The Analysis of Domestic Groups. Annual Review of Anthropology 8: 161–205. Yi, Yun- Ae

1996 Margaret G. Reid: Life and Achievements. Feminist Economics 2 (3): 17–36.

THIS PAGE INTENTIONALLY LEFT BLANK

Index

Ache of Paraguay, 108 Acheson, James, 150, 152–155, 159 agency, 31, 34 Agrarian Bank (Peru), 66–67, 71, 146 Agrawal, Arun, 149–150, 151, 153, 157–158 agriculture. See cash crops; subsistence agriculture Akerlof, George, 16 Alersey-Williams, Hugh, 86 altruistic behavior: and cross-cultural economic experiments, 99; and evolutionary biology, 99–101, 103, 104, 117, 156, 169; and free-rider problems, 156; and household economic decisions, 136–138; origins of, 91–92, 93, 97 Alvard, Michael, 109 Amster, Randall, 161 anthropology: approaches to choice, 2–5, 19, 23–24, 25, 26, 31–35, 36, 171; approaches to decision making, 1–2, 21, 31–35, 61, 81–87, 89, 140, 165–168, 171–172; and common property, 152–153, 159, 160, 163; economic experiments in, 90–93, 98, 99–103, 105–116, 117, 165, 169; holistic tradition of, 3, 28, 90, 165; and household economic decisions, 118–119, 130–134, 170; and practice theory, 35; and rational

choice theories, 2–3; rejection of risk models from other disciplines, 84–87; and risk, 61, 81–87, 167–168; and risk models from other disciplines, 81–84; and uncertainty, 61, 83. See also economic anthropology Arab-American Chaldeans, 115 Aristotle, 174n3 (chap. 2) Ashcraft, Norman, 41, 42 asymmetric information, 16 Aumann, Robert, 162 Au of New Guinea, 108, 111 availability heuristic, 78–80, 85 Axelrod, Robert, 101 Baksh, Michael, 87 Baron, Jonathan, 80 Barr, Abigail, 109 Barth, Fredrik, 32 Bator, Francis, 15 Becker, Gary, 1–3, 5, 136–138, 140, 160, 170 Beckert, Jens, 77 behavioral ecology models, 83–84 behavioral economics, 16, 114, 164, 167, 169 Belize: agricultural development in, 41–42; choice in rural Belize, 20, 21–24; Creoles of, 40, 42, 43, 121; economic conditions of, 175n3 (chap. 5); ethnic groups of, 39,

198 Anthropology, Economics, and Choice

40–41; forest work in, 42; Garifuna (Caribs) of, 40, 43, 121; households defined in, 119–123, 125, 126, 129–130; marriage in, 24, 175n1 (chap. 5); Mayas in, 24, 40–41, 42, 43, 121; and migration to United States, 175–176n4; tourist industry in, 22, 39 Bender, Donald, 131 Bentham, Jeremy, 1 Binford, Leigh, 127 Blanchard, Olivier, 76–77 Boettke, Peter, 160 Bohannan, Paul, 131 Boyd, Robert, 103 Briscoe, Simon, 86 Britain, and Belize, 22, 40, 41 Britan, Gerald, 32 British Honduras. See Belize Brox, Ottar, 159–160, 176n5 Brue, Stanley, 29, 174n1 (chap. 1) Bush, George W., 28 Cancian, Frank, 81–85, 168 capitalism: capital-intensive agriculture, 54; and fairness, 105; and household economic decisions, 139; and unpaid work, 136 cash crops: agricultural credit for, 65–67; calculating monetary value of, 43–44, 47; earning income from, 20, 21–22, 43; risks of, 21, 43, 48–49 Cassidy, John, 15 centrally planned economies, 30 Chamberlin, Edward Hastings, 103–104 Chayanov, A. V., 45–47, 51, 57–58, 132, 135–136, 174nn4–5 choice: anthropological approaches to, 2–5, 19, 23–24, 25, 26, 31–35, 36, 171; consequences of, 35–36; context of, 19; and cost-benefit analysis, 38–39; economic approaches to, 1–5, 19, 21, 24–31, 33, 36, 118, 164, 165, 167, 171; and expected utility,

6–7; issues in analysis of, 17–19; and practice theory, 35; restraints on, 22; and science-humanities continuum, 164–166; sociological approaches to, 25, 31–35, 164– 165, 166. See also decision making; rational choice theories Cliggett, Lisa, 34 cognitive psychology: critiques of economics, 166; and knowledge of concepts of probability, 18, 78; and risk, 61, 78–80, 81, 85, 88, 89, 167; and suboptimal choices, 2; and uncertainty, 61, 88 Cohen, Jeffrey, 128 common property: anthropological approaches to, 152–153, 159, 160, 163; comparative studies of, 155–159; economic approaches to, 159, 160, 161; and ethnography, 149, 150, 162, 163, 171; and government intervention, 144, 148, 149, 151, 152, 153, 155; interpretations of, 159–160; and Maine lobstermen, 152, 154–155, 176n3; and nucleated and perimeter-defended areas, 154– 155; in Peruvian Amazon, 144–149; political reactions to, 160–162; and resource conservation, 149–151, 154–155, 157–158, 160; and resource destruction, 142–143, 149, 152, 170; restrictions on, 143, 144, 145, 147, 148, 151, 152, 170; and selfinterests, 19, 160 competition, 4–5, 91, 93, 95–96, 97, 98 Comte, Auguste, 1 consumption: and economic anthropology, 5; and household behavior, 135, 136, 140, 141, 170. See also production for home consumption contingent valuation, 51, 52 Cook, Scott, 127, 173n2 cooperation: benefits of, 97–98; cooperative work groups, 124–125, 129; distinctiveness of human cooperation, 101–102, 116; and evolu-

Index 199

tionary biology, 18, 90–91, 97, 99, 101–102, 104, 116, 169; and experimental economics, 91, 93, 104, 114; genetic explanations for, 102–103, 110; and household economic decisions, 139; and households, 121; and kin selection explanation, 100, 116, 117; and Oaxacan wood carvers, 95–97, 98, 117; and rational choice theories, 92, 116, 168–169; and titfor-tat strategy, 101; and Ultimatum Game, 110, 111, 112, 113, 115 cost-benefit analyses, 38–39, 71, 73 Cruz, Isidoro, 94, 95–96, 97, 98 cultural change, 17, 31, 32, 37 cultural evolution, 45, 102, 103, 151, 175n5 (chap. 3) cultural norms: and common property, 19, 144, 154; and cooperation, 115, 116–117; and decision making, 31, 32, 33, 36, 81, 164; and ethnography, 113; and Oaxacan wood carvers, 98 cultural practices: effect on decision making, 26; history as influence on, 3, 4, 5; reductionist explanations of, 8; and substantivist approach to economic choices, 23–24; utilitarian explanations of, 6, 7, 8 cultural transmission of ideas, 99, 102, 103 culture-gene coevolution, 99 Darwin, Charles, 91, 100 Dawkins, Richard, 100 decision making: anthropological approaches to, 1–2, 21, 31–35, 61, 81–87, 89, 140, 165–168, 171–172; cognitive psychological approaches to, 61, 78–80, 88, 151; complexity of, 71, 74, 76, 88, 89, 98, 171; and cooperation, 104; and cultural norms, 31, 32, 33, 36, 81, 164; debates on, 21; economic anthropological approaches to, 21, 85, 86–87, 169; economic approaches to, 1–2,

6, 21, 23, 24–31, 61, 74–78, 87–88, 161, 163, 165–166, 171–172; and economic experiments, 114; effects of risk aversion on, 4; ethnographic approach to, 5, 12, 18, 81, 86, 89, 163, 165, 166, 172; expected utility models of, 6–7, 8; long-term consequences, 7, 13, 14, 71, 72–73, 98; short-term consequences, 7, 13, 14, 71, 72–73, 98; significance of, 35–36; and social institutions, 31, 32, 36; and unpaid versus paid work, 57; and women, 134, 174–175n4. See also choice; household economic decisions; risk; uncertainty decision-making units, 7, 13, 14, 17, 18–19. See also household economic decisions Delang, Charles, 51, 52–53, 57, 174n6 democracies, 26, 92 Denich, Bette, 32 dependency theory, 5, 22–24 Dictator Game, 114, 115 diversified activities, and risk and uncertainty, 71, 73, 89 Domestic Labor Debate, 136 domestic labor (housework), 38, 48, 53, 55–57, 58 dot-com boom of late 1990s, 16 Douglas, Mary, 85–86 dual inheritance theory, 103 dual systems theory, 139 earning income, choice in, 20, 21–24 ecological anthropology, 3–4, 83–84, 149 ecological economics, 52 economic anthropology: approaches to decision making, 21, 85, 86–87, 169; and common property, 149, 152; experiments in nonwestern field sites, 18, 169; and formalistsubstantivist debate, 4, 21, 23, 32–34, 36, 173n2; and household economic decisions, 132–133, 134;

200 Anthropology, Economics, and Choice

and models of choice, 165; relationship with mainstream economics, 4–5, 17, 81, 87 economics: approaches to choice, 1–5, 19, 21, 24–31, 33, 36, 118, 164, 165, 167, 171; approaches to decision making, 1–2, 6, 21, 23, 24–31, 61, 74–78, 87–88, 161, 163, 165–166, 171–172; assumptions of, 21, 27–31, 36, 93, 104, 137, 141–142, 169–170; and common property, 159, 160, 161; and constraints on decision making, 21, 24–31; and costbenefit analysis, 38, 39; defining of, 29, 32; experiments in, 84, 91, 93, 103–105, 114, 115, 117, 164, 165, 167; and generous behavior, 92; and global recession of 2008–2009, 15; and household economic decisions, 118–119, 134–140; and mathematical models, 17, 19, 24–25, 28, 36, 39, 74–77, 81–82, 83, 87, 88–89, 135, 137, 162–163, 164, 165, 166, 167; and migration from Mexico to United States, 14; neoclassical economic theory, 25, 33, 34, 46, 118, 135, 139, 140, 161; relationship with economic anthropology, 4–5, 17, 81, 87; and risk, 74–75, 81–82, 87, 88–89, 167; textbooks of, 28–31, 173–174n1; and uncertainty, 61, 76, 77–78, 87 economic systems, 30 education, and Oaxacan wood carvers, 69, 70, 72 Ellsberg, Daniel, 80 embeddedness concept, 21, 23, 32, 33–34, 36, 130, 165 Ensminger, Jean, 108, 109 environmental conservation, and property rights, 149–150, 153 ethnography: and common property, 149, 150, 162, 163, 171; complexity of, 93; and cultural differences, 3; and decision making, 5, 12, 18, 81, 86, 89, 163, 165, 166, 172; and de-

fining households, 18; and economic anthropology, 4, 99; and economics, 17; and fieldwork, 4, 23; and household economic decisions, 141; limitations of, 22; and models of human behavior, 93; and rational choice theories, 2–3, 5, 14; and risk, 86–87, 89, 168; value of, 5 evolutionary biology: and altruistic behavior, 99–101, 103, 104, 117, 156, 169; and cooperation, 18, 90–91, 97, 99, 101–102, 104, 116; and crosscultural economic experiments, 18, 90–91, 93, 99–103 evolutionary models, 83–84 evolutionary psychology, 3, 109–110, 113, 165 exchange value, use value distinguished from, 45 expected utility theory: and alternative outcomes, 6, 36; and costbenefit analysis, 38; and economic anthropology, 81; and economic models, 75, 87–88; estimates of, 7; and experimental economics, 84; and rational choice theories, 6–7; and risk, 18, 80, 82, 84–85, 87 experimental economics, 84, 90–93, 103–105, 164 experimental psychology, 164, 165, 167, 169 Experimental Technology Incentives Program (ETIP), 54 fairness: and experimental economics, 104–105; and markets, 105, 113; and self-interest, 104, 168; and Ultimatum Game, 105–106, 107, 109 Fehr, Ernest, 102–103 feminist theory: and autonomy and power of women, 133, 134; and economics, 17, 165; and household economic decisions, 118–119, 120, 134, 139–140, 141; and unpaid labor, 166 feudalism, 31 fieldwork, 4, 23

Index 201

firms: and microeconomics, 26–27, 30; production of, 135 Firth, Raymond, 32, 37, 130 Folbre, Nancy, 139 foraging societies: optimal foraging theory, 3–4, 45, 84; and Ultimatum Game, 107, 108, 113, 175n5 (chap. 4) Freudian psychologists, 165 Friedman, Milton, 27–28, 29, 36, 160, 161 game theory: and cooperation, 169; and economic experiments, 104, 106; and household economic decisions, 118, 138–139, 141; and prisoner’s dilemma, 176n4 gender relations: and household economic decisions, 118, 137, 138–139. See also men; women generosity: and Dictator Game, 115; and Ultimatum Game, 106, 107, 110, 175n5 (chap. 4) Gentry, Alwyn, 50, 51 globalization, 5, 23 global recession of 2008–2009, 15, 60–61, 76–77 Gnau of New Guinea, 108, 111 Granovetter, Mark, 33–34 gross domestic product, 39, 58 groups: decision making influenced by, 7, 14; supra-individual characteristics of, 3. See also household economic decisions Guatemala, 22, 32, 40, 41 Gurven, Michael, 109, 111, 115–116 Gutiérrez, Tonatiúh, 94 Hadza of Tanzania, 108, 110–111 Haldane, J. B. S., 100 Hamilton, William, 100, 101, 175n2 (chap. 4) Hammel, Eugene, 133 happiness, 6, 173n3 Hardin, Garrett: on common property, 142–144, 145, 148, 152, 154, 155, 156, 157, 159, 163, 170; on over-

population, 143, 148, 149, 151, 153; pasture analogy, 142, 143, 160, 176n1; reactions to, 149–151 Harford, Tim, 7–8, 9 Hayek, Fredrich, 162 Henrich, Joseph, 90, 92, 101–102, 103, 106–107, 111, 115 Henrich, Natalie Smith, 101–102, 103, 111, 115 herd mentality, 15–16, 78 history: approaches to choice, 165; and common property, 149, 163; cultural practices influenced by, 3, 4, 5; economic approaches to, 2, 17, 36; holistic methods of, 28 hog lots, 54–55 household economic decisions: anthropological approaches, 118– 119, 130–134, 135, 140, 141; and Belize, 119–123; economic approaches, 118–119, 134–141; and methodological individualism, 170; and Oaxacan artisan communities, 13, 127–128; and shared resources, 118, 120, 121, 129–130, 134, 135, 137, 140, 141; and women, 118–119, 120, 126, 133, 137 households: and access to common property, 147; concept of, 131, 170; defining, 18, 20, 118, 119, 120–130, 132–133, 134, 140; economic models of behavior, 59; as independent economic units, 132; interrelationship between production and consumption, 136, 140, 170; labor allocation in, 4, 43, 69–70, 71, 119, 133; and microeconomic analyses, 26–27, 30 housing bubble, 16 Hudson, Michael, 173n5 human behavior: and concept of rationality, 7–9; economic approaches to, 1, 17, 99 hunting for home consumption, calculating monetary value of, 44, 45, 46, 48

202 Anthropology, Economics, and Choice

Hurwicz, Leonid, 162 Huxley, Aldous, 40 imperialism, 22–23 inclusive fitness, 100 individuals, decision making by, 5, 7, 14, 17, 31, 32, 34, 35–36, 104, 144 industrial societies: and cooperation, 104; and households, 134, 141; unpaid labor in, 53–57 information gathering: converting uncertainty into risk with, 71, 74, 80, 89; and cultural evolution, 175n5 (chap. 3) input/output ratios, 22, 39, 43, 45 International Monetary Fund, 76–77 Jiménez, Manuel, 94–98 Johnson, Allen, 45, 87 Johnson, Craig, 150, 151, 158–159 Ju/’hoansi Bushmen, 114–115 Kahneman, Daniel, 16, 78, 79–80, 162, 167 Kantorovich, Leonid, 162 Karen (indigenous group), 52–53 Keane, Webb, 31 Keesing, Felix, 131 Keynes, John Maynard, 15, 77–78 kin networks: and anthropological analysis of economics, 130; and exchange of goods and services, 4, 113, 131; and households, 133, 140; kin selection explanations, 92, 99, 100, 102, 104, 116, 175n3 (chap. 4); as social structure, 32 Kling, Arnold, 161 Knight, Frank, 61, 72 Knightean uncertainty, 61, 76, 77, 80, 81, 83, 88, 167 Krugman, Paul, 161 Kunstadter, Peter, 132–133, 134 Kyrk, Hazel, 136 labor allocation, models of, 21 Lamalera of Indonesia, 108, 109 large-scale societies, 113–114

Leach, Edmund, 175n2 (chap. 4) Lee, Richard, 45 leisure, 6 Lesorogol, Carolyn, 162 Levitt, Steven, 114, 115, 161–162 List, John, 114, 115 Liu, Flora, 84 Lloyd, William Forster, 176n1 Lucas, Robert, 15 Lundberg, Shelly, 138 Lupton, Deborah, 85 Machiguenga, 45, 87, 90–91, 92, 106–107 macroeconomics: effect of policies on choice, 20, 21; and markets, 30–31; and no-till farming in Iowa, 62–63; as subfield of economics, 26–27, 30 Maine lobstermen, 152, 154–155 Malthus, Thomas, 149 Mankiw, Gregory, 27, 29, 174n1 (chap. 1) Mapuche of Chile, 111 markets: in Belize, 42; and cash crops, 48; and fairness, 105, 113; and formalist-substantivist debate of economic anthropology, 4, 33; and macroeconomics, 30–31; and microeconomics, 26–27, 30; and Ultimatum Game, 91, 99, 107, 109, 111, 113 Marx, Karl, 1, 105, 174n3 (chap. 2) Marxism, 17, 136, 139, 165 McCay, Bonnie, 150 McConnell, Campbell, 29, 174n1 (chap. 1) McElreath, Richard, 109, 111 Mellor, John, 47–48, 52 men, and household economic decisions, 118–119, 140–141 Mendelsohn, Robert, 50, 51 Mennonites, and Belize, 22, 41 methodological individualism: and formalists of economic anthropology, 4, 21; and models of household consumption, 140, 141; problems with, 169–171; and rational choice theory, 3

Index 203

Meyer, Donald, 75–76 Meyer, Jack, 75–76 microeconomics, 26–27, 30 migration from Mexico to United States: advantages and disadvantages of, 10–12; and artisans, 9–10; and cost-benefit analyses, 73; and household definition, 126, 129, 175– 176n4; and Oaxacan wood carvers, 68, 69, 70, 72, 73, 97; and rational choice theory, 9, 12–14; and remittances, 10, 11, 126; and wage discrepancy, 10 Miller, Steven, 122 modernization, and embeddedness, 33 motivations, 3, 18 Murphy, Arthur, 127–128 Muth, John, 15 Nash, John, 162 National Science Foundation, Cultural Anthropology Program, 91 natural selection, 100, 102 New Household Economics, 136–137, 140–141 non-timber forest products (NTFPs), 38, 48, 50–53, 174n6 no-till farming in Iowa, 62–63, 64, 71–72, 73, 74, 76, 88 Novak, Martin, 102–103 Oaxacan wood carvers: and cooperation, 95–97, 98, 117; and costbenefit analyses, 73; diversified activities of, 73; and household definition, 125–130; and household labor allocation, 69–70, 71; and information gathering, 74; openness and secrecy among, 93–97, 98; and risk, 69, 71, 72, 76, 88; and shortterm and long-term consequences, 72; and subsistence agriculture, 68, 69, 70, 72, 73; and uncertainty, 69, 71, 88 Olson, Mancur, 176n4 opportunity cost, 51–52, 53, 57, 58, 166 optimal foraging theory, 3–4, 45, 84

Orma of Kenya, 109, 111 Ortiz, Sutti, 81–82, 83, 85, 174–175n4 Ostrom, Elinor, 142, 149–151, 156–162, 163, 171, 176nn4–5 overpopulation, 143, 148, 149, 151, 153 Padoch, Christine, 123–124 paid labor: choices between paid and unpaid labor, 166–167; utility of, 17–18, 43, 47. See also unpaid labor; wage labor peasant societies, economics of, 32, 45–46, 174n4 (chap. 2) Peruvian Amazon: access to barreales in, 144–149; agricultural credit in, 65–67, 69, 71–74, 76, 88, 174n2 (chap. 3); economic conditions of, 175n3 (chap. 5); households defined in, 123–125, 126, 129–130 Peters, Charles, 50, 51 Plattner, Stuart, 91 Polanyi, Karl, 4 political economy, 3, 5, 22–23 political science, 2–3, 14, 164, 166, 168 Pollak, Robert, 138 positivism, 27 postmodernism, in anthropology, 5, 9 power: and access to common property, 144, 146–147; and dependency theory, 22; and household economic decisions, 133, 135, 137, 139–140, 141, 170; and individuals’ choices, 31, 34, 36, 118; of women, 133, 134 practice theory, 35 preferences, 3, 4 primates, and cooperation, 101–102 prisoner’s dilemma, 176n4 probability judgments: and alternative outcomes, 6, 7, 14–16, 18, 61, 75, 76, 87, 167; and cognitive psychology, 79–80, 85, 86, 88; and culturally learned conventions, 85–86; subjective probabilities, 75, 79–80, 84, 88 production for home consumption: calculating monetary values for, 17,

204 Anthropology, Economics, and Choice

43–44, 45, 46, 47–49; and feeding domestic animals, 44, 49 prospect theory, 79–80 public good: and cooperation, 116; moral appeals to, 143; public goods games, 109, 175n4 (chap. 4); and restrictions on access to commons, 145, 147–148, 156; and selfinterests, 92; and Ultimatum Game, 99 Quinn, Naomi, 84–85 rational choice theories: assumptions of, 104, 166, 169–170; and common property, 151; and cooperation, 92, 116, 168–169; criticisms of, 16–17; and economic anthropology, 4–5, 9; and ethnography, 2–3, 5, 14; and expected utility, 6–7; and formalist-substantivist debate, 32; limitations of, 25; preferences in, 3; and self-interests, 12–13, 29, 30, 31, 92, 168; strict versions of, 15–16, 18; and Ultimatum Game, 99, 106, 111; and utility maximization, 2, 143–144, 170 rational expectations hypothesis, 15 rationality: assumptions concerning rational behavior, 75, 82, 87, 106; broader meanings of, 7–9; economic idea of, 174n1 (chap. 2) real-world situations: applying economic models to, 9; and cultural norms, 115–116; and economic experiments, 61, 114, 115, 117, 164; ethnographical descriptions of decision making in, 18, 165; and group decisions, 118; and Ultimatum Game, 112 Reddy, Sanjay, 86–87 Redfield, Robert, 113, 114 Reid, Margaret, 136 replacement value, and unpaid labor, 56–57, 58, 166 representativeness heuristic, 78, 80, 85 resistance, studies of, 23, 35

ribereños. See Peruvian Amazon Richerson, Peter, 103 risk: and agricultural credit in Peruvian Amazon, 66, 67, 71; anthropological approaches to, 61, 81–87, 167–168; and behavioral ecology models, 83–84; and cash crops, 21, 43, 48–49; and cognitive psychology, 61, 78–80, 81, 85, 88, 89, 167; and cooperation and competition, 98; coping strategies for, 73–74; disciplinary approaches to, 74–87; and earning choices, 20, 21, 43; and economic anthropology, 86; and economics, 61, 74–76, 87, 88, 167, 169; and ethnography, 86–87, 89, 168; and expected utility, 18, 80, 82, 84–85, 87; and experimental economics, 84; and household economic decisions, 131; multiple risks, 71–72, 76, 88, 98, 168; and no-till farming in Iowa, 64, 71, 72; and Oaxacan wood carvers, 69, 71, 72, 76, 88; riskuncertainty continuum, 61, 64, 67, 71, 72, 76, 77, 88; theoretical approaches to, 61; variability in estimating, 72, 75 risk aversion, 4, 75–76 risk avoidance, 6 rules of thumb, 16, 84–85, 108–109 Rumsfeld, Donald, 77 Sahlins, Marshall, 8, 9, 132 Samuelson, Paul, 135, 136, 173–174n1 Sangu of Tanzania, 111 Santiago, Miguel, 94 scarce resources, 32 Scott, James, 34–35 self-exploitation, 57 self-interests: and common property, 19, 160; and communal resources, 19; and expected utility, 6; and experimental economics, 91; fairness moderating, 104, 168; and markets, 105; and Oaxacan wood carvers, 97, 98; and public good, 92; and ratio-

Index 205

nal choice theory, 12–13, 29, 30, 31, 92, 168; and Ultimatum Game, 115 selfishness: intercultural differences in, 106; mores against, 168; punishment of, 99, 103, 104, 107, 115, 117; and trait group models, 100; and Ultimatum Game, 110, 112, 113 Sen, Amartya, 24–26 Shiller, Robert, 167 Sigmund, Karl, 102–103 Skidelsky, Robert, 77–78 small-scale experimentation, 71, 74, 175n5 (chap. 5) small-scale societies, 91, 107–108, 113–114 Smith, Adam, 174n3 (chap. 2) Smith, Natalie. See Henrich, Natalie Smith social constructivism, 81, 86 social institutions: and decision making, 31, 32, 36; and publicly held resources, 19; and substantivist position on economic choices, 21, 23–24, 32–33, 34; supraindividual, 21 social networks, 33, 34 social organization, 32 social structure, 32 sociology: approaches to choice, 25, 31–35, 164–165, 166; approaches to cooperation, 168; economic approaches to, 2, 33–34; and migration from Mexico to United States, 14 Spence, Michael, 161 spiteful behavior, 100–103, 104, 110 Stepick, Alex, 127–128 structure, 31, 34 subjective probabilities, 75, 79–80, 84 subsistence agriculture: and behavioral ecology models, 84; in Belize, 40, 43, 44; calculating monetary values for, 38, 39, 43–44, 46–47, 48, 49, 166–167; and input/output analyses, 39; and Oaxacan wood carvers, 68, 69, 70, 72, 73; risks of, 49

subsistence production: imputing value to, 44–47, 48, 49, 57, 58–59, 166–167; and non-timber forest products, 50–51 substitute product value, 51, 52–53, 57 Surowiecki, James, 105 Taleb, Nassim Nicholas, 74, 174n3 (chap. 3) Tax, Sol, 32 Taylor, C. Robert, 76, 77 Thaler, Richard, 15 Third-Party Punishment Game, 115 Thomas, Duncan, 138 tit-for-tat strategy, 101, 105 Tracer, David, 111 tragedy of the commons. See common property trait group models, 100 tribal societies, economics of, 32 tropical rainforests: value of, 50. See also non-timber forest products (NTFPs) trust, 104, 105, 113 Tsimane of Bolivia, 109, 111, 115 Tucker, Bram, 84 Tversky, Amos, 16, 78, 79–80, 167 Ultimatum Game, 90–91, 92, 99, 105– 113, 114, 115, 169, 175n5 (chap. 4) uncertainty: and agricultural credit in Peruvian Amazon, 67, 71; and cash crops versus wage labor, 21; and cognitive psychology, 61, 88; converting into risk, 71, 74; and cooperation and competition, 98; coping strategies for, 73–74; disciplinary approaches to, 74–87; and economic models, 76, 88, 169; and ethnography, 89; and expected utility, 18; and global recession of 2008–2009, 76–77; and household economic decisions, 131; Knightean uncertainty, 61, 76, 77, 80, 81, 83, 88, 167; multiple uncertainties, 71–72, 88, 98, 168; and no-till

206 Anthropology, Economics, and Choice

farming in Iowa, 64, 71, 72; and Oaxacan wood carvers, 69, 71, 88; risk-uncertainty continuum, 61, 64, 67, 71, 72, 76, 77, 88; theoretical approaches to, 61 unpaid labor: calculating monetary values for, 17, 38, 39, 46, 47, 48, 51–52, 136, 166–167, 169; choices between paid and unpaid labor, 166–167; domestic labor, 38, 48, 53, 55–57, 58; economists’ ignoring of, 135, 140, 166; imputing value to, 56–59; in industrial societies, 53, 55–57, 134; and input/output analyses, 39; utility of, 17–18, 43; and women, 55–56, 136. See also paid labor unpredictable consequences, 13–14 unsubstantiated beliefs, 16 use value, 45, 49 U.S. Supreme Court, 35 utility: defining, 13; maximization of, 1, 2, 143–144. See also expected utility theory

wage labor: in Belize, 42, 43, 119; calculating monetary returns from, 20, 21, 43, 47; and Oaxacan wood carvers, 69, 70, 72, 73, 97. See also paid labor Wales, Terence, 138 Waring, Marilyn, 39, 55–57 Warner, K. E., 74 Wiessner, Polly, 114–116 Wilk, Richard, 34, 122 Wilson, E. O., 100, 175n3 (chap. 4) wind power, 55 Winking, Jeffrey, 115–116 Winterhalder, Bruce, 84 Wirth, Louis, 113, 114 women: and decision making, 134, 174–175n4; and household economic decisions, 118–119, 120, 126, 133, 137; and household labor allocation, 133; in Oaxacan artisan communities, 126; and unpaid labor, 55–56, 136 World Bank, 58 world systems theories, 5