Economists and Anthropologists—Can the Twain Meet?

Some speculations on methodological differences between two disciplines and the scope for some coordination

I met M.N.Srinivas, the doyen of Indian social anthropology (identified with sociology in India), when I was teaching at Delhi School of Economics. He used to classify social scientists in two groups, those who have a ‘book view’ and those who have a ‘field view’ (i.e. those who go out to the ‘field’ to find out the actual conditions of people), and as an anthropologist he had a partiality for the latter. So, I was pleased when one day he told me that I was a rare economist who had a ‘field view’. He was referring to the frequent village surveys I was carrying out around that time (in the 1970s and 1980s, in collaboration with Ashok Rudra, the statistician/economist) in a set of 110 (randomly selected) villages in West Bengal.

Srinivas probably did not know that just a few years before then I was very much in the ‘book view’ group. My Ph D dissertation was entirely theoretical (it can be said that I was focused then on learning the concepts and tools of non-linear differential equations and optimal control theory—slogging away on the work of the blind Russian mathematician, Lev Pontryagin—and applying them to problems of economic growth theory and international trade). I continued that kind of research when I went on to teach at MIT and published several papers on those topics in the top 5 journals. Then there was a big transition in my professional life. Much to the surprise of my colleagues, I decided to leave MIT and to go back to India to work with Indian data. However, I soon found out that the official data that I could procure there did not ask many of the questions on economic relations in rural land, labor and credit markets that I was interested in. So, I started collecting my own data from village surveys, which very few economists, even development economists, used to do those days. One economist told me that what I was doing was more like economic anthropology (this may have been his polite way of telling me that I was not a proper economist anymore).

Around this time, I also started reading up quite a bit on the work by Indian social anthropologists. At Delhi School of Economics, I interacted a great deal with my friend, the renowned social anthropologist André Beteille (who had a French father and a Bengali mother). I closely studied his ethnographic work on class and caste in Tamil Nadu. He also told me a lot about his teacher, the anthropologist Nirmal Bose (who may be known in the rest of India more as a close intellectual associate of Gandhi in his later days), and encouraged me to read his books. André even gave me a copy of an insightful book by Bose on the ‘The Structure of Hindu Society’ that he had translated from Bengali with an Introduction and critical notes. Later, I probably was the only economist who was asked to contribute to a book of essays in André’s honor edited by Ram Guha and Jonathan Parry.

As I was discussing with social anthropologists and reading up their work, I soon became aware of some large differences in the methods that economists and anthropologists follow. I also found out that these two groups hardly talked to one another on substantive issues even when they were working on related themes, say on the peasant society, and in the rare occasions when they tried to talk they often talked past one another. The thought came to me that maybe I should put them together in a conference where they’ll not be allowed to talk past each as far as possible. With the sponsorship of the Social Science Research Council of New York, I organized the first such conference at Bangalore in 1984, and following up on this conference I edited a book titled “Conversations between Economists and Anthropologists” in 1989. This book was about how economists and anthropologists look at the problems of measuring economic change or development in rural society.

Nearly two decades later my Berkeley colleague Isha Ray and I co-edited a sequel, a sort of Conversations II, following up on a conference in Goa that we had organized. This time the focus was on how economists and anthropologists look at issues of cooperation in the management of the local environmental commons (like forests, grazing lands, fisheries and irrigation). This volume is titled “The Contested Commons: Conversations between Economists and Anthropologists” (Blackwell, 2008). In the Introduction to this volume, we speculated about some methodological differences between the two disciplines. Here I’ll draw upon some of those speculations.

Reading the work of economists and anthropologists at least three differences in their general methodological approach strike one:

· Individual Autonomy versus Social Embeddedness

Much of Economics starts with the assumption of an autonomous individual with a given set of stable preferences and explains social phenomena in terms of individual choices and motives. Such individuals and explanatory devices are, however, quite alien to the anthropologist. To the latter, individuals are deeply embedded/situated in social and political relationships governed by norms and moral commitments, and so not all social phenomena are reducible simply to individual characteristics. Preferences often reflect the inner workings of culture and power in society. Autonomy in a social vacuum is not meaningful, as individuals are taken as relational creatures with multiple allegiances and overlapping identities.

In the context of the public use of common environmental resources, for the economist free riding in the form of littering or polluting or degrading is quite understandable, even ‘normal’, when the individual calculates the cost and benefits of a given individual action. But, as my friend the sociologist Erik Olin Wright pointed out in his contribution to our book, to the sociologist/anthropologist such action is ‘pathological’ -–sociologists in such contexts use a term, coined by Émile Durkheim, ‘anomie’ meaning normlessness-- as it violates the basic social understanding that some norms are part of the shared collective commitment and are not subject to individual cost-benefit calculation.

Recent advances in behavioral economics allows for the salience of some cases of departures from given preferences, with their being ‘endogenous’ or subject to influence by social forces, and similarly cases of departures from rational, self-regarding behavior allowing for cognitive failures and behavior influenced by, say, cultural norms of reciprocity and fairness. But economists will want these departures to have some patterns, so that economic propositions can have some predictive power. Also, while economists consider the constraints of norms and conventions that emerge from a social process (maybe even as the unintended consequences of thousands of uncoordinated decisions), they are usually reluctant to give up the agency that they believe resides in the individual, at least not to the extent that the anthropologist would. Even when preferences are taken as endogenous, economists’ models will find it difficult to grapple with cases where preferences are not fully known before action but develop through pragmatic social interaction, something anthropologists can easily live with.

· Outcomes versus Processes

Economists’ empirical investigations, including those using the most sophisticated econometric methods, are confined to studying outcomes, they’ll seldom give you any empirical clue about the process by which those outcomes have been reached, or the mechanisms and the relational dynamics that sustain them. Anthropologists, on the other hand, intensively study the people who are involved in these processes and relations, often live with them and ask them why they did what they did, and have insights from the field that are more grounded than the theoretical speculations behind economist’s models. In particular, the economists’ empirical exercises are insufficient to capture the power dynamics behind particular outcomes; beyond measurable inequalities they cannot easily capture the various material and symbolic dimensions of power that are articulated through social institutions or how the latter lead to people often internalizing the limits of what is permissible under a given power structure.

· Parsimony versus Complexity

As in physical sciences, economists often use the principle of parsimony in looking for the minimum assumptions under which something can be explained. This is primarily because economists are often looking for generalizability, and making the least assumptions allows them to bring a lot of things under the same tent, but this also gives their quest a bit of elegance. Sometimes this is called a ‘thin’ theory of human action, in contrast with the ‘thick description’ of the anthropologist. Also, with the economist’s interest in policy they want to change one variable at a time, keeping the others the same, so that they can zero in on the effect of a particular policy change. This kind of abstraction is often unacceptable to anthropologists immersed in the messy, grubby reality at the ground level. They think pursuit of parsimony makes one miss the multi-dimensional multi-layered nature of reality, the context-specificity or situational contingency of particular problems, and the path-dependence of history. If this makes generalizability or policy advice difficult that does not particularly bother the anthropologist in their frequent concern more with the particular and the unique.

In spite of these methodological differences there is a great deal of scope for coordinating and combining the respective strengths of the two disciplines. For example, as carrying out field surveys is now more common among economists than before, they can benefit if before launching large-scale surveys, they talk to anthropologists and sociologists (or even think of including the latter in their survey team) who usually have intimate knowledge of the area and the people. This may help in deciphering what are the important questions to ask and what is the appropriate local way of framing the same generic question. And after the survey, interaction with them may help the economist in understanding the mechanism or process through which the observed/measured data generate a particular finding.

Similarly, anthropologists may sometimes also be interested in generalizations, beyond local particularities, and may thus benefit from getting involved in issues like sampling design or statistical inference. Some anthropologists have become famous for generalizations about society, even though they garnered their original insight from studying a particular village or community. Take M.N. Srinivas, the social anthropologist I started this post with. He is most well known for his idea of the process of social change in India--his idea of what he called ‘sanskritization’, the process by which lower-caste or tribal groups attempt to raise their social status in the caste hierarchy by adopting the customs, rituals, beliefs, and lifestyles of upper-caste groups. Even though he may have got his idea from studying the village he called ‘Rampura’, he clearly meant this idea to be more widely applicable beyond that village.

One example of a creative way of combining the strengths of Economics and Anthropology was suggested by the economist Vijayendra Rao: what he called ‘participatory econometrics’ includes, apart from participant observation by the researchers, the participation by the respondents or the people who are being studied in the design of the large-scale survey itself, involving intensive focus group discussions, and also participatory appraisals by them of the researchers’ findings. This is one way of getting better insights than traditional econometric methods, and yet the findings can be more generalizable than the traditional case study of the anthropologist.

Even apart from field surveys, analytical thinking in both disciplines can gain from each other. I have seen in Berkeley how my brilliant friend and colleague George Akerlof (a recipient of the Nobel Memorial Prize in Economics), who for a time attended courses in the Sociology Department and used insights from sociological concepts of group norms and identities in writing path-breaking papers in labor economics and human decision-making in families, schools, organizations and workplaces.