Commentary on Henrich et al.
Word Counts:
Abstract: 49 words
Main Text: 1000 words
References: 211 words
Total Text: 1260 words
Abstract: The data of Henrich et al., when combined with other research, suggest that sharing behavior probably varies systematically across cultures, situations, and individuals. Economic policies founded on recognition of this "radical contingency" would, I argue, nurture economic pluralism rather than attempting to bring the world under one system.
I have followed the project of Henrich et al. with great interest since attending a lecture about it a few years ago by Samuel Bowles, who amusingly compared the roster of authors to the cast of a Cecil B. DeMille epic. The reach of this study is truly impressive, as is the thoroughness with which the authors have addressed potential objections. This study is already a landmark in the joining of economic theory with anthropology, and, to a psychologist who studied under Amos Tversky, it appears to be the coup de grace in the behavioral critique of Homo economicus.
Henrich et al. emphasize the failure of the pure self-interest model across all the societies that they and others have studied, the greater variability across small-scale societies than has been seen in large-scale societies when procedures are held constant, and the importance of group membership and key group-level variables, as opposed to measured individual differences, as predictors of behavior. If we combine their study with others , however, I claim the picture that emerges is just that sharing behavior is radically contingent.1 Adopting the useful classification of effect types in the target article, I use "radical contingency" to refer to systematic variations in a behavior across all three of the following types of variables: (1) cultural groups, (2) situational contexts, and (3) individuals. Let us consider each in turn.
Cultural groups. An important contribution of Henrich et al. is that it demonstrates that sharing behavior in the games they studied varies widely across communities. This is crucial because previous studies had not revealed much cultural variation in particular for the ultimatum game. The variation in sharing behavior may be even stronger than claimed in Henrich et al. if we consider the economically-trained to be a cultural group, because such training has been shown to induce behaviors such as free riding (Marwell & Ames, 1981; Frank et al., 1993).
Situational contexts. Even within a community, very different behaviors may be evoked by changes in the situation or framing in which participants are given a task. Henrich et al. did not manipulate context in this way, though they do note that the ability of participants to see a task as similar to aspects of their daily lives may help to determine the response, and that similarities in daily experience and in such construals within a community may account for variation across communities. We know, from studies the authors cite, that large swings in the tendency to share can result from changes in presentational context (e.g. Hoffman et al., 1994). Liberman et al. (2004), for example, found a swing from 1/3 to 2/3 cooperation in the prisoner's dilemma when it was described as the "Community Game" instead of as the "Wall Street Game". Recent experiments have shown that subtle changes in presentation such as whether a set of preferences is presented in rank or pairwise format can strongly affect social preferences when criteria strongly compete (Davies et al., in preparation).
Individuals. A notable feature of all the data on sharing behavior is the substantial presence of within-group variation. Henrich et al. report failing to find reliable predictors of individual differences. Indeed, in public goods games, individual variation appears to be greater in large-scale societies than in those studied by Henrich et al., with bimodal percentages of students opting for the extremes (full and no contribution). It seems very likely that correlates of individual differences in small-scale societies could be found as well if one were to measure subjective variables such as attitudes and beliefs. In large-scale societies, individual differences may reflect adherence to ideologies.
A radical contingency model of sharing behavior requires going beyond the evidence in Henrich et al., but it also differs from their interpretation by, for example, including the possibility that a norm of self-interest can prevail within a community. There is mounting evidence for the importance of such a norm in contemporary U.S. culture (Miller, 1999), and Ferraro et al. (2005) have argued that the assumptions of economics as a discipline may bring about such norms as self-fulfilling predictions, by, for example, shaping institutional arrangements. Henrich et al. also do not emphasize situational and individual variables.
If the propensity to share is viewed as radically contingent, the consequences for policy appear sharply at odds with current practice. Assuming that the selfishness axiom holds universally bolsters efforts to impose economic orders such as the "Washington consensus", often through transnational institutions and/or military intervention. But if, as Henrich et al. indicate, locally varying conditions select for different norms, and Homo economicus does not characterize people generally, then national or global institutions might better foster a plurality of economic arrangements. The autonomous municipalities associated with the Zapatistas in Chiapas (Mexico) are examples of such arrangements.
Combining (a) the observation from earlier data that ultimatum game behavior shows little variation across cultures among university students with (b) the greater variety of behavior seen in small-scale societies, implies that a set of shared assumptions has emerged across large societies. This seems related to globalization. Sociologists debate whether such convergence reflects inherent commonality of preferences or the influence of one dominant world system (see, e.g. Chase-Dunn & Grimes, 1995; Meyer et al., 1997). But the fact that it seems to be happening is at odds with the economic diversity that Henrich et al. characterize as resulting from culture-gene co-evolution. If there is no diversity, then there can be no selection.
System globalization also makes it harder to accommodate individual preferences for arrangements that could otherwise be satisfied through voluntary association. In such a world, behavior is less likely to reflect individuals' and groups' true beliefs, preferences, and constraints, simply because there is little room for variation. The challenge for an economics rooted in an understanding of radical contingency is to provide for both diverse arrangements and the translocal coordination necessary to foster human freedom and to sustain our global ecology.
Notes
1. This term has appeared in various disciplines with somewhat different meanings previously. I am not alluding to any particular previous usage.
References
Chase-Dunn, C. & Grimes, P. (1995). World-systems analysis. Annual Review of Sociology 21:387-417.
Davies, T., Shah, R., Ling, K. & Trochet, R. (in preparation) Intuitive preference aggregation: Tests of independence and consistency.
Ferraro, F., Pfeffer, J. & Sutton, R.I. (2005). Economics language and assumptions: How theories can become self-fulfilling. Academy of Management Review 30:8-24.
Frank, R.H., Gilovich, T.D. & Regan, D.T. (1993) Does studying economics inhibit cooperation? Journal of Economic Perspectives 7:159-171.
Hoffman, E., McCabe, K., Shachat, K. & Smith, V. (1994) Preferences, property rights, and anonymity in bargaining games. Games and Economic Behavior 7:346-380.
Liberman, V., Samuels, S.M. & Ross, L. (2004) The name of the game: Predictive power of reputations versus situational labels in determining prisoner's dilemma game moves. Personality and Social Psychology Bulletin 30:1175-1185.
Meyer, J.W., Boli, J., Thomas, G.M. & F.O. Ramirez (1997) World society and the nation-state. American Journal of Sociology 103:144-181.
Marwell, G. & Ames, R.E. (1981) Economists free ride, does anyone else? Journal of Public Economics 15:295-310.
Miller, D.T. (1999) The norm of self-interest. American Psychologist 54:1-8.