Evolutionary Psychology and Economics Experiments
From Burnham and Kurzban’s open critique of Henrich et al (2005):
Tinbergen provided a framework that explained both maximizing behavior and persistent failures to maximize in particular settings (Tinbergen 1968), and this framework has become the standard for the study of the behavior of all animals except humans.
As an example of Tinbergen’s approach, consider work showing that herring gulls fail to maximize in environments constructed to include artificial eggs. Careful experimental studies were able to pick apart the mechanisms that gulls use for egg selection (Baerends & Drent 1982a; 1982b). These mechanisms (1) arose by natural selection, (2) advanced the inclusive fitness of the individuals in natural environments, and, crucially, (3) failed spectacularly in particular, artificial settings.
Tinbergen’s theory and its applications suggest a similar explanation for human economic behavior that fails to maximize, and this approach is already providing novel insights. For example, it has long been known that humans fail to maximize in many experimental settings involving time discounting (Ainslie 1974; Rachlin 1970). A recent study manipulated the mechanism of discounting and caused an increase or decrease in the deviation from maximization (Wilson & Daly 2004). By illuminating one aspect of the cognitive architecture behind discounting, this work suggests that the apparently puzzling economic behavior is simply caused by adaptive mechanisms interacting with particular and peculiar environments.
Similarly, a growing body of research investigates the mechanisms that modulate prosocial behavior as a function of anonymity. Models based on signaling (Smith & Bliege Bird 2000) or reputation (Panchanathan & Boyd 2003; Trivers 1971) predict the adaptive value of psychological mechanisms sensitive to cues of anonymity. Indeed, people modulate their behavior as a function of anonymity (Burnham 2003; Hoffman et al. 1996b; Rege & Telle 2004) and prosociality is more likely when actions are observed.
In fact, the data in the target article would not be surprising at all if they took place in a repeated, non-anonymous setting with an ability to generate reputations. Is this failure to maximize – like that of gulls with artificial eggs and people discounting in the laboratory – caused by mechanisms interacting with specific environmental cues? If so, it might be possible to create prosociality using cues to social presence. In particular, a powerful cue is likely to be the presence or absence of eyes, which is used to modulate behavior across many species (Call et al. 2003; Hampton 1994; Hare et al. 2001).
This hypothesis that eyes will produce prosocial economic behavior has been tested and confirmed in two studies. Contributions to a public good game increased by 29% in the presence of human eyes (Burnham & Hare, in press; see also Kurzban 2001). Similarly, contributions in a dictator game were increased 32% by the presentation of eyespots (Haley & Fessler 2005). We are optimistic that the continued application of Tinbergen’s framework to human economic decisions may provide both proximate and ultimate explanations for prosocial behavior.