preview

The Rational And Self Interested Agent Used As A Model Of Human Behaviour

Decent Essays

Meet Homo economicus, the perfectly rational and self-interested agent used as a model of human behaviour in neoclassical economics. Lacking the emotion and cognitive biases that typically cloud optimal decision making, Homo economicus is able to consistently pursue self-interest without any regard for the welfare of others. The troubling observation here is that economists have been modelling human behaviour in a manner effectively consistent with psychopathy. The classic, albeit Hollywood inspired psychopath – intelligent, capable of self-control, driven to succeed and dominate others – is disturbingly similar to our description of Homo economicus. Indeed, using a variety of anonymous, one-shot cooperation games, Yamagishi et al. (2014) …show more content…

This is problematic, and invites the following questions: 1) How to justify modeling economic outcomes in a manner not accurately reflecting most real-life decision making, and 2) How to arrive at an economic, or even evolutionary explanation for our (nearly) global violations in utility maximization. I address these points in brief below.
Economists have long known people aren’t the tidy, rational, utility maximizing non-cooperators that they’re modeled as in neoclassical economics. For this reason, the early behavioral economic literature was concerned with modifying existing models to better reflect psychological factors (e.g. Kahneman & Tversky, 1979). Most people, for example, do not follow Bayesian statistics in situations under risk, and so prospect theory serves as a useful addition when modelling loss aversion. Thus the assumptions underlying Homo economicus can be interpreted as a pragmatic choice that allows economists to make simplifying approximations appropriate for the problem under review. This notion, that economists are conscious of the empirical limitations of their models, suggests that Homo economicus can and should serve only as a preliminary step towards more sophisticated representations of real-life decision making. If it’s understood that the assumptions underlying the models are at best useful approximations, then perhaps the issue is with the breadth of application rather than the models themselves.
Though why is it that we don’t

Get Access