A recent paper by Edward Cartwright, published in the Journal of Sports Economics, looks at the extent to which social norms can reduce or eliminate doping in professional sport. It does so by applying models of guilt aversion and reciprocity from psychological game theory.
In the model of reciprocity first-order beliefs are critical. These are an athletes beliefs about she thinks others will do. In short, if an athlete thinks others will dope then she will dope. Conversely, if an athlete thinks others will not dope, and she is sufficiently reciprocal, then she will not dope. Reciprocity can lead to an equilibrium with no doping but only under strong conditions. In particular, we need every athlete to value reciprocity and this seems unlikely. One ‘selfish’ athlete is enough to break any social norm against doping.
In the model of guilt aversion second-order beliefs are critical. These are an athletes beliefs about what she thinks others expect her to do. In this case, if an athlete thinks others expect her to not dope, and she is sufficiently guilt averse, then she will not dope. Again, this is enough to create an equilibrium of no doping. Crucially, this equilibrium does not unravel if there is one selfish athlete. Guilt means an athlete will not dope irrespective of what she expects other athletes to do.
The results in the paper have some interesting policy applications. Primarily, they suggest that we should focus on creating positive second-order beliefs. In practical terms this means bringing athletes up with a belief that others expect they will not dope. Doping in sport is currently framed, through, for instance, the whereabouts system, in a way that almost suggests guilt until proven innocent. This could easily lead to negative second-order beliefs in which an athlete believes others expect her to dope. More trust and positivity around young athletes may, therefore, be warranted to generate social norms against doping.
The full version of the paper is available here.