We analyse political competition when the source of political conict stems from groups that have di¤erent misspecied stochastic models of how policies a¤ect outcomes. We focus on groups that di¤er in the simplicity of their model of the world; a “simple” group,that believes that only a subset of policy instruments a¤ect outcomes, competes with another group that believes a wider range of policy instruments matter. Whoever wins the election chooses that periods policies. Upon observing the policy outcomes, both groups update the weight they place on the di¤erent policy instruments, albeit within the connes of their model. We show that in all steady state equilibria the simple group is always in power with a strictly positive probability. Moreover, it always espouses policies that are more extreme compared to those of the other group, on all policy instruments it considers.

Please sign up for meetings here: docs.google.com/spreadsheets/d/1Tf4YtDeDdmv3Dv379EyhWTL6lszs2Dy6yiff7yJeJAY/edit#gid=0