Does the winner take it all? Redistributive policies and political extremism

Abstract

We show that regional heterogeneity of underlying fundamentals (e.g. economic history, geography, natural resources, social capital) can lead to extreme voting in federations. When the outcome of federal policies { such as transfer schemes, market regulation, migration targets { depends on these fundamentals, the set of regions that benefits or loses from such a policy is fixed exogenously (for each given policy). This creates a strategic incentive for the median voter to elect extremely protective federal representatives, who will distort redistribution in favour of their constituency by influencing the choice of the policy magnitude. Interestingly, the benefits of selecting tough negotiators outweigh those of belonging to the ruling coalition. We test our predictions by looking at parties' performances at national and European Parliament elections from 1990 onwards, and find that strategic voting is indeed U-shaped: winning and losing member states vote more extremely than those in the middle. Our online survey provides further evidence.