The Spatial Political Economy of Discontent

The recent rise and distinct geography of populism highlights the need for high resolution data on the economic and political landscapes and improved spatial political economy models that explain their interrelation. This paper shows that divergent development generates political externalities in lagging regions. To do so, it develops a dynamic spatial political economy model that integrates redistributive taxation and agglomerated economic growth in a standard economic geography framework. It finds that divergent development induces skill-biased labor mobility towards faster growing locations, simultaneously reducing their willingness to pay redistributive taxes and increasing their electoral influence on redistributive policy. To empirically validate and calibrate the model, the Spatial Political Economy in Europe Database (SPEED) is introduced, containing newly georeferenced electoral maps, political party classifications and gridded (per capita) GDP estimates for most European countries in the 17th release of the Constituency-Level Electoral Archive (CLEA). Instrumental variable regressions exploiting geographically-determined differences in economic growth potential confirm a strong constituency-level causal relation between underdevelopment and radical vote shares in the past two centuries. Counterfactual simulations suggests that policies that enhance labor mobility or income redistribution may both increase radical vote shares at least in the short run, as they risk fueling backlash in lagging and leading regions respectively.