Place of registration and place of residence: the detrimental impact of transport cost on electoral participation

Few studies have tried to assess the empirical impact of transport cost on electoral turnout. Unfortunately, these researches suffer from different limits, especially limites related to the endogeneity of the voting station localization. Our study, based on French data, overcomes the main usual empirical difficulties. In particular, the French case provides a very valuable opportunity for testing the impact of transport cost on individual decision of turnout, because voters can be registered in another municipality than their residential municipality. As such, some of them have to travel important distance in order to cast their ballot. And this distance is totally exogenous to the electoral manipulation of places of voting location potentially made by local administration. So, we show that distance, and in fine cost of voting, have a highly significant impact on electoral turnout: at average distance (122 km) a 1% increase of distance induces a 0.05% decrease at the first round of 2012 presidential election and 0.04% at the second round. This result is robust to many tests: if we change the empirical method carried out or the election studied or if we control for the weight of the largest distance, the results remain the same.