Do Elections Build States? Evidence from Liberian Electoral Administration

• In Event: State Capacity and State Building

Abstract

In contexts where the state otherwise has limited reach, effective electoral administration permits the projection of state authority and increases levels of state-citizen interaction. I examine how increased access to electoral participation consequently affects citizens’ attitudes towards formal political institutions. Studying the geographical expansion of polling places between the Liberian elections of 2014 and 2017, I leverage unusually rich data sources in a difference-in-differences design. First, to estimate the effects of reduced distance to vote on electoral participation, I use record linkage techniques to match large-scale deanonymized administrative surveys with the complete voter registers from each election. Second, I employ a geolocated panel survey of citizens enumerated shortly following each election to examine the effects of electoral access on beliefs and attitudes. Third, to account for confounding in the targeting of polling places, throughout I compare individuals in localities close to a new polling place to those close to a ‘potential’ new polling place, which I construct drawing on internal documents of the electoral commission. I find that increased electoral access notably increases citizens’ trust in formal state institutions, much more so than in elected offices of government. The results suggest that effective electoral administration in weakly-institutionalized settings can significantly contribute towards broader state-building outcomes.

Author

• Jeremy Bowles, Stanford University