Essays in political and public economics.

Summary This thesis is in the field of Political and Public Economics. Through the use of quasi-experimental methods and administrative data, its goal is to achieve a better understanding of the behavior of citizens and elected officials.The first two chapters, in collaboration with Vincent Pons, study voter behavior. The first chapter assesses the extent to which voters vote expressively or strategically and the consequences for electoral outcomes. We exploit the threshold that determines the qualification of candidates for the second round of the French legislative and cantonal elections. Using the Regression Discontinuity Design (RDD) method, we compare the electoral results according to whether two or three candidates are running. The results show that the presence of a third candidate significantly increases voter turnout and reduces the number of votes obtained by the first two candidates. More importantly, in about one-fifth of the elections, the presence of a third candidate causes the loss of the candidate who was ideologically closest to the first two. We thus show that a large fraction of voters prefer to vote expressively rather than strategically and that, in this context, majority rule often leads to suboptimal results.The second chapter explores the type of information that may affect the behavior of voters and candidates, by assessing the impact of a specific piece of information: the ranking of candidates. Using an RDD, we isolate the impact of the ranking of candidates in the first round on the results of the second round, in French elections where up to three or four candidates can qualify for the second round. We show that being marginally ranked first, second or third in the first round significantly increases the probability that a candidate will continue to run in the second round and win. The impact remains significant even when only two candidates qualify (and thus there is no room for strategic voting), suggesting that the desire to vote for the winner plays a key role in voters' voting decision.The third chapter studies the behavior of elected officials and in particular the reluctance of mayors to cooperate on public policy. I exploit the 2010 law, which in France forced non-integrated communes to join an intercommunality. I study the causal impact of their integration using the double difference method. My results show that communes resisted in order to avoid the local economic consequences of integration: urban communes resisted in order to avoid the construction of new housing, while rural communes resisted in order to avoid the loss of local public services. I note first that the communes forced to join an intercommunality experienced a significant increase in the number of building permits issued per year. This effect is mainly manifested in densely built-up communes that face a high demand for real estate: communes where more construction would increase congestion, and where residents are therefore likely to be subject to the NIMBY (Not In My BackYard) syndrome. Secondly, I show, through the exploitation of data on municipal day-care centers and libraries, that rural municipalities have suffered a decrease in the number of local public services. Finally, the results show that communes forced to integrate benefited from higher tax revenues and better access to public transport, but that these benefits were not sufficient to offset the negative effects of integration.