Four Essays on Fiscal Decentralisation and Secessions.

Summary Between 1945 and 2008, the number of internationally recognized countries increased from 74 to 193 (Spolaore, 2008). More recently, several countries have experienced a growing wave of decentralization. In OECD countries, for example, the number of subnational governments reached 140,000 in 2014. Moreover, these subnational governments are increasingly influential in these countries (OECD, 2014). Given these trends toward increasing decentralization, this thesis studies two aspects of it: tax competition, and endogenous border choice. In terms of tax competition, this thesis studies the auctioning of new investments by a multi-location firm. The purpose of this analysis is to study the strategic behavior of the firm in this type of competition. Indeed, contrary to the existing literature that only considers firms that produce in one location, the first chapter of the thesis shows that the firm can modify the allocation of its investments by differentiating them, to attract higher subsidies. In the second chapter, the thesis studies how the addition of infrastructure costs for regions prior to the auction affects competition between regions as well as the firm's behavior. In terms of endogenous border choice, this thesis provides two analyses: one empirical, and one experimental. In the third chapter, the thesis studies the decision of voters in 213 Quebec cities to leave a municipal merger that was imposed on them a few years earlier. The analysis reveals that voters choose to secede more when the income and language differences between their city and other cities in the same merger are greater. The analysis also reveals that these two factors are not independent. Indeed, income differences have a more pronounced effect on the secessionist vote when language differences are also high. Given the importance of language as an ethno-linguistic group in Quebec, the results in this chapter suggest that voter choice is sensitive to ethnic differences, and not just to differences in taste for public goods, as suggested by Alesina, Baqir and Hoxby (2004). Finally, the last chapter presents the results of a laboratory experiment on the link between decentralization and secession. The literature on the subject suggests the existence of two contradictory effects. Decentralization could counteract secessionist movements by allowing regions to make more decisions at a local level, but it could also provide additional resources to secessionist movements, thus reinforcing the tendency towards separation. The results of the experiment show that the total effect of decentralization is to decrease the probability of voting for secession.