Essays on Institutions, Natural Resources and Taxation

This thesis presents three essays that aim at analysing the impact of internal and external factors on institutions, which, in turn, may affect development outcomes. The first essay relates to the literature on civil society, aid and democratization. It investigates whether projects implemented by civil society organisations (CSOs) improve the democracy level of the recipient countries. In particular, it evaluates the effectiveness of projects funded by the United Nations Democracy Fund (UNDEF), a trust fund established in 2005 by the United Nations. An empirical analysis based on the propensity score matching method is implemented on a sample of developing countries. The treated group encompasses countries that benefited from projects implemented by CSOs and funded by UNDEF between 2006 and 2011. The findings indicate that the average treatment effect on the treated (ATT) is positive and significant only when countries receive UNDEF-funded projects for three rounds or more. This suggests that UNDEF should continue to fund civil society and non-governmental organisations, but it should also implement a long-term planning and solicit their projects over time. The second and the third essays contribute to the literature on the resource curse, the counter-intuitive finding that countries highly endowed in exploitable natural resources perform worse than those who lack this asset. Both studies confirm that natural resources may be a curse or a blessing for a country, depending on the quality of its institutions. In the second essay, the synthetic control method is used to compare the evolution of the democracy level of countries that experienced giant oil discoveries with the weighted democracy level of countries that do not incur the same event and have similar pre-event characteristics. Focusing on 12 countries that have reached the time after which the rate of oil discoveries starts to decline (peak of oil discoveries), it shows that the variation in oil endowment has a negative effect in the long run in most of cases, but countries with a high level of democracy in the pre-event period are not affected. The third essay analyses the effect of resource abundance on fiscal capacity, defined as the amount of taxes that a government could potentially raise given the structure of the tax system and its available powers of enforcement. Using panel methods, it tests two hypotheses: a) resource rents reduce the incentives to invest in fiscal capacity, thereby resource-rich countries have less developed tax systems; b) political institutions placing limits on the executive powers promote common interests and, thus, raise the incentives for investing in fiscal capacity. The negative effect of natural resources on fiscal infrastructures is therefore mitigated or neutralised in countries with a higher level of executive constraints. The empirical analysis demonstrates that resource rents are negatively associated with fiscal capacity, measured as the share of non-resource taxes on income, profits, and capital gains in non-resource total taxes. However, countries with a high level of executive constraints are able to neutralise or even reverse this effect, depending on the type of resource endowments. The paper gives also an insight into the specific channels through which natural resources work on tax systems, suggesting that they affect fiscal institutions that make the state accountable and transparent to its citizens.