This thesis tackles issues related to regional integration, trade costs and aid, with empirical work related to the East African Community (EAC). The common thread is the impact of various types of trade costs on the structure and functioning of the economies of EAC member states. The first chapter introduces the literature and chapters 2, 3 and 4 constitute the core of the thesis. Chapter 2 develops a three-good, two-country duality-based general equilibrium model to investigate the effects of different types of aid and preferential trade on welfare and relative prices. The model is innovative in two ways: (i) a regionally tradable good is introduced, the price of which is determined endogenously, (ii) a regional infrastructure good, bought with aid monies, is brought in which lowers trade costs within the region. Using comparative statics, the properties of the model are explored in terms of the effects of tariff and aid shocks on welfare and relative prices. Chapter 3 develops a Computable General Equilibrium (CGE) model calibrated on Uganda and Kenya to gauge the importance of chapter 2’s results. The regional spill-over is now modeled as regional public capital serving as an input in both countries’ production functions. The simulations show how Kenya effectively exports some of the standard aid-induced real exchange rate appreciation to Uganda through a regional trade channel. Distributionally, Kenya’s urban and Uganda’s rural households win—which corresponds to regional comparative advantage patterns. Abolishing the intra-regional tariff increases welfare in Uganda and reduces it in Kenya, showing the ambiguous welfare results of Customs Unions known since Viner. Chapter 4 gauges the importance of border effects in Eastern Africa by testing the law-ofone- price (LOP) hypothesis on a consumer price data-set covering 24 goods in 39 cities in 4 countries. Using level regressions a significant border effect is found, whilst distance also plays a big role, both between and within countries. Neither the nominal exchange rate, nor non-tariff barriers (NTBs) reduce the border effect very much, even though both variables are significant. Looking at specific goods, markets for staple foods are the most integrated. As for the impact of the Customs Union between Tanzania, Uganda and Kenya (since 2005), there is a positive integration effect for the Kenya-Uganda border. Finally, Kenya’s political crisis at the end of 2007 can be linked to higher departures from the LOP throughout the region and can thus be said to have had clear knock-on effects for the landlocked EAC countries that depend on it as a transit country.

La réussite du processus d'intégration régionale dépend essentiellement de sa réception et de sa sanction dans les ordres juridiques internes. la crainte principale réside dans l'ineffectivité des règles communes dans les Etats parties. Ceux-ci peuvent en effet refuser de tirer toutes les conséquences du transfert des compétences qu'ils effectuent librement au profit des organes communautaires.Si la CEMAC peine aujourd'hui à trouver le juste équilibre entre les objectifs ambitieux qu'elle affiche et la capacité des Etats membres à les réaliser en commun, l'OHADA arrive, dans un environnement difficile, à tracer une voie qui est porteuse d'espoirs. / X