Résumé : This thesis offers three papers on elite power and its macroeconomic impact. In the first chapter (co-authored), we investigate how internal distribution motives can affect the implementation of an important macroeconomic policy: capital controls. In order to do this, we study one of history's largest debt repatriations, which took place under strict capital controls in 1930s Germany, providing a wealth of quantitative and qualitative historical evidence. We show that the authorities kept private repatriations under strict control, thus avoiding detrimental macroeconomic effects, while allowing discretionary repatriations in order to reap internal political benefits. We formalize this mechanism in a model in which elite capture can affect optimal debt repatriations and the management of official reserves under capital controls. The second chapter introduces elite power in a model of endogenous capital controls and foreign debt default, where the elite may a have a disproportionate effect --relative to its population share-- on the social planner's decisions. The social planner allocates scarce foreign exchange reserves between the private sector, public good provision, and private debt repatriations, an operation that only benefits the elite. Even when complete default on foreign debt is the best policy for the country as a whole, some debt will be repatriated, due to the elite seeking arbitrage profits and interlinkages between international trade and domestic production.The third chapter proposes a model of endogenous quality of institutions (QI), chosen as a costly investment by a productive elite. The elite produces output similarly to the non-elites, but can also set distortionary taxes on production and divert tax receipts for its benefit. A credit market with imperfections determines the allocation of capital - where funded projects differ by productivity. QI influences both the frictions on the credit market (and therefore productivity) as well as state capacity in managing taxation. The elite can choose policies (tax rate and investment in QI) that reflect different development paths, ranging from a weak state (no taxes, lowest QI), pure production income economy (no taxes, high QI), inefficient rent extraction (high taxes, low QI), to intermediate cases of both rent extraction and investment in QI. The emerging non-linear relation between investment and QI shows net capital inflows or outflows to be uncorrelated to QI and that domestic policy choices are affected by external factors such as the prevailing world rate of return. This result shows that QI should not be treated as an exogenous parameter in models of economic development and in policy making, as it can hide the reverse causality between capital flows and domestic policy choices, where a country's development path can be influenced by capital flows, not just the opposite. This calls for a greater consideration of domestic political constraints in development policies and in modelling open economies.