Government corruption and economic growth

This dissertation examines the role of government in economic growth. Specifically, I investigate the effects of corruption and strategic interaction between corrupt governments on growth. The dissertation consists of three chapters. The first is a broad overview of the literature on corruption, economic growth and fiscal decentralization. I first outline the various approaches used in the study of corruption by economists. I review the theoretical and empirical literature on corruption and economic growth, with a particular focus on models where issues of decentralization play a part in the debate on corruption. Although this literature is not vast, it is of particular interest as my third chapter explores the consequences of decentralization on growth in the presence of corrupt governments.;In the second chapter, I study the consequences of corrupt government behavior on an economy that is closed. I first allow this economy to exhibit transitional dynamics. The aim is to ascertain how this type of government behavior affects an economy's transition to its steady state, how a corrupt economy's steady state compares to that of a benevolent economy and how a change of regime from one type of government to another affects the economy. The results suggest that the presence of the corrupt government lowers both the steady state and the transitional levels of capital, consumption and output. In addition, the corrupt economy's transition to the steady state is slower than that of the benevolent economy. I also show that depending on the elasticity of inter-temporal substitution of consumption, a regime change from a benevolent economy to a corrupt economy can lead to a short term reduction in consumption. I then allow long run growth in this economy and study the consequences of corruption on economic outcomes. As would be expected, it is shown that an economy inhabited by a corrupt government has lower levels of growth, consumption and savings.;The third chapter explores the consequences of strategic interaction between corrupt governments in an endogenous growth model. Capital is assumed to move freely between two regions, each inhabited by a self-seeking government and a competitive private sector. The model is solved for both Nash and cooperative equilibria and the results are compared to those of two benevolent governments. This allows a comparison between a decentralized and a centralized corrupt government. Unlike the typical result in the tax competition literature, inter-governmental competition leads to an increase in public good provision which is welfare improving and decreases the balanced growth levels of corruption while increasing the growth rate. This result provides strong evidence that decentralization may be an effective constraint on corrupt government behavior. I test this implication of the model using a cross-country instrumental variable estimation to correct for the possible endogeneity of decentralization. Several variables for decentralization and corruption are used to maximize robustness. I find strong evidence in support of the theoretical model's prediction that decentralization is associated with lower levels of corruption.