Three Essays on Corruption and Government Finance

There are three chapters in my dissertation. In chapter one I study the impact of government corruption on tax avoidance by corporations with tax liability in China. I begin by developing a model of a firm\u27s choice of tax avoidance based on the level of corruption in the firm\u27s province. I show that a higher level of government corruption and a higher tax rate are predicted to increase a firm\u27s tax avoidance. My empirical estimates show that a one-standard-deviation increase in government corruption corresponds to a 6% increase in tax avoidance by firms, based on data from nearly 600,000 firms from 1998 to 2007. When I separate the sample by type of ownership of firms, I find that domestic private firms tend to avoid a larger portion of their tax liabilities than foreign or state-owned firms. I also find that tax avoidance increases when the effective tax rate increases. The size of the firm, age of the firm, and whether the firm exports from China also influence tax avoidance. In chapter two, a joint work with Sergey Mityakov, we study the impact of competition on tax avoidance by corporations in Mexico. Using more than 2,000 firm level data, we show that firms tend to avoid more taxes in less competitive markets, that is more concentrated market. A one standard deviation increase in competition leads to a 7.7% increase in tax avoidance. By splitting the sample based on ownership structure, we show that domestic firms avoid more taxes. When we separate the firms according to the plant size, we find that larger firms tend to avoid more taxes. We find robust and consistent results by using different competition measures and other robustness tests. In chapter three, I theoretically demonstrate how fiscal decentralization affects corruption. The theory predicts that fiscal decentralization reduces corruption. This result is then tested using a panel data set of 31 provincial level government from 1998 to 2007 in China. My estimates suggest that fiscal decentralization in government expenditures and government revenue is negatively correlated with corruption. Using leader and location dummy variables, I find that central government leaders can influence the corruption level in China. I also find that more developed regions in China tend to be less corrupt