| This dissertation comprises two essays that elaborate on different aspects of the relationship between government expenditure and long-term economic growth. The first essay explores how the size of government, as measured by the level of spending, affects growth. Theoretical models suggest a nonlinear relationship; however, testing this hypothesis empirically in cross-country studies is complicated by the endogeneity of government spending and the accurate identification of turning points. This paper examines the nonlinear hypothesis by incorporating threshold analysis in a cross-country growth regression. The methodology utilizes a sample-splitting framework and follows an objective strategy for identifying and testing changes in the slope. Using a broad panel of countries over the period 1971--2005, the results show evidence in favor of a nonlinear effect, but not of the form predicted by theory. When total government spending is low, there is no statistically significant effect on economic growth. However, after passing a certain threshold (26 percent for developed countries and 33 percent for developing countries) government spending exhibits a negative effect on growth. This pattern remains the same even when productive government spending is singled out.;The second essay develops a dynamic macroeconomic model to explore how variations in the composition and financing of government expenditures affect economic growth in the long-run. The model is used to analyze how public investment spending funded by taxes or borrowing affects long-term output growth. We also examine the effect of varying the composition of public expenditure, shifting between consumption and investment spending, or re-allocating between different types of public investment. In addition, we use alternative parameterizations of the model to explore how the effects on growth change under extreme initial fiscal conditions such as high average tax rates, debt ratios and public consumption spending. The model is calibrated to reflect economic conditions in the seven largest Latin American economies during the period 1990 to 2008. We find that, where tax rates are not already high, funding public investment by raising taxes may increase long-run growth. If existing tax rates are high, then public investment is only growth-enhancing if funded by restructuring the composition of public spending. Interestingly, using debt to finance new public investment compromises growth, regardless of the initial fiscal condition. |
