Content area
Abstract
Investments in human capital accumulation, government consumption and total government expenditures present a striking negative correlation with capital shares. This correlation is robust to alternative specifications, lists of controls, and exclusion of outliers. Causality tests strongly support the hypothesis that the direction of causation runs from capital shares to the government spending variables. We present a political economy model of interest groups that can account for these correlations. In contrast, a median voter model predicts positive correlations between capital shares and the government spending variables. [PUBLICATION ABSTRACT]





