The literature connecting growth and policies is much smaller. Glaeser et al. (1995) include a large vector of city-level political variables in our growth regressions. We find only two political variables that predict urban growth: the amount of debt per capita of the city in 1960 and spending on sanitation. We interpret the debt effect as the result of reverse causality. Cities expecting large growth borrow because they know that they will be able to repay that borrowing in the future. The sanitation effect is provocative, but hardly an overwhelming stylized fact. Rauch (1995) also shows that the municipal reforms in the early part of this century have little effect on growth.