Abstract
The authors estimate vector autoregressions for three categories of state government expenditure: education, highways, and welfare. Each regression contains state expenditure on that category, federal aid to the state in that category, total federal aid to the state, state general revenue, and personal income. The authors calculate impulse response functions for these vector autoregressions to analyze changes in aid and expenditure over time. The authors find that deviations in expenditure and category-specific aid have large and long-lasting effects on future expenditure. Category-specific aid has a much stronger effect than aid, revenue, or income, suggesting a flypaper effect in which aid is spent disproportionately in the category in which it is granted. This effect persists for up to 5 years, suggesting that there are structural dynamic links between category-specific aid and expenditure by states on those categories.
Get full access to this article
View all access options for this article.
