The effects of government spending shocks: Evidence from U.S. states

Using panel structural VAR analyses and a recently released dataset on quarterly gross domestic product by state, this paper finds that an increase in state spending raises output, employment, and the real wage. Our baseline estimates imply that the multiplier effect of a spending increase on output is 1.3 contemporaneously and 1.2 over three years. The effects, however, vary considerably depending on the economic environment and institutional context. Specifically, we find that (i) the spending multiplier is larger during recessions than expansions (ii) the spending multiplier is relatively larger during periods of low debt than in episodes of high debt; (iii) spending shocks have relatively larger effects on states with “medium” balanced-budget requirements than those with either weak or strong balanced-budget requirements; and (iv) the spending multiplier is relatively smaller for states with either lax or strict debt restriction provisions compared to those with moderate provisions. These differences in the magnitude of the spending multiplier across states with different balanced budget rules and debt restriction provisions are rather small and not always statistically significant.