Working Paper: NBER ID: w9748
Authors: Brian Knight
Abstract: Local public goods financed from a national tax base provide concentrated benefits to receipient jurisdictions but disperse costs, creating incentives for legislators to increase own-district spending but to restrain aggregate spending due to the associated tax costs. While these common pool incentives underpin a variety of theoretical analyses, which tend to predict inefficiencies in the allocation of public goods, there is little direct evidence that individual legislators respond to such incentives. To test for reactions to such incentives, this paper analyzes 1998 Congressional votes over transportation project funding. The empirical results provide evidence that legislators respond to common pool incentives: the probability of supporting the projects is increasing in own-district spending and decreasing in the tax burden associated with aggregate spending. Having found that legislators do respond to such incentives, I use the parameter estimates to calculate the efficient level of public goods, which suggest over-spending in aggregate, especially in politically powerful districts, and large associated deadweight loss.
Keywords: No keywords provided
JEL Codes: H4; H7; D7
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| own-district spending (H72) | legislative support (K16) |
| tax burden (H22) | legislative support (K16) |
| own-district spending (H72) | efficient level of public goods (H49) |
| tax burden (H22) | efficient level of public goods (H49) |