Brock Mendel May Be Writing the Ultimate Local Multipliers Paper...
Brock Mendel http://people.fas.harvard.edu/~jbmendel/LocalMultiplier030612.pdf:
After considering the theory of the local multiplier, I proceed to estimate it using local data. I map firm-level revenue surprises into the counties where those firms are headquartered and regress employment growth in those counties on these revenue surprises. Under the restrictions proven in the theoretical section, this regression consistently identifies the impact of surprise government spending on employment growth. I include the effects of surprises to revenues for firms located in a neighboring county. I find that one year of one job is created for every $48,000 in spending. The validity of this procedure implicitly assumes that firms conduct all of their operations in the county containing their headquarters, which is obviously untrue. I therefore conclude that my estimate is an upper bound and not a point estimate…
[…]
The literature has identified multipliers for at least six different types of expenditures; to argue over which of these is “the” correct multiplier is to miss the point. There is a literature identifying the effects of federal government spending on the military, another using a narrative approach to identify the effects of exogenous changes to the tax code, a third using a VAR approach to identify innovations in total government spending, a fourth looking at the stimulative impact of tax rebates delivered in different formats, and fifth examining the effects of a liquidity shock to a state governments, and a sixth considering exogenous changes in federal grants to states. I carefully review all of these below; the important point here is that there is no ex-ante reason to expect all of these to identify the same multiplier. In fact, it is more surprising that so much of the local multiplier literature does manage to find consensus…
[…]
A very recent literature considers the impact of government spending on local economic outcomes. While my review of the global multiplier literature was necessarily a partial one, the following review covers every paper I have found on local multipliers as of December 21, 2011…
[…]
The data include 27,542,324 contracts totaling 4.27 trillion dollars over ten years…
