Cyclical Government Spending: Theory and Empirics[Mafia and public spending: evidence on the fiscal multiplier from a quasi-experiment]

This paper shows that part of what is usually labelled discretionary government spending actually varies systematically over the cycle. I exploit the pervasive gap between ordinary least squares and two-stage least squares local government spending multipliers to estimate how cyclical the systematic part of government spending is. Estimating a structural open-economy New Keynesian model on United States state-level data, I find that when employment decreases by , the systematic component of government spending decreases by . I also find that the empirical specification in Nakamura and Steinsson, ‘Fiscal Stimulus in a monetary union’ (American Economic Review, 2014) does a good job in recovering the true impact multiplier effect, but that it overestimates the long-run cumulative effect.