Job Market Paper: Regional Government Consumption and Investment Multipliers

This paper is the first to use the regional fiscal multipliers framework to separately identify government consumption and investment multipliers. I run panel IV regressions on a U.S. state-level dataset, constructed from transaction-level data on U.S. federal military procurement contracts, and find that regional government consumption multipliers are larger than regional government investment multipliers. Also, spillovers are more negative from government consumption. I interpret these results with a New Keynesian DSGE model of a two-sector, two-region monetary union, with two key mechanisms: (1) investment is more intertemporally substitutable, and (2) consumption exhibits greater home bias. Finally, I estimate a quantitative model of the United States as a 51-region monetary union, to measure the effects of the 2021 Infrastructure Investment and Jobs Act (IIJA). While panel estimates of regional multipliers are larger than time-series estimates of aggregate multipliers, the quantitative model yields larger aggregate multipliers than a model that abstracts from sectoral or regional differences.

JEL Codes: E32, E62, H30, H57.

Media Mentions: Chosun Biz (Korean, English), Aju News (Korean)