Measuring the Output Responses to Fiscal Policy

This 2012 American Economic Journal: Economic Policy paper by Alan Auerbach and Yuriy Gorodnichenko estimates how government spending multipliers vary with the state of the business cycle, using a smooth-transition structural VAR (STVAR) fit to postwar quarterly U.S. data. The authors find that government spending is considerably more effective at raising output in recessions than in expansions: in their baseline estimates the cumulative multiplier runs roughly 0 to 0.5 in expansions but 1 to 1.5 in recessions, consistent with both older Keynesian intuition and newer zero-lower-bound DSGE models. Disaggregating spending, they find multipliers differ sharply by component, with military/defense spending showing the largest and most cyclically sensitive response (peaking near 4 in recession). Purging fiscal shocks of the portion that professional and government forecasters could already predict – using real-time forecasts from the Michigan RSQE model, the Survey of Professional Forecasters, and Federal Reserve Board staff projections – further raises the estimated recession multiplier, since much of what a standard VAR treats as a “shock” is in fact anticipated. The results are specific to the postwar U.S. sample and to the historical persistence of fiscal shocks within it, and the authors caution that lumping together historically distinct types of recessions (financial-crisis recessions vs. monetary-contraction recessions) may obscure further heterogeneity in the multiplier.

American Economic Journal: Economic Policy | Macro Paper Warehouse

Measuring the Output Responses to Fiscal Policy