Macroeconomic Model Spillovers and Their Discontents

The Great Recession underlined that policies in some countries can have profound spillovers elsewhere. Sadly, the solution of simulating large macroeconomic models to measure these spillovers has been found wanting. Typical models generate lower international correlations of output and financial asset prices than are seen in even pre-crisis data. Imposing higher financial market correlations creates more reasonable cross-country spillovers, and is likely to become the norm in policy modeling despite weak theoretical underpinnings, as is already true of sticky wages. We propose using event studies to calibrate market reactions to particular policy announcements, and report results for U.S. monetary and fiscal policy announcements in 2009 and 2010 that are plausible and event-specific.