Aggregate Effects and Measuring Regional Dynamics

Empirical models of regional adjustment often control for aggregate effects when estimating the impact of regional shocks. It is, however, difficult to filter out the effects of aggregate shocks— such as national recessions— because the incidence of these shocks varies across space and time. We argue that an approximate factor structure is better suited to controlling for this form of spatiotemporal heterogeneity than existing methods. Applying the method to US states, we find that region-specific labor demand shocks are primarily absorbed through changes in participation; that the migration response to these shocks is limited; and that recoveries are highly protracted.