Measuring Employer-to-Employer Reallocation

We revisit measurement of Employer-to-Employer (EE) transitions, the main engine labor market competition and employment reallocation, in monthly Current Population Survey (CPS). follow Fallick Fleischman (2004) exploit a key survey question introduced with 1994 CPS redesign. detect sudden sharp increase incidence missing answers to this starting 2007, when U.S. Census Bureau change methodology, Respondent Identification Policy (RIP). show evidence selection into answering EE by both observable unobservable worker characteristics that correlate mobility. propose model procedure impute question, thus after introduction RIP. Our imputed aggregate series restores close congruence business cycle, especially onset Great Recession, exhibits much less dramatic drop 2008-2009 full recovery 2016, eliminates spurious appearance declining dynamism US 2000. also offer first (large negative) impact COVID-19 crisis on reallocation. Institutional subscribers NBER working paper series, residents developing countries may download without additional charge at www.nber.org.