Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Using the Quarterly Workforce Indicators database, we document that a significant amount of the decline in labor market turnover during the last two decades is accounted for by the decline in employment spells that last just one or two quarters. This phenomenon is pervasive: short-term employment spells have declined across industries, firm size categories, demographic groups, and geographic regions. Using a search-and-matching model in the Diamond-Mortensen-Pissarides tradition that incorporates noisy signals about the quality of a worker-firm match, we argue that improved screening by workers and firms can account for much of the decline in short-lived employment spells.