Considering employment change since its trough across NECTAs, the variation in the contributions of the education and health services sector are stark, ranging from a low of -0.9 percentage point in Norwich to a high of 3.2 percentage points in Worcester. The manufacturing sector has actually fared worse since the trough in both Worcester and New Bedford than in Providence, yet Worcester and New Bedford have seen stronger recoveries than Providence in total employment.

These trends suggest that formerly manufacturing-intensive cities can recover to pre-recession employment levels more successfully by trying to capitalize on favorable national employment trends in nonmanufacturing sectors, most notably education and health services, than by anticipating a return of lost manufacturing jobs as the overall economy improves.

This paper seeks to discover why Rhode Island experienced a more severe recession than any other New England state and why the state continues to lag the region in some measures of labor market health. We find that two key factors can explain the state's last-place rank in the region for employment growth between 2008 and 2009, a time when the bulk of job losses occurred both nationally and regionally: (1) the industrial composition of the state's manufacturing sector prior to the recession, which left it vulnerable to particularly steep labor demand shocks and (2) the severity of the state's prior-year house price declines. We also find that Providence as a metropolitan area or NECTA had the most severe recession among a 10-NECTA comparison group, lending robustness to the state-level comparisons. Concerning the economic recovery, we observe that employment remains farther below its pre-recession peak in Rhode Island than in any other New England state because it fell farther during the recession, not because it grew more slowly than in the region's other states during the recovery.