What are the long term employment effects of the recession?
Danny Yagan│ April 15, 2019
Policy brief written with the assistance of Emma Fernandez
Summary of: Employment Hysteresis from the Great Recession. Yagan, Danny. (2018). Journal of Political Economy.
Employment rates have not recovered from the Great Recession
The U.S. unemployment rate rose from five to ten percent during the Great Recession, and by the end of 2015 returned to its pre-recession level of below five percent. Focusing on this standard measure would seem to indicate that the U.S. labor market has recovered from the recession. However, while the unemployment rate has bounced back, the U.S. labor force participation rate (the share of adults who are employed or seeking employment) and therefore the U.S. employment rate (the share of adults who are employed) did not. It declined more than three percentage points between 2007 and 2015 and did not recover, as millions of adults exited the labor force. Only half or less of this drop in the labor force participation rate can be explained by demographic change. I explore whether and how the Great Recession caused this decline in America’s employment rate.
Spending effects are greater than the sum of their parts
Long-term labor market effects can occur because of factors unrelated to the Great Recession and the underlying shocks that gave rise to it. For example, technical and trade changes can influence hiring patterns. I use regional variation in recession severity as a natural experiment to isolate the effects of the Great Recession. The results suggest that the recession caused most of the employment rate decline.
Not all parts of the U.S. experienced the Great Recession equally, as seen in Figure 1 below. Phoenix, for example, experienced a large unemployment shock, while San Antonio was hit relatively mildly by the recession. To compare regional severity, I categorize all U.S. local areas based on the magnitude of the local unemployment shock from 2007 to 2009.
Using anonymous individual data from over a million federal income tax records, I find that workers who were living in 2007 in areas that suffered severe unemployment shocks (like Phoenix) were less likely to be employed in 2015. A one percentage point higher local recession shock caused the average American to be 0.39 percentage points less likely to be working in 2015. These effects were largest for workers with lower initial earnings. Extrapolating from local areas to the aggregate United States, the results suggest that the Great Recession caused 76% of the decline in the overall U.S. employment rate.
Recession Impacts by Geography
In contrast to the conventional view that recession impacts are short-lived, my research shows that the Great Recession’s employment impacts were persistent. Thus despite normal signals that the recession is over, it is possible that the U.S. is still not at a full employment level consistent with stable inflation. This paper opens opportunities for further research to model labor force exit during and after economic fluctuations, test for long-term employment impacts of previous recessions, and examine any continued and future impacts of the Great Recession.
Persistent employment impacts of recessions amplify the potential bang-for-the-buck of significant stimulus early in a recession to the extent that stimulus avoids long-lasting labor market damage. Continued stimulus policies could even be warranted to spur reentry of those who are still non-employed years after the recession’s end. Alternatively, job training or skill development policies may be able to combat persistent non-employment—helping to replenish skills lost during joblessness and to prepare workers for jobs in the new economy.