Research Highlights Article
July 19, 2024
The distributional consequences of recessions
How do local labor markets recover after economic shocks?
Source: Mathias_Berlin
As long as economists have had the data, they’ve noticed that national economies regularly go through recessions, and that after each downturn, the pendulum swings back, suggesting little lasting impact. However, macroeconomic cycles and national statistics may be hiding substantial regional variation.
In a paper in the American Economic Journal: Applied Economics, authors Brad Hershbein and Bryan A. Stuart studied the relative recovery of local labor markets after recessions in the United States. They find that local labor markets hit the hardest persistently lag behind their more fortunate counterparts and experience long-term decreases in the employment–population ratio.
The authors’ findings are based on the analysis of local economic activity—such as employment, population, and earnings—between 1970 and 2019.
Several previous studies on local labor markets have found that employment rates completely recover from recessions. However, these approaches tend to rely on an estimation technique known as vector autoregression (VAR), which is prone to bias. The authors show that given the small samples of data available for studying recessions, VARs overestimate the amount of recovery in local economies.
“It turns out that there's bias, and that bias leads to the false conclusion that there's convergence,” Hershbein told the AEA in an interview. “Things don’t get back to where they were, and you have these gaps between places after they've been shocked.”
A map of severe recessions
The chart below shows the frequency of severe recessions, by metropolitan area, from 1973 to 2009. A severe recession is defined as experiencing a logarithmic employment change worse than the median area for a given recession.
Hershbein and Stuart used an event study regression that avoids this potential bias. Their approach relates the evolution of local economic activity to sudden employment changes caused by recessions. However, this method does not estimate the overall recovery of local markets. Rather, it shows how much areas with more or less exposure to the initial shock diverge after a recession.
Using this event study approach, the authors compared the relative employment, population, and earnings changes in labor markets across US metropolitan areas.
They found that declines in employment due to recessions are extremely persistent. Across the five recessions in their sample, a 10 percent decrease in employment during a recession—which is roughly the difference between areas in the 90th percentile and 10th percentile of employment loss during the Great Recession—led to an 11 percent decrease in employment nearly a decade after the recession trough.
These persistent gaps were not concentrated in specific jobs, but spread across industries. However, the relative employment and earnings losses were concentrated among the lowest wage earners. For workers in the bottom of the income distribution, it appeared that recessions were followed by a greater loss of earnings in locales harder hit by the recession.
The evidence also suggests that local markets were not adjusting the way that many economic models would predict. In theory, people will move away when local economic conditions deteriorate, reallocating their labor in more efficient ways and allowing the employment rate to fully recover.
But unlike previous studies, the authors saw limited population response in the data, which led to persistently lower employment–population ratios and changes to the composition of the local workforce. In particular, the population loss was not driven because people left. It was primarily driven by fewer people moving in, especially young people.
If places get hit by a recession, not only do they lose jobs, not only are they more reliant on social welfare transfers, but they have fewer young people moving in. That has big implications for economic dynamism.
Brad Hershbein
“If places get hit by a recession, not only do they lose jobs, not only are they more reliant on social welfare transfers, but they have fewer young people moving in. So the places get older, and they tend to create fewer entrepreneurship opportunities,” Hershbein said. “That has big implications for economic dynamism.”
Contrary to some previous work, the findings also indicate that this persistence is not a new phenomenon. Recessions since the early 1970s all led to persistent gaps between areas that experienced more or less severe employment losses.
Overall, the authors’ work indicates that economic shocks to local labor markets can cause persistent regional gaps in employment rates and exacerbate inequality more than previous economic research would suggest.
♦
“The Evolution of Local Labor Markets after Recessions” appears in the July 2024 issue of the American Economic Journal: Applied Economics.