Meet the Author

Jayson Gerbec, Miriam Singer, and Adam Smith |

Research Interns

Jayson Gerbec, Miriam Singer, and Adam Smith are former research interns in the Research Department of the Federal Reserve Bank of Cleveland.

09.27.10

Economic Trends

Job Churning in Regional Labor Markets

Jayson Gerbec, Miriam Singer, and Adam Smith

A recently developed data source, the U.S. Census Bureau’s Longitudinal Employer-Household Dynamics (LEHD) database, is providing new information on the dynamics of U.S. labor markets. The LEHD provides quarterly data on worker and employment flows and allows for a rich description of both worker and job turnover at relatively fine levels of geographic detail. A key feature of data on employment flows is that they show the amount of job creation and job destruction that occurs greatly exceeds the net employment change. That is, there is considerable job and worker “churning” that is hidden by reports that focus on net employment change.

The measurement of worker reallocation is important to economists for a number of reasons. In particular, unemployment rates are related to worker flows, and one key driver of worker flows is the underlying flows in jobs across employers. As employers change size, they shed and add workers, resulting in hirings and separations. As we’ll show below, there has been some decline in excess reallocation in local labor markets, and this is consistent with the general trend in worker flows reported in research by the Cleveland Fed.

We measure the amount of job churning that has occurred in 190 of the largest metropolitan labor markets in the United States—the top quartile of U.S. labor markets with respect to size—since 2001. The measure we use is the excess job reallocation rate, which is calculated as job creation plus job destruction minus the absolute value of net employment change, all divided by the level of employment for the given time period. It reflects the amount of a metro area’s employment flows that exceeds its overall employment growth. Net employment growth is often less than a percentage point annually for a location, while excess job reallocation is roughly 10 times as large.

The figure below shows the distribution over time of excess employment flows across metropolitan areas. It shows the median, 25th and 75th percentiles, and the minimum and maximum values for the first quarter of each year from 2001 to 2009. Median excess job reallocation rates decline from 9.7 percent in 2001 to 8 percent in 2006, and then experience a slight uptick to 8.2 percent in 2009. The interquartile range is typically 2 to 2.5 percentage points with little discernible pattern either in the expansion or contraction of the range over time. However, there is considerable spread in excess reallocation rates across metropolitan areas, with some areas experiencing annual rates above 12 percent of employment and others below 6 percent.

There is also persistence in excess reallocation rates across metropolitan areas. That is, some metropolitan areas have a tendency to have higher reallocation rates while others have lower rates. The 8-year average excess job reallocation rates for the largest 190 MSAs go from a low of 5.7 percent to a high of 13.1 percent. The Fourth District’s four largest MSAs (Pittsburgh, Cleveland, Cincinnati, and Columbus) have similar levels of excess job churning—they all fall within one standard deviation of the mean excess reallocation rate. Cleveland has the lowest reallocation rate of the four cities (8.3 percent), while Pittsburgh has the highest (9.6 percent).

While the 190 metropolitan areas in our sample represent the largest labor markets in the nation, they differ quite a bit in size, ranging from 70,000 to 5.9 million workers. Excess reallocation is only moderately correlated with the size of the local labor market, with larger metropolitan areas having somewhat higher levels of reallocation.

What else could be behind these differences in metropolitan area reallocation rates? It is likely a combination of industry structure and firm heterogeneity. In particular, differences in both the size and age distributions of firms are likely sources of variation in the magnitude of employment flows across metropolitan areas.