The Bureau of Labor Statistics (BLS) provides labor turnover and vacancy data for four broad census regions (the Midwest, Northeast, South, and West) and the entire nation as part of its Job Openings and Labor Turnover Survey (JOLTS). According to these data, which begin in December 2000, there is significant variation in the way labor turnover behaves in the different regions. In the Northeast, for instance, employers hired an average of 745,000 workers a month, while in the South it was more than 1.7 million. Similarly, the South has accounted for most of the job openings and separations over the past eight years, whereas the Northeast lagged behind other regions.
These figures should not be surprising given the South's greater employment numbers: Since 2001, average total employment in the South has accounted for about 35.5 percent of U.S. nonfarm payroll employment, compared to 23.6 percent for the Midwest, 21.9 percent for the West, and 19 percent for the Northeast. But the curious behavior emerges when we look at each region's share of various categories of labor turnover. The contribution of the South is greater than its employment share in virtually every category. Its share of hires, separations, and job openings are all about 38 percent of the U.S. total. The West shows a similar pattern, although much less pronounced. The Northeast, on the other hand, seems to have contributed less than its employment share in all measures of labor turnover and job openings. The Midwest presents a balanced picture in terms of hires and separations. Its share of job openings, however, has been lagging significantly behind its employment share. As a result, the South's employment share has grown over this period (from 35 percent to 36 percent), more rapidly than the West (from 21.5 percent to 22.2 percent), whereas that of Northeast and Midwest have been shrinking somewhat (from 19.2 percent to 18.5 percent and from 24 percent to 23 percent, respectively).
Time series data on turnover and job openings reveal some interesting similarities and differences across regions as well. During the last recession, for example, all four regions experienced sharp declines in job openings. Even though each region started to recover later, only in the South and West had job openings reached their pre-recession levels by 2007. Interestingly, in the Midwest, job openings have been virtually flat for the past three years.
Hiring activity also declined in all four regions during the last recession. Once again, although each region had begun to recover by the end of 2003, hiring remained below pre-recession levels in the Midwest and Northeast. This picture, in conjunction with the separations data, implies that job reallocation has declined in the Northeast and Midwest over time. While high levels of separations and hiring could be inefficient if resources are being spent unnecessarily to reallocate workers across different firms, regions, and states, they could also indicate an active search by both workers and firms to find their best matches in the labor market.
Each region's relative shares could give further interesting details about regional labor markets. To this end, we construct relative shares of labor turnover and job openings for all four regions, adjusting each for the region's employment share. This is basically a ratio of two shares: the share of a region's job openings, hires, or separations over the share of a region's employment. So, for example, a ratio for hires greater than 1 for a particular region means it is hiring a higher number of workers than the U.S. average. Several features of the data stand out when interpreted in this way: First, the South leads U. …