The U.S. Labor Market: Status Quo or a New Normal?

Featured in print Digest

[There is no] compelling evidence that there have been changes in the structure of the labor market that are capable of explaining the [recent] pattern of persistently high unemployment rates.

The recession of 2007 to 2009 caused such high and persistent unemployment that it led many to conclude that the labor market had undergone structural changes, making it difficult or impossible to return to pre-recession employment levels. But in The United States Labor Market: Status Quo or A New Normal? (NBER Working Paper No. 18386), Edward Lazear and James Spletzer suggest that cyclical, not structural forces, are behind the surge in unemployment from 4.4 percent in the spring of 2007 to 10 percent in the fall of 2009, and the slow decline since then.

"[T]he current recession does not appear fundamentally different from prior ones, except that it is worse," they conclude. They fail to find "any compelling evidence that there have been changes in the structure of the labor market that are capable of explaining the pattern of persistently high unemployment rates." Instead, they note that "the evidence points to primarily cyclic factors."

The authors note that there are a number of ongoing, long-term industrial and demographic shifts in the labor market, but that none of these factors can explain the recent rise in unemployment. For example, the relative decline in U.S. manufacturing jobs has been under way for a half century, and the rise in female employment dates back to the second half of the twentieth century. The U.S. labor force is also aging, but again this is a long-term trend.

The authors' evidence suggests that long-term trends played a limited role in the recent recession and other past recessions. For example, in each of the business cycles between December 1979 and March 2012, the rise or fall in unemployment can be explained by changes in gender-specific unemployment rates, not by shifts in the gender composition of the workforce. Similarly, the aging of the workforce does not correlate very strongly with shifts in the unemployment rate during business cycle sub-periods. Since November 1982 the changing age composition of the workforce has lowered the unemployment rate by eight-tenths of a percentage point. This trend is reinforced by changes such as the rising education of the workforce and the shift toward service jobs, which have worked to lower unemployment over the last four decades.

The authors suggest that the rapid rise in unemployment during the 2007-9 recession can be explained almost entirely by the rise in unemployment within industries. Some industries such as construction, manufacturing, and retail trade saw unemployment soar. But these were the same industries that saw large decreases in unemployment during the recovery. The construction sector, for example, accounted for 19.4 percent of the increase in the national unemployment rate during the recession; this sector also accounted for 21.5 percent of the decline in the unemployment rate during the recovery.

The same phenomenon has occurred with mismatch -- the difference between vacancies and the number of unemployed in an industry, occupation, or location. Industrial mismatch rose during the 2007-9 recession, and then declined just as quickly. Occupational mismatch -- always higher than industrial mismatch and less sensitive to the business cycle -- rose during the recent recession but has since fallen below its pre-recession level. "There is no evidence that the recession resulted in a long-lasting skills gap that would require retraining experienced workers to work in different industries," the authors conclude. "Turning unemployed manufacturing and construction workers into nurses and teachers would not provide those workers with immediate jobs; there is already a surplus of unemployed even in the low unemployment industries."

There are at least two areas where this recession appears different than previous ones. First, the long-term unemployed make up a larger share of total unemployment than in past downturns, even those with comparably high unemployment rates. Second, there are more vacancies per unemployed person than even a couple of years ago. This shift of the Beveridge curve -- which measures the relationship between job openings and the unemployment rate -- may suggest that some permanent structural change is under way and is keeping the unemployed from filling the jobs that are available. The authors conclude that the reason for such a shift, if it has indeed occurred, won't be known until unemployment returns to normal levels.

--Laurent Belsie