The unemployment rate exceeded 8 percent for more than three years, leading commentators and policy makers to speculate that there has been a fundamental change in the labor market. The view is that today’s economy cannot support unemployment rates below 5 percent—like the levels that prevailed before the recession and in the late 1990s. Those in government may take some comfort in this view; it lowers expectations and provides a rationale for the dismal labor market.
Excuses aside, this issue is also important for central banks. The Federal Reserve and other central banks have policy choices to make if the high rates of unemployment reflect cyclic phenomena. But if the problem is structural—perhaps reflecting a mismatch between skills needed by business and skills possessed by the unemployed—there is little the Fed can do.
Research I’ve done with James Spletzer of the U.S. Census Bureau shows that the problems in the labor market are not structural. They reflect slow economic growth, and the cure is a decent recovery.
In 2007, the unemployment rate was 4.4 percent. Two years later, it reached 10 percent. The structure of a modern economy does not change that quickly. The demographic composition of the labor force, its educational breakdown, and even the industrial mix did not differ much between 2007 and 2009.
More specifically, from 2007 to 2009 unemployment grew dramatically in a few industries, and these changes contributed to the rise in overall unemployment. But the changes were similar to those experienced in prior recessions. As unemployment rates declined somewhat after 2009, the pattern played out in reverse. Industries that saw the largest increases in unemployment were the ones with the largest decreases as overall unemployment fell.
Between November 2007 and October 2009, the national unemployment rate rose 4.9 percentage points. Of that increase, 19 percent was accounted for by increases in unemployment in construction, 19 percent by increases in unemployment in manufacturing, and 13 percent by increases in unemployment in retailing. Although every industry experienced rising unemployment, half the increase occurred in those three.
Those same industries experienced the largest reductions in unemployment as the overall rate declined. Between October 2009 and March 2012, the overall rate fell by 1.9 percentage points. Of that fall, 22 percent was a result of declines in construction unemployment, 31 percent in manufacturing unemployment, and about 8 percent in retail. Once again, those three industries accounted for more than 50 percent of the reductions in the rates of unemployment.
“Mismatch” is another measure of structural maladies in the labor market. Mismatch can take a number of forms, but the most important is industrial or occupational mismatch, in which industries that have many job openings have few unemployed workers with the requisite skills, and industries with many unemployed workers do not have job openings.
For example, suppose demand in health care is growing, providing openings for workers with the needed skills. At the same time, manufacturing is declining, but workers who are well suited to manufacturing may not be able to move easily into health care.
Spletzer and I find that mismatch increased dramatically from 2007 to 2009. But it decreased just as rapidly from 2009 to 2012. Like unemployment itself, industrial mismatch rises during recessions and falls as the economy recovers. The measure of mismatch that we use, which is an index of how far out of balance are supply and demand, is already back to 2005 levels.
Whatever mismatch exists today was also present when the labor market was booming. Turning construction workers into nurses might help a little, because some of the shortages in health and other industries are a long-run problem. But high unemployment today is not a result of the job openings being where the appropriately skilled workers are unavailable.
Even within industries, there may be some chronic mismatch between vacancies and skills available. Our results on occupational mismatch suggest there is a shortage of skilled managers and professionals in most industries.
That is not to imply that we are back to where we should be. The unemployment rate is still well above the 5 percent level that we can aspire to maintain.
The reason for the high level of unemployment is the obvious one: overall economic growth has been very slow. Since the recession formally ended in June 2009, the economy has grown at 2.2 percent per year, or 6.6 percent in total. An empirical rule of thumb is that each percentage point of growth contributes about half a percentage point to employment.
The economy has regained about four million jobs since bottoming out in early 2010, which is right around 3 percent of employment—just the gain that would be predicted from past experience. Things aren’t great, but the failure is a result of weak economic growth, not of a labor market that is out of sync with the rest of the economy.
The evidence suggests that to reduce unemployment, all we need to do is grow the economy. Unfortunately, current policies aren’t doing that. The problems in the economy are not structural and this is not a jobless recovery. A more accurate view is that it is not a recovery at all.