Job Polarization and Jobless Recoveries

While the hollowing out of middle-skill positions has been ongoing for 30 years, it happens almost uniquely in recessions.

Two recent trends in the U.S. labor market -- job polarization and jobless recoveries -- are not only evident in the aftermath of the recent recession but also are linked, according to Nir Jaimovich and Henry Siu writing in The Trend is the Cycle: Job Polarization and Jobless Recoveries (NBER Working Paper No. 18334). Polarization -- the loss of middle-skill jobs that concentrate on routine tasks -- happens in spurts and essentially only during recessions. Because these jobs don't return once recessions are over, the rebound in the labor market is slow, and the result is a sluggish or "jobless" recovery.

"Jobless recoveries are observed only in these disappearing, middle-skill jobs," the authors write. "The high- and low-skill occupations to which employment is polarizing either do not experience contractions, or if they do, rebound soon after the turning point in aggregate output."

While the hollowing out of middle-skill positions has been ongoing for 30 years, it happens almost uniquely in recessions. Fully 92 percent of the loss of these jobs occurs within 12 months of NBER-dated recessions. Averaged over the last three decades, these jobs have accounted for more than half of all U.S. employment - so their disappearance represents a major drag on recoveries.

The authors compare three early recessions (ending in 1970, 1975, and 1982) with three later ones (ending in 1991, 2001, and 2009). On average, employment took ten months to recover in the early downturns versus 21 months in the later recessions. In the most recent (and sharpest) recession, employment still hasn't recovered to pre-downturn levels.

This difference can't be explained by differences in the severity of the declines. On average, output regained half of its pre-recession high in seven months during the early downturns; it took only slightly longer -- nine months -- in the later downturns. What has changed is the behavior of employment in routine occupations.

Routine occupations bear the brunt of the loss in all of the downturns. In the 1982 recession, these occupations experienced an employment loss that was greater than the total employment loss in the recession, because employment in the non-routine occupations was actually growing. In early recessions, however, these occupations recovered; in later recessions, they have not, not even in the long term. Thus, the authors conclude, "jobless recoveries" are observed only in these disappearing routine occupations and only since job polarization began.

This phenomenon is not accounted for simply by the cyclical behavior and secular decline of manufacturing in the United States. Similarly, it is not merely a result of the employment experience of workers with low educational attainment. The authors write that "a trend in routine-biased technological change can lead to job polarization that is concentrated in downturns, and recoveries from these recessions that are jobless."

--Laurent Belsie