Deeply Conflicted Verdicts on Job Market Progress

The two BLS employment surveys gave us two starkly contrasting messages about job market trends in October. The survey of employers delivered the reassuring news that job gains remain strongly positive and on an even keel. Private firms added 212,000 workers to their payrolls in October, somewhat faster than the average pace of private job gains over the previous 6 and 12 months. At the same time, new estimates of job gains in August and September lifted estimated employment growth in those two months by 70,000. Government payrolls dipped slightly in September, and indeed have shrunk modestly over the past year. Nonetheless, the latest report on payrolls conveys an unmistakable impression of continued strong demand for workers in the private sector. Given the current age structure of the population, the nation needs about 80,000 new jobs every month to keep up with the growth of the working-age population. The employer survey shows job gains that comfortably exceed that threshold.

In contrast to the good news in the employer survey, the household survey served up a giant helping of bad news. According to household respondents, employment fell 735,000 in October. Since June employment in the household survey has shrunk about 122,000 a month. In the past 12 months total employment has increased only 240,000. The main reason the unemployment rate has dipped in the past year is that the number of labor force participants has fallen by 737,000.  The number of adults in the work force fell 720,000 in October and has shrunk by an average of 250,000 a month since June. The labor force participation rate dropped 0.4 percentage points in October, reaching a 35-year low. The falling participation rate meant that October’s unemployment rate increased “only” 0.1%.

In view of the shifting age composition of the U.S. population, we should expect some drop in participation even if the nation enjoyed full employment. The huge baby boom generation is reaching retirement age, and its withdrawal from the labor force would be expected even if wages were climbing and jobs plentiful. Since the onset of the Great Recession in early 2008, however, the labor force participation rate has fallen considerably faster than can be explained by an older population. Roughly half the drop in the participation rate is due to aging, but a bit more than half is due to the discouraging job prospects faced by laid-off workers and young adults who want to begin a career. If we focus solely on adults who are between 25 and 54, when participation rates are highest, the latest BLS statistics show that participation has fallen to 80.5% of the population, a drop of 0.4% since September and the lowest level since March 1984. The employment-population rate of 25-54 year-olds also reached a 29-year low in October. Compared with other wealthy countries, the U.S. now has below-average employment and participation rates among its prime-age population.

The October government shutdown affected estimated employment gains in both the employer and household surveys. Over 400,000 federal government employees were directly affected by the shutdown, and it is likely that a sizable number of private employees were indirectly affected. However, the laid-off federal employees were treated as paid employees in the employer survey, whereas many of them were classified as on temporary layoff (and consequently unemployed) in the household survey. It seems plausible to think private payroll gains in October would have been even larger were it not for the shutdown. This implies that the employer survey is providing evidence of considerable strength in private-sector demand for workers. In the November jobs report the federal workers temporarily laid off in October will again be reported as employees in the household survey, erasing some of the employment losses reported in the October household survey.

The shutdown cannot explain the very different patterns of employment growth shown in the employer and household surveys over recent months. The employer survey shows payrolls increased more than 1.7 million since January. The household survey shows an increase in employed adults of less than 0.25 million. Other indicators of the strength of the economy suggest the payroll survey is giving us the more accurate indicator of job market gains. Nonetheless, the gap between the signals provided by the two surveys is deeply puzzling.