School's Out for Summer, and the Timing Messed With May's Jobs Data

Statistical quirks explain some of the big swings in U.S. employment numbers.

Prising: Workforce Participation Rate at Historical Low

Remember how unexpectedly bad the May U.S. employment report was? With Friday's release of June employment data, it turns out the weakness was a fluke after all—and it looks like a big part of it was due to seasonal quirks related to the timing of the end of the school year.

Every month, the Bureau of Labor Statistics compiles and publishes data tracking how many people are moving between employment and unemployment and, more broadly, how many people are entering and leaving the labor force. The agency also adjusts for seasonal factors to smooth the data—school-year timing being one of them.

In May, we saw a decline in the unemployment rate and a drop in labor-force participation. The change suggested that unemployment was falling for "bad reasons," as discouraged workers gave up looking for work. Reinforcing that notion, the share of unemployed workers leaving the labor force spiked, on a seasonally adjusted basis.

The unadjusted data painted a different picture, as John Robertson, an economist at the Federal Reserve Bank of Atlanta, pointed out in a blog post last month.

When schools finish up in May, more students start looking for work, which adds people to the labor force. So normally, in data that aren't adjusted for seasonal fluctuations, fewer people drop out of the workforce from jobless rolls during the month.

That big slowdown in dropouts we usually see in May didn't happen this year. So when the normal seasonal adjustment was applied, it magnified the flows out of the labor force. That, in turn, helped push the unemployment rate down 0.3 percentage point, to 4.7 percent, as the participation rate dropped 0.2 percentage point.

0708 new chart

In June, however, the slowdown in labor-force dropouts played out more fully, as the chart shows. The unemployment rate jumped back 0.2 percentage point, while the participation rate rose 0.1 percentage point.

Taken with the strong rebound in payroll gains, to 287,000 in June from a revised 11,000 the prior month, the data suggest weakness in the U.S. job market was exaggerated due to statistical quirks in May. Policymakers can take some comfort from the fact that the employment situation in the second quarter of this year appears to have been much better than initially reported.

Given the timing of college graduations, "it's always tough this time of year" to sort through the data, says Sarah House, an economist at Wells Fargo Securities in Charlotte, N.C.

    Before it's here, it's on the Bloomberg Terminal.