Select Page

According to the official Labor Department report on jobs and unemployment for May, the economy added 2.5 million workers in the month and the unemployment rate fell to 13.3% (from 14.7% in April.) Economists surveyed by Bloomberg had projected a 6 million drop in jobs and a rise in unemployment to 19.7%.

A huge surprise and stocks rallied on the numbers.

However, there’s good reason to doubt these numbers. The Bureau of Labor Statistics at the Labor Department flagged some of those problems even as it issued the figures. (The market, for today at least, was happy with the headlines.)

The jobs and unemployment figures are the result of two very different processes. And right now there are problems with both the jobs and unemployment data.

First, there’s a persistent problem in the way that jobs are being counted in the payroll data being reported by companies Some workers, nobody knows how many but its suspected that there total is significant, are being counted as on payroll–that is employed as far as the jobs numbers are concerned, when their hours or pay are actually minimal. Certainly less than back before the coronavirus recession. In other words, people are going back to work for job counting purposes but they’re not making much money at those jobs. That’s a problem since it’s wages that get spent to drive the economy not the number of jobs on paper. The last report from the Federal Reserve showed a huge drop in income from wages. That income hole was filled by government payments that included enhanced unemployment tbeenfits (which expire on July 31) and those $1200 checks (one-time only so far.) Today’s report showed that workers’ average hourly earnings fell by 1.0%, a rate well below the economist consensus.

Second, the Current Population Survey of 60,000 households that is used to calculate the unemployment rate has developed several response problems. First the rate of response to the survey has dropped by 15% in the most recent months. That makes the results less robust. Second, respondents to the survey have miss-classified workers. “As was the case in March and April, household survey interviewers were instructed to classify employed persons absent from work due to coronavirus-related business closures as unemployed on temporary layoff,” the Labor Department said in today’s report.“However, it is apparent that not all such workers were so classified. BLS and the Census Bureau are investigating why this misclassification error continues to occur and are taking additional steps to address the issue.”

Her’s the important take away: “If workers who were recorded as employed but absent from work due to ‘other reasons’ (over and above the number absent for other reasons in a typical May) had been classified as unemployed on temporary layoff, the overall unemployment rate would have been about 3 percentage points higher than reported (on a not seasonally adjusted basis).”

In other words the unemployment rate could/should have been 16.3% and not the reported 13.3%. That would still be much better than projected by economists but higher rather than lower than the unemployment rate of 14.7% in April.

I think this quote from Betsey Stevenson, an economist at the University of Michigan who was the chief economist at the Labor Department in the wake of the 200y7-2009 Great Recession gets at the important problem in bidding up stocks on current data.  “The problem is that Wall Street is used to predicting job loss due to a typical recession, not one in which people are temporarily sent home en masse,” she told Bloomberg.