Revealing the True Details Behind the U.S. Employment Report (Released in Early June 2020)

First, there was no manipulation of the data. However, during the month of May 2020, BLS employees compiling the data for the Household Survey (used to compute the national unemployment rate) counted an excess number of workers that did not show up for work as being employed in May 2020.

Of course, in a typical month many workers may be legitimately employed even if they are not on the premises. This occurs, because many of these workers may be on vacation, on maternity leave, on military duty, absent from work due to bad weather or absent from work due to a labor dispute. Still the results for May 2020, raised some eyebrows because over the past 4 years, (during the month of May) if we average the number of workers out of work but still counted as being employed (due to these factors), the number has averaged 4.1 million (on a non-seasonally adjusted basis, compared to the figure of 8.4 million in May 2020.

After adjusting for these non-seasonally adjusted results, the Bureau of Labor Statistics (BLS) concluded that the actual number of unemployed individuals should have been closer to 25.8 million in May 2020 instead of the reported figure of 21 million. This extra number of 4.8 million unemployed workers implies that the true unemployment rate for May 2020 should have been equal to 16.3% instead of 13.3%.

What Else Did We Learn from the May Employment Report?

Out of the 21 million number of people that were counted as being unemployed in May 2020, (using the actual reported number and not the corrected figure), 11.9 million people that were originally unemployed in April 2020, remained unemployed in May 2020. This figure is up from a reading of 3.4 million reported during the prior month.

On the good news front, we observed less individuals that were employed in April move to the unemployed category, i.e., just 4.9 million in May 2020 versus a figure of 17.5 million in April 2020. Stated another way, in April 2020, we observed an outsized 17.5 million people suddenly moving from being employed during the prior month to now suddenly being placed in the unemployed category.

Some additional good news comes from those individuals that were not in the labor force. When an individual is defined as “not in the labor force,” they are not counted as being employed (because they don’t have a job) and are not counted as being unemployed (because they are not actively looking for a job). The number of individuals that moved from this "not in the labor force," category to suddenly being counted as being unemployed rose to 4.2 million in May versus a comparable reading of 2.2 million in April 2020. When individuals re-enter the labor force and start looking for a job, it implies that they have increased confidence of finding employment, which encourages them to look for work. Sometimes, they find a job the same month, and therefore move straight into the employed category, but sometimes, they temporarily move into the unemployed category. In either case, seeing this number rise in a given month, does suggest an improvement in labor market conditions.

What about the Establishment Survey?

Unlike the previous survey, that provides information on about 60,000 Households per month, this survey simply counts the number of Non-Farm Payroll jobs in any given month by mailing out questionnaires to employers.

Although some may think that the reported error that was discovered by the BLS within the Household Survey may mean that the +2.5 million non-farm payrolls jobs gained in May 2020 was incorrect --- this is clearly not the case. Those positive results stand as no irregularities were found in that separate survey. More importantly, over long periods of time, the BLS has informed us that the Establishment Survey tracks movements in real GDP growth more closely than the Household Survey.

These overall results do suggest that U.S. labor market conditions are slowly improving. But with 25.8 million workers still pounding the streets and looking for work or waiting to return to their previous jobs, it may be a bit premature to bring out the brass trumpets to celebrate full victory, just yet. One month of small improvement does not define a trend.

Stay tuned for next month’s report for more details on how U.S. labor market conditions are evolving as the economy continues to gradually re-open!