We expect an increase of 3.1 million in total employment, led by workers who have been called back to their jobs, and a decrease to 12.2% in the unemployment rate when the jobs data for June is released on Thursday.
But those figures should be looked at with a skeptical eye. Pandemic economics is now organized around near real-time economic data that tends to de-emphasize lagging economic reports that often appear out of date by the time they are published.
By the time the June jobs report comes out, we will have already started estimating the damage to the labor market wrought by pullbacks and shutdowns of business because of an intensification of the pandemic that will appear in the July estimate.
What we will be watching?
- Data collection: In the May survey, the collection rate for the establishment survey was 69%, lower than the normal rate, while the collection rate for the household survey was 67%, about 15% lower than the pre-pandemic collection rate. This means that we can expect significant revisions in either direction across the entire data sets that underscore the establishment and household surveys. If one looks at the variance in the consensus forecast, one will note a high of 9 million in the change in total employment and low of 500,000 — both in contrast with the Bloomberg consensus of 3 million. The standard deviation of the consensus forecast is 1,577%, which should tell one all one needs to know about the general level of uncertainty about the labor market five months into the pandemic.
- Misclassified workers: The period of March through April was characterized by difficulties at the Bureau of Labor Statistics classifying employed people absent from work because of pandemic-related business closures as unemployed or as a temporary layoff. One suspects that this is a sophisticated coding error that will take some time work through. For this reason, one may want to take the unemployment rate estimate with a grain of salt. The standard deviation inside the Bloomberg consensus forecast is .88%, which is very large compared to past forecasts, with the high estimate of 15.5%, the low forecast 10.1% and the median estimate at 12.4%.
- Alternative data: Many of us are now looking at private data sources such as Homebase and LinkUp. If one uses that data in addition to other alternative data combined with traditional economic or financial data to create quantitative forecasts, one might forecast a slightly more optimistic outcome for the June employment report.
- Traditional data: That being said, we will continue to focus on traditional data in addition to the novel data outlined above. This includes the U-6 underemployment estimate, aggregate hours worked, employment to population ratio and the labor force participation rate. Finally, given the large shock to the domestic labor market, we will look at the median duration of unemployment and average weeks of unemployment.
For more information on how the coronavirus is affecting midsize businesses, please visit the RSM Coronavirus Resource Center.