March Jobs: An Alternative Measure?

Consumers displayed caution in their spending and saving habits even before the COVID-19 crisis took hold. The March payroll survey was taken during the week of March 12, before many shelter-in-place and stay-at-home warnings were issued. Anyone who worked one hour that week will be included in the payroll survey and not yet have hit the ranks of the unemployed. That said, we know that something major happened during the first few weeks of the month when initial unemployment claims surged first to 282,000 and then to 3.3 million. Could the number crunchers at the Bureau of Labor Statistics (BLS) decide to update their estimates on the birth and death rates of firms to more accurately capture the March turning point? We asked BLS directly and this is the answer we received:

“BLS will be reviewing our birth-death model and all other aspects of our methodology very carefully. Should any changes to our methodology be necessary they will be made clear to the public and available upon release of the data. These changes are typically spelled out in a box note within the Employment Situation news release.”

Hmm? That sounds like BLS statisticians may be weighing a change to how they announce the initial payroll data. I called former BLS Commissioner Erica Groshen to get some perspective. She underscored that she has no knowledge of the current decision making but that BLS has done a lot of work on the problem with detecting turning points in the wake of the Great Recession. The analysis at the time suggested that BLS was better off sticking to a methodology that captured firm birth and death estimates during normal times, or an expansion, than during a turning point. But, and these next two points are important, BLS has a lot more computing power than it did when she was there and may be better able to capture the magnitude of the March inflection. BLS was likely still weighing that decision on Monday. Groshen noted that the births of firms, which drop precipitously in bad times, are more important than the payroll calculation at inflection points than firm deaths, but even that could change.

Groshen also said that statisticians may want to wait until next month to release an update on births and deaths of firms, when they will have a better baseline for the data. The other option could be to introduce a shadow series that is more volatile but better captures the turning point and what that will ultimately mean for policy makers trying to understand the speed and breadth of this recession.

The former commissioner worried that the response rate for the establishment survey, which is a survey of actual firms, could fall as workers shift to work-at-home arrangements and the responsibility of filling out such surveys changes hands. The response rate for the household survey, which is literally a survey of households and includes our calculation of unemployment, could be much higher. I can’t thank her enough for taking the time to talk with me on these points.

Another issue: The 2020 Census was hiring and is ramping up. There are a lot fewer field offices, about half as many as there were in 2009-10. Look for a separate line on Census hires in the March data.

So where does that leave us? We are currently hoping to see a note on changes in how the BLS estimates the payroll data, but not assuming it for the initial report for March. That leaves us with a drop of 125,000 payroll jobs in March. Government sector jobs are expected to rise by 25,000, which means private sector losses will reach 150,000 if no adjustment is made to the data. The largest drop in jobs is expected to be in leisure and hospitality as major conferences were cancelled in late February. Traditional retailers will continue to be hit with the exception of Big-Box discounters and grocery stores. We started to hoard toilet paper, hand sanitizer and bleach wipes before shelter-in-place orders went into effect. We are expecting the number of professional hires to slow to a crawl. Health care jobs are expected to remain strong for March; the closing of dental offices and smaller clinics mostly occurred after the survey week.

Construction has been buoyed by unusually warm winter weather in recent months and will be fighting seasonal adjustment in March; this is the time of year it usually picks up. That said, this is an industry that was strong and was even able to get waivers for its workers to be classified as essential when shelter-in-place orders went into effect.

Average hourly earnings are expected to remain unchanged, as the loss in low-wage jobs is expected to be offset by higher paid salary jobs. The year-over-year gain in hourly wages is expected to drop to 2.7%, the lowest level in more than two years. Hours worked are also expected to contract. Hourly workers, mostly servicing the convention business, were let go or suffered a loss in the number of hours that they worked.

The unemployment rate lags and is expected to rise to 3.8%. That figure is expected to surge well into the double digits in April. Participation in the labor force is expected to tail off a bit. Who will be able to say they are actively looking for work when entire states are on lockdown? As the drop in employment accelerates, the details of the household survey are expected to carry more weight than the headline payroll figures.

Bottom Line
We already know that the fall in employment is apocalyptic. The survey data will eventually catch up to that reality. Even though that may come as soon as the April employment report in one month, the interim period will feel like an eternity.

Media Contact
Karen Nye
T +1 312 602 8973

Other Inquiries
Na Tasha Lowe
T +1 312 754 7368