Gordon's Guesstimate: March Jobs

Authored by Gordon Gray, AAF’s Director of Fiscal Policy

In January and February, the U.S. economy added 546,000 jobs to American payrolls. Over the nearly 11 years since the official end of the Great Recession, the economic recovery created 21 million new jobs and drove the unemployment rate from 9.5 percent to 3.5 percent. And it’s now over.

This morning’s jobs numbers will show the beginning of the economic storm that’s already arrived on U.S. shores in the form of the COVID-19 pandemic. But when the numbers are released later this morning, they will show just the leading edge, and observers should be prepared for a jobs report that looks somewhat less dire than what contemporaneous indicators suggest. It’s therefore worth clarifying what the leading economic indicators can and can’t tell us.

The most frequent indicator that labor-market observers tend to look at is weekly unemployment insurance (UI) benefit claims. These data include initial claims and continuing claims, measuring the number of people newly applying for unemployment benefits and the total rolling number of individuals receiving benefits at any one time. These data are very noisy, so much so that monthly rolling averages are often not more descriptive of labor market movements than any single week print. But in the last two weeks, 10 million Americans have filed for unemployment benefits. The 6,648,000 new UI claims reported on Thursday is nearly double the claims reported on March 26^th, and nearly an order of magnitude higher than prior initial UI claims record of 695,000. No event in modern American history has ever precipitated such immediate and widespread job loss.

There is not a straight line between UI claims and the employment report, however. Most important, the employment report reflects the partial results of surveys of representative employers and households for reference periods covering the 12^th of every month. As more complete survey data are included over subsequent months, the jobs numbers are revised. When the “employment situation” report comes out today, it will thus reflect the labor market, as measured by partial surveys, in the first half of the month. Note that anyone who worked even a single day during the reference period will be counted as employed. Contemporaneous UI claims were somewhat elevated during that week, but well below the historic levels seen since. Accordingly, today’s jobs report should barely cover the leading edge of the economic storm that is manifestly upon us.

Does this mean the report is meaningless? No. It won’t tell us much about what the future holds, but these reports do provide substantially more detail about the labor market than more timely data. They should inform policymakers and help them scale and tailor the policy response. Today’s jobs report will not affect Federal Reserve policy or move markets, but going forward, the report will be critical in evaluating the economic consequences of the pandemic.

Today’s jobs numbers will essentially reflect a very strong labor market largely a week away from its most precipitous collapse in American history. Where the labor market was on this slope during the reference week is difficult to assess. The consensus among economists appears to be that today’s report will show the first payroll decline since 2010. But it is also possible payrolls may still be positive. ADP reported that private sector payrolls shed 27,000 jobs in March. A small cross-current is the addition of about 17,000 temporary Census takers to public rolls, as of the second week in March. The contemporaneous ISM manufacturing employment index fell considerably for the month, but again, that is not an accurate characterization of the particular week covered by the employment report. This guesstimator is assuming today’s report will show a 5,000 payroll gain, a slight increase in unemployment to 3.6 percent, and an 8-cents earnings increase.