Gordon's Guesstimate: June Jobs

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

Last month’s job report showed the single largest one-month gain in history and was the most conspicuous upside economic surprise the nation has seen as it recovers from the recession induced from the COVID-19 pandemic. It would take 8 more months of similar employment gains to replace the jobs lost in March and April, however. The persistent question, of course, is whether the economy can sustain such gains. At least for June, that seems like a safe bet, despite storm clouds in the offing.

The troubling increases in infections and hospitalizations in nominally reopened states, and observed weakening in some higher-frequency data will likely not be captured in the June Employment Situation Report. Recall that for the household survey, which captures the unemployment rate among other data, the reference period is the calendar week with the 12th day of the month. In the establishment survey, which captures the monthly change in employment, the reference period is the pay period including the 12th. Thus, for the June report, the report will reflect survey responses as of the second week in June, right about the time caseloads began to pick up in some states.

Because of this timing, the report will reflect the incremental improvement in the economy identified in the contemporaneous indicators. From the May to June reference period, initial unemployment insurance claims total over 7.1 million. But for the recent past, this level of new unemployment insurance (UI) claims would be shocking, but the trend has been a steady decline in new claims. Indeed, average new UI claims for the 4-week period between the May and June reference periods is down by over 1.2 million claims from the preceding 4-week period.

Additional indicators point to a pickup in employment running to the millions, not least of which is the ADP survey. Over the long term, the ADP and Bureau of Labor Statistics (BLS) survey results line up rather closely, but in any given month the two series can wildly diverge. These comparisons, however, are based on monthly employment changes that are at least an order of magnitude smaller than the flows the next several months will likely witness. It would be surprising if this morning's employment report were directionally or a multiple apart from ADP. Other contemporaneous data are further suggestive of continued backfilling of the jobs lost, on net, in March and April.

The closure of data collection centers, lower response rates, and the necessary suspension of in-person interviews have challenged the BLS in its data collection, though the agency has performed a creditable public service in clearly identifying these challenges. In particular, the agency has been highly transparent in explaining its methodology and the implications of the issues the agency is facing in data collection, particularly with respect to the classification of certain workers in the survey data. As has previously been discussed, the unemployment rate would be about 3 percentage points higher if these classification errors were eliminated.

This morning’s report will almost certainly indicate a million-plus pickup in employment. Forecasters are reportedly predicting anywhere from 2 to 7 million new jobs in the June report. Such a broad range of estimates is suggestive of the tremendous uncertainty that continues to attach to these figures. This guesstimator is assuming a 3 million pick-up in employment. I assume that the net effects of this gain, improvements in classification, and an increase in labor force participation will contribute to the U-3 remaining flat. The influx of more workers will likely pull the earnings data down, I’m assuming by 5 cents.