The Surprising May Employment Report

Chief Economist Scott Brown discusses current economic conditions.

In contrast to expectations of further deterioration, the May Employment Report suggested significant improvement in labor market conditions. No doubt, the economy has turned the corner as states have re-opened. However, the report was sharply at odds with the number of individuals receiving unemployment benefits. Even with the May job gains, we still have a long way to go for a full recovery. For the financial markets, the direction of change is what matters.

The headline data on jobless claims are for regular state programs. They don’t include separate programs for government workers, newly discharged veterans, pandemic assistance, and others. The number of individuals receiving unemployment benefits in all programs totaled about 30 million for the week ending May 16, a little more than 11 million more than in mid-April (corresponding to the timing of the payroll surveys). The Bureau of Labor Statistics reported that nonfarm payrolls rose by 2.5 million (median estimate: -8.0 million) in the initial estimate for May (3.1 million in private-sector payrolls). Payrolls fell 19.5 million from February to May. So how can 30 million be collecting unemployment benefits? What to believe? Jobless claims data are actual counts for the most part. Payrolls are an estimate, subject to statistical error. Still, most economists were expecting very sharp payroll gains for June (reflecting a re-opening in state economies) – maybe some of that showed up a little early. The jobless claims data still bear watching closely. Drilling down into the report, dentist offices accounted for 8% of the gain in nonfarm payrolls. Leisure and hospitality accounted for half.

Scott Brown
Click here to enlarge

The unemployment rate fell to 13.3% in May (median forecast: 19.8%), vs. 14.7% in April. The May figure was understated by about three percentage points due to a classification problem (April was understated by about five percentage points). The “insured unemployment rate” from the jobless claims report is continuing claims as a percent of nonfarm payrolls (from the end of the previous quarter) – 12.9% for the week ending May 16 (not seasonally adjusted). That reflects claimants for the regular state programs. Adding in the other programs, the insured unemployment rate would have been 20.6% – again a sharp contrast to the BLS’s figure.

The ISM surveys remained in contraction in May, with the headline figures inflated by an increase in supplier delivery times (which reflect pandemic disruptions rather than underlying strength in demand). In manufacturing, new orders, production, and employment continued to contract sharply. In the non-manufacturing survey, business activity, new orders, and employment continued to decline, but not as much as in April. Unit motor vehicle sales rose to a 12.2 million seasonally adjusted annual rate, up from 8.7 million in April (but still down 30% from a year earlier).