There are risks on both sides of our forecast given that some states continue to recover from the COVID-19 crisis while others, notably in the South, are seeing a sharp uptick in cases. A weekly Household Pulse survey from the Census Bureau is suggesting that June’s 4.8 million increase in jobs could have been fully erased. The reliability of this survey is not known, but it predicted payrolls well with its first test in June.
Still, more established surveys paint a less negative picture. The Dallas Fed’s service survey fell sharply in July but service surveys from the Philly and Richmond Feds showed modest improvements. Most manufacturing surveys were firmer, too. Initial claims ticked up in the July payroll survey week, but the 4-week average is still down from June’s survey. Labor-market perceptions improved in the Conference Board’s consumer confidence report despite an overall slip.
On balance, we expect a moderate increase in payrolls, with goods and most service industries continuing to recover, if not as strong as in June. However, we expect to see some losses in retail and leisure/hospitality and are forecasting a small decline in private-sector service jobs. We expect seasonal adjustments (assuming layoffs in education) to lift the number of government payrolls, with many teachers already having lost their jobs.
A rise of 350 thousand in employment would usually see unemployment edge lower. However, if the labor force continues to bounce back from recent declines, it could exceed employment growth, which an unexpected deterioration in the COVID-19 situation will restrain. We look for an increase in the labor force of 750 thousand, lifting unemployment by 0.2% to 11.3%.
Average hourly earnings have also been highly volatile in recent months, surging due to heavy losses in low-paid jobs and then correcting as some of those jobs returned. While low-paying work leads the downside risks this month, we do not expect the losses to be large enough to have a major impact on average hourly earnings. We expect an unchanged outcome, reflecting a subdued underlying picture. This would cause y/y growth to slip to 4.7% from 5.0%, which would still be above pre-crisis levels.
The workweek in June was still inflated by past losses of low-paying jobs, which tend to see shorter hours. Losses of such jobs may prevent a decline in July, leaving the workweek unchanged at 34.5 hours.