Solving the Strong Labor-Market Puzzle 

Despite the Federal Reserve’s tight monetary policy, sticky inflation and slowing economic growth, the U.S. labor market continues to show resilience in early 2024. But that will likely soften later this year. 

The most recent Bureau of Labor Statistics (BLS) report estimates that nonfarm payroll employment increased by 303,000 in March, while the unemployment rate dipped incrementally to 3.8%. A smattering of other data also provide support for the notion that labor demands remain robust. For instance, initial claims for unemployment benefits have remained low. In the last week of March there were 221,000 claims, only 3,000 above the
pre-pandemic average for 2019. 

One puzzling aspect of the monthly employment data has been a large, ongoing discrepancy between payroll employment, which the BLS obtains from survey responses of about 275,000 businesses (Establishment Survey), and the number of employed persons obtained from survey responses of about 42,000 households (Household Survey.) Because the two surveys come from different samples and use different concepts of employment, different estimates are to be expected, particularly from month-to-month. Yet the surveys typically have been close over longer periods—that is until recent years.

The Establishment Survey initial estimate for jobs added in March was less than the increase in employment reported from the Household Survey, but that is an anomaly. During the 12 months ended in March 2024, the cumulative job growth implied by the Household Survey (after adjusting for conceptual differences with the Establishment Survey) was about 2 million less than the job growth estimate from the Establishment Survey. 

As it turns out, most of the discrepancy can be explained by an apparent undercount of the population in the Census Bureau’s latest estimate. To estimate employment from the Household Survey, the BLS expands the responses of the survey participants in proportion to the estimated size of the population. So, a smaller estimate for the population will result in a smaller estimate for the number of employed persons. In January 2024, the Congressional Budget Office (CBO) published its revised population estimates. The CBO’s estimate for 2023 is 2.2 million higher than the Census Bureau’s estimate. The difference is mostly the result of the CBO’s estimate of illegal immigrants based on Department of Homeland Security data. So, it appears that the Establishment Survey better accounts for the effects of an immigration surge than the Household Survey, and that the Establishment Survey’s higher job estimates are a better indicator of current labor-market tightness. 

However, an outlook for continued strength in demand for workers is called into question by weakening trends in other labor-related data, even though they appear favorable at current levels. Take, for example, the number of job openings. It peaked at 12.2 million in March 2022 and has since trended downward to 8.8 million in February 2024. 

Likewise, job opening rates, job-hiring rates, job-quits rates and wage-growth indicators all show deteriorating multi-year trends extending into the first quarter of 2024. That indicates that a notably softer demand for labor is likely to develop later in 2024.