The Fed Will Key Off Unemployment Amid Mixed Jobs Data

The 272,000 increases in non-farm payrolls in May far exceeded investors’ expectations and continues a pattern of strong jobs gains over the past 12 months from the BLS employment survey. During that span, nonfarm payrolls have increased on average by 230,000 per month.

This pace far exceeds what most economists assume steady-state nonfarm payroll growth to be at about 100,000 jobs. It is the pace of hiring between the business cycle peaks of 2007 and 2020, and also matches the growth of the working age population.

Accordingly, many investors now question whether the Fed will lower interest rates in the balance of this year. However, the full story of what is happening in the labor market is complex because other indicators paint a different picture.

For example, employment measured by the household survey fell by 408,000 in May. This survey, which counts the number of people employed, shows that jobs have increased by an average of only 30,000 per month over the past 12 months.

As shown in the chart below, the two measures of employment are in synch much of the time, but there has been a growing disparity between them over the past couple of years.

Employment Measures: BLS versus Household Survey

So, what accounts for this disparity, and does it matter to the Fed?

Jim Glassman, an economist formerly with JPMorgan Chase who monitored the Fed, explains the differences in a recent memorandum for his clients. The BLS survey is the most commonly cited jobs measure, because it covers 666,000 workers and is less volatile compared with the household survey that covers 60,000 households. BLS estimates of jobs associated with business startups and failures are based on a birth-death model, which then get recalibrated in annual benchmark revisions.

One difference between the two measures is the household survey includes domestic work and self-employed, as well as workers who are paid off the books.

Another difference is that people who work multiple jobs may be counted more than once in the BLS survey. This is an important factor contributing to the widening disparity in the two reports, because moonlighting has rebounded in the past few years after it weakened during the pandemic. This shows up in the data as an acceleration in the number of gig workers and a slowdown in the number of full-time workers.

Economists also suspect that increased immigration might be boosting nonfarm payrolls, but it is difficult to know the magnitude. The wild card is the impact of undocumented immigrants. Don Luskin of Trend Macro contends this is the main source of increased job growth recently in a Wall Street Journal opinion piece.

Finally, another consideration is that estimates of private-sector employment by the BLS and the ADP typically agree on broad trends: Both data sets show private-sector payrolls have increased by about 200,000 per month over the past year and a half.

Weighing these considerations, Glassman does not believe the disparate employment trends are a conundrum for the Federal Reserve. He observes: “No one knows what the optimal pace of hiring is, given the unknowns about immigration and moonlighting. And that’s why the Fed’s annual statement of its goals includes no numerical goal for the pace of job growth that is consistent with the maximum employment mandate. (Bottom line: “maximum employment” is best characterized by unemployment.)”

The picture about the job market comes into clearer focus when one looks at the trend in the unemployment rate: It has been creeping up for a year now and did so again in May, when it approached 4 percent for the first time in two years despite strong growth of nonfarm payrolls.

Unemployment Rate, July 2019 to May 2024

Requests for unemployment insurance benefits have also been turning up marginally. This information tracks people who already are in the job market, had a job, and have credits with unemployment insurance system. The data are monitored closely, because they are timely (reported with a one-week lag) and are based on actual counts rather than surveys.

Additional evidence about the jobs market softening comes from the Jobs Opening and Labor Turnover Survey (JOLTS) report. Jobs openings fell more than expected in April and pushed the number of available jobs per jobseeker to its lowest level in nearly three years. The ratio of openings for unemployed workers declined to 1.24 in April. It is now well below its post pandemic peak of nearly two to one and is closely watched by Fed chair Jerome Powell as a measure of labor market tightness.

The bottom line is that it is reasonable for investors to conclude the strong payrolls report for May rules out any Fed easing this summer. However, the gradual softening in the labor market underway including an uptick in unemployed workers still leaves the door open for the Fed to lower rates later this year.

Meanwhile, investors will get a glimpse of what Fed officials are thinking in this week’s FOMC meeting when the next quarterly FOMC projections are released. The projections will likely show most FOMC members now foresee one-two rate cuts this year versus three in the March survey.