The crucial U.S. jobs report has developed a bad habit: It keeps exaggerating how many new jobs are created each month.
The government’s early estimate of new job gains has overstated the increase in hiring in 15 of the past 17 months dating to the start of 2023.
The U.S. labor market is still quite strong, to be sure, but a less reliable jobs report could cause Wall Street or the Federal Reserve to misjudge how fast the economy is either growing or slowing. The Fed’s decision on whether to cut interest rates could also be affected, with big implications for the economy.
The problem has been especially pronounced this year. The jobs report has overstated the gain in new jobs by an average of 50,000 a month from January to May.
See historical revisions to U.S. jobs report
Consider some recent examples. The report originally showed a 272,000 increase in new jobs in May, for instance, but was revised down last Friday to 218,000. That’s a 54,000 reduction.
Or take January. The preliminary estimate of a stunning 353,000 increase in new jobs was later revised down to 256,000. That’s a 97,000 drop.
The problem has not gone unnoticed by Wall Street economists or the Federal Reserve.
At the Fed’s last meeting in early June, for instance, several senior officials said the employment report “may have overstated actual job gains,” according to a summary.
The root of the problem, asserts chief economist Richard Moody of Regions Financial, are low response rates to the government survey that determine how many jobs are created each month.
The BLS’s so-called establishment survey derives its estimate of how many jobs are created by polling some 600,000 business sites around the country.
Yet since the pandemic the percentage of businesses responding to initial survey each month has fallen dramatically.
In 2023, the initial response rate sank to a 21-year low of 64.5%. And just 61% of business sites have responded promptly to the questionnaire through the first six months of this year.
When fewer businesses respond, Moody contends, the preliminary jobs report is prone to be less accurate.
“Low collection rates lessen the reliability of the initial estimate of job growth in any given month and leave the door open to sizable revisions in subsequent months,” he wrote earlier in the year.
That’s exactly what has been happening.
Most businesses surveyed eventually turn in their answers to the Bureau of Labor Statistics, the agency responsible for compiling the jobs report. The response rate for the government’s final estimate each month is more than 90% after late responses trickle in.
The BLS then finetunes its monthly job estimates to be more precise. Large revisions are not uncommon historically given the difficulty in tracking employment in such a huge economy, but they tend to be smaller and more sporadic.
The agency contends a low response doesn’t make its employment estimate less accurate based on the agency’s own research. But the BLS has been trying new ways to try to boost the initial response rate.
What worries some Wall Street
DJIA
SPX
economists is that downward revisions in monthly jobs gains tend to foreshadow a weakening economy.
Other signals also point in that direction.
The unemployment rate has risen steadily to a 2 1/-2 year high of 4.1% in June from a 54-year low of 3.4% in 2023, for one thing.
The number of laid-off workers collecting benefits, meanwhile, has risen to a highest level since 2018 if the pandemic years are omitted.
Whether those trends intensify — and foreshadow a soft patch in the economy — remains to be seen.
Even after big downward revisions this year in the employment report, the economy has averaged a strong 222,000 new jobs a month from January to June. That’s well above the the 183,000 average from 2010 to 2019.
While the labor market has undergone “considerable cooling,” Fed Chairman Jerome Powell said Tuesday, “it’s still strong.”