After six months of haranguing the Federal Reserve to slash the federal funds rate despite a persistently strong economy, President Donald Trump finally got a data point proving his case that the labor market is likely softening. The Bureau of Labor Statistics announced in its July jobs report that, despite the economy adding a net 73,000 jobs, major revisions to its May and June figures showed a quarter-million fewer new positions in those months than the agency originally reported.

Trump promptly fired Erika McEntarfer, former BLS commissioner, and subsequently accused the BLS of repeatedly publishing “RIGGED” and “record-setting revisions” to benefit Democrats before and after his 2024 victory.

Recommended Stories

• Sephora urged to drop Huda Beauty line after founder accuses Israel of causing World Wars
• New Paramount head doesn't want to 'politicize' company as merger closes
• Alan Dershowitz refused a pierogi a second time at farmer's market

“Those big adjustments were made to cover up, and level out, the FAKE political numbers that were CONCOCTED in order to make a great Republican Success look less stellar!!!” Trump posted on Truth Social.

As was the case with Jim Comey’s tenure as FBI director in 2017, Trump had every reason to fire McEntarfer. And as with his belated termination of the utterly clownish Comey, Trump was right about firing the less clownish, but still incompetent, McEntarfer.

But Trump did so for the wrong reasons.

Though McEntarfer was indeed an appointee of former President Joe Biden, a bipartisan consensus of 86 senators, including Vice President JD Vance, a former senator, confirmed the economist as BLS commissioner in January 2024. Under her half-year tenure, the BLS’s revisions actually occasionally favored Trump, such as the infamously awful (and incorrect) October 2024 report saying the Biden administration only oversaw 12,000 new jobs added to the economy.

And recent revisions haven’t been quite record-setting; in an aggregation of jobs reports going back to World War II, statistician Nate Silver found that the May and June revisions were, respectively, only the 66th and 54th largest in the BLS’s history, with no other jobs reports published under McEntarfer’s tenure cracking the top 100 worst revisions.

But McEntarfer failed to solve the more important question of why still-sizable revisions were necessary: namely, the BLS’s cratering survey response rate.

The monthly jobs report, the first major economic record released to summarize the month prior, relies on two separate surveys, the Current Employment Statistics program and the Current Population Survey. The BLS is responsible for administering the CES, also known as the payroll or establishment survey, to some 119,000 businesses. The Census Bureau administers the CPS or household survey to 60,000 households. Whereas the CES tracks the broader macroeconomic trends of overall employment, average hours worked, and corresponding average earnings, the CPS zooms in on more demographic data.

Furthermore, while the CES supplies the gold standard economic gauge of nonfarm payroll employment, the CPS’s broader measure includes agricultural workers and self-employed workers.

Because the jobs survey is meant to be preliminary, we always expect some margin for revision. After all, our monthly schedule of a jobs report release on the first Friday of every month means that the BLS beats the South Koreans and the Canadians with publishing our preliminary employment report by about a week or two, and Norwegians by at least three weeks. Every other major OECD peer doesn’t release their monthly jobs report until at least a month after the BLS releases ours for a given month.

But the BLS is relying on increasingly insufficient data to do so.

Before the COVID-19 pandemic, around 83% of households contacted by the Census Bureau by telephone or in person responded to the CPS. When McEntarfer took office, the response rate fell to 70%, and by July, it plunged to 67%.

In absolute terms, the CES response rate was even worse, falling from 60% before the pandemic to barely 42% this spring. But to her credit, McEntarfer presided over a stabilization of the CES response rate, which is actually at the same level as when she was confirmed.

Trump is not wrong to fire McEntarfer for failing to improve the response rate. But that’s not the end of the story.

Trump must hold accountable the Commerce Department, which oversees the Census and its free-falling CPS response rate. And Commerce Secretary Howard Lutnick has not been a passive observer in the 3% decline in the CPS response rate; he has actively made it worse.

Back in February, Lutnick terminated the Bureau of Economic Analysis Advisory Committee and the Federal Economic Statistics Advisory Committee. This was not to fulfill Trump’s cost-cutting ambitions. The committees, which had an annual combined budget of $200,000, which is dust mites in D.C. terms, relied on unpaid economists volunteering to help improve and verify the accuracy of the federal government’s economic data, crucially inviting private sector and even foreign economists to bounce new ideas off career government wonks.

In particular, FESAC was laser-focused on improving the response rates of surveys conducted for the BLS, Census, and Bureau of Economic Analysis. In the year before FESAC’s untimely demise, the committee weighed plans to digitize and modernize the CPS and CES for economists from the Federal Reserve Bank of Chicago, Goldman Sachs, and the national statistics offices for the United Kingdom, Germany, and Canada.

TAXPAYERS NEED GOOD GOVERNMENT DATA

Nobody inside or outside the BLS or the Commerce Department can explain why Lutnick killed FESAC and BEAAC. What we do know is that if Trump wants the U.S. dollar to remain the world’s reserve currency and the long end of the bond yield curve to decrease before we have to refinance more of our debt, investors must maintain trust in the government’s employment data.

By all means, Trump had good reason to fire a BLS commissioner who failed to solve the bureau’s most pressing problem. But his next nominee better have a solution, and Lutnick put a plan in place to replace the statistical advisory committees. Otherwise, McEntarfer’s successor won’t last any longer.