The Damage to Economic Data May Already Be Done

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

If you have been closely following the ongoing Bureau of Labor Statistics story—in which Donald Trump fired then-Commissioner Erika McEntarfer after being displeased by the bureau’s July jobs report and selected the Heritage Foundation economist E. J. Antoni to succeed her—you will have heard an unusual consensus about the airtight political independence of the agency and the people who work there. Among BLS employees, including former Commissioner William Beach, whom Trump appointed in his first term, a fierce loyalty to the data is bone deep.

Antoni does not appear to share that spirit of independence, nor does he seem to have a great deal of talent for economics or statistics, according to economists from across the political spectrum. Even so, his power to avoid future reports that embarrass Trump appears to be limited. In an interview recorded on August 4, before his nomination, Antoni proposed eliminating the monthly release of employment data, but the administration has already insisted that that won’t happen. BLS data may not be completely tamper-proof, but they’re pretty close. The sharpest economic minds in this country, both inside and outside the bureau, pay meticulous attention to the deepest layers of the data, many strata below the headline-unemployment rate and change-in-payroll employment. Deceiving them all would be very hard to do.

Unfortunately, that might not matter. Antoni doesn’t have to manipulate any data to undermine the reliability of the government’s economic statistics. That damage might already have been done.

I was a career press official at the Department of Labor who prepared a series of labor secretaries for their TV appearances early on the first Friday morning of every month. The release of the jobs report—“Jobs Day”—is a marquee event in this little corner of the federal government, when the press and the financial world’s attention is fixed on the plaza of the Frances Perkins Building, in Washington. I lasted only one Jobs Day into the tenure of Trump’s labor secretary, Lori Chavez-DeRemer, before taking DOGE’s buyout deal. I decided to leave the government in large part out of fear of precisely the kind of demands for oaths of political loyalty that were being threatened then and are now being implicitly exacted on every career civil servant at the BLS.

Brian Klaas: Will Trump get his Potemkin statistics?

Most labor secretaries, understanding the power of jobs data to create or destroy value in the financial markets, have taken a sober and restrained approach to these press appearances. Then there’s Chavez-DeRemer. One of her prime talking points has been that “native-born workers have accounted for all job gains since Inauguration Day.” Every single one. Not a single Russian surgeon or Canadian blackjack dealer got a job after January 20 of this year. In fact, the BLS makes no such assertion. The claim is absurd on its face—the kind of political catnip that a Cabinet secretary in the Trump administration is expected to put forward without shame, as a kind of homage to the boss.

The existence of an independent BLS commissioner is predicated on the idea that someone needs to talk about the labor market who is never tempted to say such things. It’s a public service, primarily for investors. Might a member of the Cabinet say something iffy as a result of her political loyalties? That’s not ideal, but here’s someone else you can listen to who doesn’t have that problem. Until now, this arrangement allowed the president’s representative to attempt to convince the public of the effectiveness of his priorities while reinforcing the objective, nonpartisan genesis of the underlying data. If the BLS commissioner is now every bit the political animal that the labor secretary is, then what is the purpose of the BLS commissioner?

I am not a statistician; perhaps Antoni can mandate methodological deviations that bias the numbers in Trump’s preferred direction. But I don’t think he needs to. Confidence in the bureau is already badly weakened. This is about more than just our trust as consumers of the jobs report, because we are also its producers. To create its reports, the BLS needs businesses and citizens to take the time to respond to surveys about changes to their payroll and about who is going to work or looking for a job in their household. Even before Trump won the election last November, the trend in survey responsiveness was declining, posing an existential threat to the robustness of the data.

The appointment of a transparent partisan to the head of the BLS is unlikely to improve matters. Why should we take the time to report our economic circumstances to the government if we believe the government isn’t interested in the truth? If fewer Americans think that contributing to the creation of these reports is a valuable use of their time, the civil servants at the BLS will struggle to produce reliable numbers, regardless of what policies Antoni puts into place. The damage to our understanding of the economy would be far more consequential than a month of bad jobs numbers.