We try to avoid politics in this newsletter, but it’s impossible to discuss the nominee for commissioner of the Bureau of Labor Statistics without it feeling at least a little political. Rest assured, we would write the exact same column if the parties of the players were reversed.
The July jobs report was bad enough that President Trump (metaphorically) killed the messenger, firing the commissioner of the BLS later that day. This raised some alarm bells. America has the best economic data agencies in the world, and it’s critically important that their output remains unaffected by political considerations.
But hey, no worries. There’s a deep bench of qualified conservative economists that could be nominated to run the BLS.
Except there actually are some worries, because the president skipped those qualified conservative economists and nominated E.J. Antoni instead.
E.J. Antoni is not qualified to run the BLS. He has a PhD from Northern Illinois (no problem there; good football program, too), but he graduated just five years ago, has never worked with the kind of large statistics collection processes relevant to BLS operations, and has just one academic citation (about 1,300 fewer than his predecessor).
So he’s green. Which isn’t ideal, but that alone isn’t a huge cause for concern. The bigger problem is that he’s simply too partisan for the position.
And we don’t mean partisan as in he’s conservative. Bill Beach, who Trump appointed to run the BLS during his first term, is also a staunch Republican. Like Antoni, Beach was an economist at the Heritage Foundation, a conservative think tank. Prior to running the BLS, Beach served as Chief Economist for the Republicans on the Senate Budget Committee.
[As an aside, here’s what Bill Beach had to say about the firing of the BLS commissioner and quality of the data…]
We mean the kind of partisan where you incorrectly interpret data (presumably by intent) in order to push a narrative. Antoni’s mistakes are frequent, egregious, and include: not understanding how the import price index works, using a dual Y axis to make unconnected series appear coincident, inappropriately fitting lines onto data series to suggest data manipulation, and asserting that we’ve been in recession twice in the past three years.
Here’s a thread from Brian Albrecht, a free market economist, on some of Antoni’s more glaring offenses.
Conservative economist Stan Veuger, quoted in the Washington Post, summed it up nicely:
What’s wrong with a partisan BLS commissioner?
Statistical agencies need to be politically independent. These data are important and need to be trusted, and even the possibility of political meddling in the data (which we don’t necessarily expect, more on that below) would diminish that trust.
Regarding the importance of BLS data:
Investors pay close attention to it, including data on jobs and inflation. There’s a reason markets move on these releases.
The Fed bases interest rate decisions on these data.
The consumer price index is used to make cost of living adjustments for social security benefits.
Regional employment data are critical. There are 50 states, each of which pursues different economic policy. Knowing how those economies perform lets us know which policies are or are not working. The south, for instance, has been kicking ass the past decade or so, while the northeast has been on the other end of the kicking. That tells us something about the implications of tax, zoning, and other regulatory policymaking.
Businesses use BLS data to inform their operations, both for investment and geographic selection decisions.
Construction firms have contractual agreements to use producer price index data to adjust their prices over the duration of a project.
Could a commissioner manipulate data?
Look, it’s entirely possible that Antoni comes in and does an upstanding job of running the BLS. But just for a moment, let’s assume he wants to manipulate the data.
The BLS is big. Dozens of people work on the monthly jobs report. We’d know if a new commissioner came in and said, “no, not enough jobs, need to pump that higher.” Whistles would blow.
The more probable downside scenario is that changes are made or orders are given that workers find so objectionable that they resign. In the short term, the already short-staffed agency becomes even shorter-staffed, degrading the quality of data. In the long term, positions are filled with workers who may not object to morally questionable methodological alterations.
What else could go wrong?
Antoni’s comments about suspending the monthly jobs report, arguably the most important economic data series, come to mind. Granted, he said that before he was nominated, but there are signs that the administration intends to make major changes to the jobs report.
So we stop producing the monthly jobs report, spend a few months in the dark, and then some brand new data series—one we have no context for—comes on line to replace it.
This brings to mind the Chinese approach to economic data. When their youth unemployment rate hit a record high in June 2023, they stopped publishing it for a couple months and then brought it back with a new methodology that just happened to produce a rate 7 percentage points lower than the previous one.
We tend to prefer the U.S. approach to economic data production.
Are we overreacting?
Hopefully!
But also, no. We aren’t. Even if Antoni is a good—or even great—BLS commissioner, this doesn’t serve the president’s interests. Let’s say Antoni comes in and does the job with integrity, doesn’t exert political pressure, maybe even improves some methods. If the economy happens to boom during the first few months of his tenure, do investors trust that data? Does the Fed?
It’s like if the Russians sent a KGB operative to judge an Olympic figure skating competition. Even if he’s perfectly unbiased, it’ll be hard to trust his scores if the Russian skater takes gold.
But is the BLS already manipulating data?
No, and Antoni—to his credit—acknowledges this. The big objection to the BLS right now is the revisions to jobs data, and as Jeremy Horpedahl shows here, the revisions have: 1) gotten smaller over time and 2) have tended to benefit Republicans by showing stronger initial employment gains.
That said, BLS methods could definitely be improved. Falling response rates (which has happened in most countries, not just the U.S.) is a real problem, albeit one that probably requires more funding to fix. Any improvements to BLS data production, however, should be done gradually, apolitically, and by someone experienced with this kind of work.
What’s next?
This week is packed with economic data, including yesterday’s all-important consumer price index release. We’ll cover that and a lot more in Week in Review, our every Friday post that covered everything you need to know about the economy in a breezy, 5 minute read.
Week in Review is only for pay subscribers. If that’s not you and you want it to be, just click the subscribe button.
Should he manipulate BLS data, to what extent can we detect the manipulation (and how quickly) via other data sources? i.e. ADP jobs reports, etc.