During a campaign rally in North Carolina last August, Donald Trump accused the Biden administration of cooking employment statistics, claiming his political opponent was trying to inflate job growth numbers until after the election. The comments came after the Labor Department’s Bureau of Labor Statistics issued a sizable downward revision to its employment estimates as part of the agency’s annual benchmarking to improve the accuracy of the monthly jobs reports.
Trump’s unfounded accusations were par for the course in a campaign that frequently boosted conspiracy theories, but they may have been a harbinger for how his administration would approach federal economic statistics.
Since taking office, Trump’s political appointees have criticized how some economic measurements like gross domestic product (GDP) are calculated. Simultaneously, the administration has shuttered several advisory committees of independent economists, statisticians, academics, and industry representatives that helped federal economic agencies maintain the accuracy—including keeping after-the-fact revisions to a minimum—and relevance of those metrics. The moves have raised concerns about the administration’s commitment to the political independence and integrity of numbers that literally move markets and inform nearly every element of economic decision-making.
The Bureau of Labor Statistics (BLS) alone produces dozens of statistical products including major economic indicators like the consumer price index and monthly jobs reports but also measurements like National Longitudinal Surveys, which track cohorts of people over the course of their lives to capture not only employment trends but also patterns of major life events, collecting unique data that are used by researchers in a variety of fields beyond economics.
Former officials and advisory committee members told The Dispatch they’re concerned that the administration could undermine confidence in these key economic statistics. The risk, they said, isn’t necessarily from partisans trying to skew the data. Rather, the danger lies in a slower atrophying of data integrity as already under-resourced agencies face budget cuts and lose access to valuable expert guidance that has improved statistical accuracy across decades of economic change.
Earlier this month, the Trump administration terminated multiple economic advisory committees that assist the federal government’s three main economic statistical agencies: BLS, the Bureau of Economic Analysis (BEA), and the Census Bureau. The closures began with the Commerce Department ending the Federal Economic Statistics Advisory Committee (FESAC)—a panel that advised all three agencies—and the BEA Advisory Committee. Commerce Secretary Howard Lutnick said in a public notice on the committee’s websites that their purposes “have been fulfilled.” Two weeks later, the BLS informed the members of two of its advisory panels—the Technical Advisory Committee (TAC) and the Data Users Advisory Committee (DUAC)—that the committees had been terminated.
The latter cancellations, first reported and confirmed by The Dispatch, were done so quietly with the administration deleting the committees’ webpages instead of publicly posting a termination notice. Echoing Lutnick’s comments, a spokesperson for the Labor Department told The Dispatch that the BLS committees’ purposes have been fulfilled and that there will be “no effect on the functionality of BLS.”
Steve Landefeld, who served as BEA director from 1995 to 2014 and was involved in the creation of FESAC in 1999, told The Dispatch the idea that the committees have fulfilled their purpose misses entirely the nature of statistical work. “It just takes my breath away, the thought that we had one simple problem, and we’d fixed it,” Landefeld said. “I wish that were so.” Measuring things like GDP, inflation, and employment is a massive undertaking that involves constant adjustments and technical challenges to accurately capture economic change. Landefeld likened BEA’s job to a tailor trying to measure someone for a suit while they’re going for a run.
“The committees were and would continue to fulfill their function because the issues, the challenges, the questions, never go away,” Michael Horrigan, a former associate commissioner at BLS until 2019 who oversaw the jobs and inflation measurement programs, told The Dispatch. “They’re always there. There’s always room for improvement.”
Susan Houseman, a labor economist who chaired the BLS’ Technical Advisory Committee for more than a decade, told The Dispatch, “You can’t just be doing the same surveys, doing things the same way as you always have been; they’re constantly updating things, improving things.” She said BLS leaned on the committee to both conceptualize and implement necessary changes, noting the employment data revisions that prompted Trump’s August comments are exactly the kind of thing the TAC would help the agency improve.
While the issues the statistical agencies seek advice on are often technical and marginal, the value of small improvements accumulates over time. For example, Landefeld emphasized the advisory committees’ work supporting the expansion of surveys for the services sector and improving measures to capture quality advances in the IT sector. “It just tremendously improved the accuracy of our estimates of the services sector and of overall real GDP,” Landefeld said of the efforts. “That’s important, because if you want the economy to grow up to its potential, you got to know what its potential is.”
Agencies are currently grappling with how to measure the value of artificial intelligence. Avinash Collis, a professor at Carnegie Mellon University who researches how to measure the digital economy, served on FESAC and told The Dispatch that he provided guidance on how the statistical agencies could structure their nascent AI measurements to allow for data comparison across metrics.
The committees were terminated pursuant to Trump’s executive order on reducing the federal bureaucracy and improving efficiency. Government agencies can often fall prey to bureaucratic bloat and end up creating senseless roadblocks. But these advisory committees cost next to nothing—the members were volunteers reimbursed only for their travel expenses. “The cost-benefit analysis on these committees was wildly in favor of the taxpayer,” David Wilcox, a senior fellow at the Peterson Institute for International Economics and the director of U.S. economic research at Bloomberg Economics, told The Dispatch.
Wilcox, who served as the most recent FESAC chair, described the committee as a unicorn, a force multiplier. Agency principals worked closely with the advisers to address real-time problems, Wilcox explained, which in turn attracted committee members with world-class expertise, including Nobel Prize winners, since they knew their input would be taken seriously. “It wasn’t that these statistical products were fully baked, and they just wanted to come and educate the committee about what the agencies were about to do,” he said of BLS, BEA, and Census Bureau officials. “It was, ‘Let us show you our homework at an intermediate stage. Look over our shoulder and tell us how you would confront these difficult problems.’”
These dynamics were on full display at the BEA Advisory Committee’s final meeting, which is available on YouTube, where two of the government’s highest-ranking statistical officials, BEA Director Vipin Arora and the Commerce Department’s Chief Data Officer Oliver Wise, spent hours engaging with members. “We need your advice on all of this,” Arora told the committee when discussing how the agency should prioritize statistical products in light of worsening budget constraints.
“I’m not worried about what happens in the next year or two, or even three, with the termination of these committees. I’m worried more about over the longer term without having a systematic vehicle for providing input and expertise.”
Susan Houseman
Landefeld, whose decades-long tenure at BEA predated the committee system, explained how feedback on the statistical products and data used to be delivered more circuitously from outside researchers. Bringing outside perspectives directly into BEA’s work resulted in not only technical innovations but also provided a salutary credibility to the agency leadership’s efforts to implement improvements. “It’s one thing to try and lead from the head of the statistical agency,” Landefeld said. “It’s a much more powerful thing if you can lead with the support of some of the leading academic and nonprofit researchers in the country.”
The officials and former committee members The Dispatch spoke with stressed that the statistical agencies’ staff are highly competent and will continue doing quality work despite the elimination of the advisory committees. Horrigan, the former BLS associate commissioner, said agency staff will still access some external advice, at economic conferences, for example, but it would be on a more ad hoc basis.
“I don’t want to exaggerate this. The BLS staff are great. They’re going to do a good job,” Houseman said. “But we do genuinely help them and provide real benefits to them, provide quite valuable input. So, I’m not worried about what happens in the next year or two, or even three, with the termination of these committees. I’m worried more about over the longer term without having a systematic vehicle for providing input and expertise.”
The loss of committees also comes as the agencies have been dealing with declining budgets—a problem that has been building for years. Their appropriations have not kept pace with inflation even as the costs of producing measurements have increased. Several committee members and former officials cited the decline in survey response rates—a basic collection point for economic data—as a key cost factor, forcing agencies to spend more than in the past to generate the same number of responses. If budgets and personnel are cut further—the Commerce Department is reportedly planning to slash 20 percent of its staff—the agencies will be forced to stop producing some economic measurements and or cut back on things like survey sample size.
“These agencies are pretty freakin’ efficient,” Horrigan said. “If they get a budget cut, the answer is going to be very simple. They’re going to eliminate surveys. They’re going to eliminate products, period.”
Data accuracy is a matter of degree, not a binary right or wrong. Eliminating statistical products and cutting costs for current metrics would reduce both the accuracy of individual measurements and the overall picture of the economy available to decision-makers. “Suppose a statistical agency had to cut costs,” Wilcox explained, “one way they could do that is by cutting the sample size in a particular survey by half. Instead of surveying 100 households, or 100 businesses, they could survey 50. Well, what effects would that have? That would mean that the statistical noise embedded in the survey inevitably would be greater.”
The consequences of a fuzzier sense of economic reality would be felt far and wide. “It would adversely affect the Federal Reserve as they make the tough decisions that are inherently part of monetary policy, having to decide in real time, based on perceived trends in inflation and the pace of economic activity, whether to raise interest rates, lower interest rates, or keep them the same,” Wilcox said. “Business leaders having to make decisions about where to invest, how much to employ, how much more capital to invest in. Households who want to know where are the best employment opportunities likely to be in the future. Where should I move to? What form of college or higher degree should I get? All these are examples of decisions that need to be made where better information, more granular information, more timely information, more accurate information is really valuable.”
Beyond potential budget cuts, Trump officials’ recent comments about GDP have also worried observers. In what appeared to be an attempt to get ahead of any effects of Department of Government Efficiency (DOGE) cuts, Elon Musk and Secretary of Commerce Howard Lutnick have both questioned what goes into GDP numbers. Reacting to a report on projections of slower economic growth, Elon Musk said last month that government spending should be stripped out of GDP. Lutnick seemed to follow Musk’s lead, saying a few days later, “You know that governments historically have messed with GDP,” adding that he was going to separate government spending to “make it transparent.”
Neither Lutnick nor Musk seemed to know that BEA already produces GDP measurements without government spending—the government and the non-government numbers produce nearly identical growth results. There’s been no reported follow-up on any plans to change how GDP is calculated, but the apparent impulse to tinker with the numbers first rather than the White House’s policies has some worried about the continued political independence of economic statistics. Given the past statements from the president and other administration officials, it’s at least an open question as to how the White House would treat the data results if the economy sours.
The Trump administration is in the process of introducing a new category of federal employee that could reclassify thousands of positions as political and remove their civil service protections. Trump officials like Office of Management and Budget Director Russ Vought don’t believe that any executive branch agencies should be allowed to operate independently from the president’s agenda.
But for now, Landefeld and Houseman both emphasized that the agency staff, who are almost entirely career civil servants, are committed to producing independent, nonpartisan statistics and would call out any political meddling in their work. “There’s lots of external and internal checks, which would make it hard to manipulate the numbers,” Landefeld said. “You couldn’t keep that conspiracy going at all. There’s just too many people involved who would talk about it if that were to happen.”
