A White House display turned into a broader dispute over which employment data deserve public trust. The key issue is not simply the line on the chart, but whether outsiders can examine how it was drawn.
Donald Trump embraced Stephen Moore’s jobs graph in the Oval Office as evidence of stronger job creation in the United States than under President Joe Biden. But the Heritage Foundation economist’s chart, presented after a weak Bureau of Labor Statistics jobs report, rests on unpublished Census Bureau figures and a comparison that cannot yet be fully checked.
The graph was presented as evidence of a stronger economy, yet its underlying data and comparison have drawn scrutiny. The dispute matters because a presidential jobs claim can change sharply depending on the dataset, the starting date and even the definition of employment being used.
The Oval Office chart display
According to The New York Times, Trump called reporters to the Oval Office in August 2025 to display charts prepared by Moore, a conservative economist affiliated with the Heritage Foundation. Trump called them “all new numbers” and said they depicted his economic record more favorably than the latest federal jobs report.
The appearance followed weaker-than-expected employment figures from the Bureau of Labor Statistics, or BLS. Trump criticized that report, claimed without offering evidence that the numbers were rigged, and fired BLS Commissioner Erika McEntarfer.
That sequence made Moore’s charts more than a routine political visual. They became part of a larger disagreement over whether widely used federal labor-market statistics should be accepted when they conflict with a president’s preferred account of the economy.
Why public data access matters
Moore said the charts used Census Bureau data that had not been published, according to the Times. That does not automatically make the figures wrong, since researchers and agencies can work with preliminary or specialized data before a public release.
It does, however, limit what outside analysts can test. Until the data are available, economists cannot fully examine the inputs, definitions, calculations or adjustments that produced the visual comparison.
That distinction is crucial when a chart is offered as proof of a president’s job-creation performance. A claim carries more weight when independent analysts can reproduce the result and measure it against established indicators.
The BLS publishes widely used monthly employment measures through its Current Employment Statistics survey, including total nonfarm payroll employment. Those estimates can be revised, sometimes substantially, as more complete information becomes available—a process that can complicate political messaging but is intended to improve accuracy over time.
A starting date can reshape results
A graph can be technically accurate while still giving readers a distorted impression if its baseline is unclear. A comparison that begins before or after a recession, a pandemic shutdown, a major data revision or a population change can produce a very different picture.
Trump’s first term makes that problem especially clear. Payroll employment increased before the COVID-19 pandemic, then fell sharply during the 2020 shutdowns. The recovery began before Trump left office and continued during Biden’s term.
A single uninterrupted line covering those years has difficulty separating the effects of presidential policy from an extraordinary public-health and economic shock. The question is not only how many jobs were gained or lost, but which period is being measured and why.
Early results in Trump’s second term also cannot be treated as the product of one administration alone. They may reflect conditions inherited from the prior administration, Federal Reserve policy, business investments made months earlier and revisions to earlier employment data.
The pandemic complicates every scorecard
BLS payroll data show that the United States added jobs in most months before the pandemic. Then, in April 2020 alone, payroll employment fell by more than 20 million jobs—an unprecedented loss.
The rebound that followed generated unusually large monthly gains in part because workers were returning to jobs lost during the shutdown. Those gains can be cited as evidence of a strong recovery, while full-term losses can be cited as evidence of a weak record. Both arguments can draw on real data.
Neither framing, by itself, resolves the more useful question: how did employment perform relative to the economic circumstances at the time? Presidents influence economic conditions, but they do not directly hire most of the workers counted in the monthly jobs report.
Economists therefore look beyond a payroll total. Other measures include labor-force participation, unemployment, wage growth, hours worked, the share of working-age people who have jobs, population growth and whether hiring is concentrated in particular sectors.
Competing measures need clear labels
Trump and Moore can argue that their chosen measure captures labor-market changes that conventional payroll data miss. Supporters may contend that official figures do not adequately show who is gaining jobs or whether growth has reached particular groups of workers.
Critics have a direct response: a chart challenging the BLS should disclose its method and rely on data others can independently review. Without that transparency, it is difficult to tell whether it identifies a meaningful trend or selectively supports a desired conclusion.
There is also a difference between criticizing a particular monthly report and alleging that the federal statistical system is manipulated. The BLS publishes its methodology, issues revisions and is staffed by career employees whose data are used by businesses, investors, state governments and economists across the political spectrum.
The unanswered test for Moore’s graph
Moore’s graph may remain part of Trump’s economic message, especially if it is used to argue that standard jobs reports miss a broader workforce change. Its persuasive power at a news conference, though, is separate from whether its underlying claims can be verified.
The deciding test is whether the Census data, formulas and comparison choices are made public and evaluated. Until then, the chart is best understood as an argument rather than a settled finding about Trump’s job-creation record.
For readers weighing competing claims, the useful checks are straightforward: identify the data source, examine the starting date, ask how employment is defined and determine whether the underlying figures are public. In this debate, those choices—not just the graph itself—are the central evidence.
Leave a Reply