The Numbers We Argue About:
What BLS Jobs Data Really Shows.
The monthly jobs report is one of the most-cited numbers in American life, and one of the least understood. Both parties wave it around when it flatters them and dismiss it when it doesn’t. The truth is more interesting than the talking points: the labor market that produced the “Great Resignation” has quietly returned to normal, and the data that measures it has developed a real reliability problem.
Two years running, the Bureau of Labor Statistics has issued the largest downward payroll revisions in the history of its modern series — nearly a million jobs erased in the latest one. Those revisions document weakness in numbers first reported during the Biden administration. That is an accuracy story, not a partisan one, and it matters precisely because the data is supposed to be above the fight.
Then the fight came anyway. After a weak July 2025 report, President Trump (R) fired the BLS Commissioner and called the numbers “RIGGED.” Multiple fact-checks found no evidence of manipulation — the Commissioner cannot alter the computer-generated figures. This is what happens when a measurement everyone depends on becomes a thing everyone suspects.
- 2.0%quits rateMarch 2026 JOLTS — back at the pre-pandemic norm, down from a 3.0% record in late 2021 — BLS / FRED JTSQUR
- -911,000jobsthe 2025 preliminary benchmark revision — largest on record back to 2002 — BLS CES
- 4.6%NY joblessNew York unemployment (April 2026) vs. 4.3% nationally; NYC ~5.6% — BLS
- Aug 1, 2025BLS chief firedTrump dismissed Commissioner Erika McEntarfer after a weak July report — fact-checks found no manipulation
The Bureau of Labor Statistics publishes the headline jobs numbers every month, and almost no one reads past the top line. In April 2026, the BLS reported the unemployment rate at 4.3%, nonfarm payrolls up 115,000, and labor-force participation at 61.8%— the lowest reading since October 2021. Each of those numbers comes from a different survey, with different sample sizes and different margins of error, and each gets revised as more complete information arrives.
That revision process is normal and necessary. What is not normal is the size of the corrections the agency has had to make over the last two years — and the political firestorm that erupted when one bad month collided with an election-year nerve. To understand why the jobs data became a battleground, you have to understand two things that are actually true about the labor market, and one thing that is not.
In July 2022, the BLS itself published a careful analysis in the Monthly Labor Reviewtitled “The ‘Great Resignation’ in perspective.” The phrase had been coined by organizational psychologist Anthony Klotz, and by late 2021 it described something genuinely unprecedented: the quits rate hit 3.0%in November and December 2021 — a record since the JOLTS series began in December 2000 — with roughly 4.5 million people quitting in November 2021 alone and about 47.8 million quits across all of 2021.
The BLS analysis made two points that complicate the headline. First, the quits surge rose too fast to be explained by ordinary labor-market tightening alone — it carried pandemic-specific drivers, from health fears to childcare disruptions to stimulus-cushioned savings. Second, quitting is not new: the article noted that quit rates in manufacturing were actually higherin the 1960s and 1970s. The “Great Resignation” was real, but it was a spike, not a permanent rewrite of how Americans work.
“The surge in quits rose too quickly to be explained by a tightening labor market alone — pandemic-specific factors were at work, and quitting on this scale was not without historical precedent.”
BLS, 'The Great Resignation' in perspective · Monthly Labor Review, July 2022 (paraphrase)
Four years on, the spike is gone. The quits rate fell to 2.4% across 2023, slid to 2.0% by July 2024, and sat at 2.0%in the latest JOLTS release for March 2026 — at or below the roughly 2.3% pre-pandemic norm. Economists now describe the labor market not as the “Great Resignation” but as the “Great Stay”: low hiring, low firing, and workers staying put.
FRED's Quits: Total Nonfarm Rate (JTSQUR) tracks the share of workers voluntarily leaving jobs each month. After peaking at a series-record 3.0% in late 2021, the rate has settled back to 2.0% — at or below its pre-pandemic level.
The national averages hide enormous variation by state, and New York is one of the clearest laggards. BLS data put New York’s unemployment rate at 4.6%in April 2026 — above the 4.3% national rate — with New York City running near 5.6%. The broader U-6 measure, which counts the underemployed and discouraged, reached 8.0% in New York in 2025. Private-sector jobs in NYC grew just 0.8% across 2025, the fewest first-half hires since 2009 outside of the 2020 shutdown year.
The composition of what growth there was tells its own story. More than 77%of New York’s 2025 job growth came from government-funded health care — a category propped up by Medicaid and public spending rather than private expansion. Meanwhile the people leaving are not the people arriving: per IRS migration data, New York City residents who departed between 2019 and 2023 earned roughly $68 billion more in aggregate income than those who moved in. The top combined NYC-plus-state income-tax rate, 14.776%, is the highest in the nation.
- →Unemployment: 4.6% statewide (April 2026) vs. 4.3% national; NYC ~5.6%
- →U-6 underutilization: 8.0% in New York (2025)
- →NYC private-sector job growth: just 0.8% in 2025 — fewest first-half hires since 2009 (ex-2020)
- →Over 77% of 2025 NY growth came from government-funded health care
- →1M+ residents have left New York since 2021
- →Departing NYC residents (2019–23) earned ~$68B more than arrivals
- →Top combined NYC + state income tax: 14.776% — highest in the nation
But here the honesty rule cuts against the easy narrative. The same benchmark-revision machinery that erased jobs nationally did the opposite for New York in 2025: the revision actually raised the state’s job count by +109,600, according to Bloomberg’s September 2025 analysis. And the tax-flight thesis — that high taxes are driving the wealthy out — is contested; researchers note that cost-of-living, remote work, and retirement also drive migration. The numbers above are real; the single-cause story some draw from them is not settled.
The labor market that produced the 'Great Resignation' has cooled into a 'Great Stay': hiring and quitting are both running below pre-pandemic norms, and growth is increasingly concentrated in a handful of states and sectors.
Every year, the BLS reconciles its survey-based payroll estimates against near-complete unemployment-insurance tax records covering almost every job in the country. This is the annual benchmark revision, and it is the moment the monthly guesswork meets the harder count. For two consecutive years, that reconciliation has produced the largest downward corrections in the modern history of the series.
The 2024 benchmark came in at a preliminary -818,000— the largest downward revision since 2009 — later finalized at -598,000. The 2025 benchmark was worse: a preliminary -911,000, the largest on record going back to 2002. In plain terms, the economy had created hundreds of thousands fewer jobs than the headline reports during those stretches had suggested. Because those stretches fell during the Biden administration, the revisions document a weakness in Biden-era reporting — which is exactly why this should be read as a data-integrity problem rather than a partisan trophy.
Why do the corrections keep getting bigger? Statisticians point to falling survey response rates, a noisier post-pandemic economy, and the difficulty of modeling new and dying businesses in real time. None of those explanations is partisan. All of them point to the same uncomfortable conclusion: the monthly number arrives with a wider error band than the confident headlines imply, and the public is rarely told so.
On August 1, 2025, the data became personal. That morning’s jobs report showed payrolls up just 73,000, with the prior two months — May and June — revised down a combined 258,000. Within hours, President Trump (R) fired BLS Commissioner Erika McEntarferand, on Truth Social, called the figures “RIGGED.”
In my opinion, today's Jobs Numbers were RIGGED in order to make the Republicans, and ME, look bad.
The accusation ran into a structural problem: the BLS Commissioner does not produce the numbers by hand. The payroll and household figures are generated by a documented statistical process from survey data collected by career staff and field economists, and the Commissioner sees the final results shortly before release — with no ability to reach in and change them. That is not a political defense; it is how the agency is built.
The clearest rebuttal came from inside the conservative camp. William Beach, who served as BLS Commissioner during Trump’s own first term, publicly defended the agency, explaining that a commissioner simply cannot rig the numbers because the figures are computer-generated from the underlying survey data.
“A BLS Commissioner can't rig the numbers — the figures are computer-generated from the survey data, and there's no lever to pull to change them.”
William Beach, former BLS Commissioner (Trump first-term appointee) · paraphrase
In September 2025, the Department of Labor’s Inspector General opened a probe of BLS data-collection practices — a legitimate inquiry into why the estimates have grown less reliable, and a different question entirely from whether anyone manipulated them.
Strip away the noise and a few things hold up under scrutiny. The “Great Resignation” was a real, record-setting event — and it is over; the quits rate is back to its pre-pandemic norm. The benchmark revisions are real, the biggest on record, and they document genuine weakness in payroll numbers first reported under the Biden administration. New York’s labor market is genuinely soft relative to the nation, with growth leaning on government-funded health care.
And a few things do nothold up. There is no evidence the BLS “rigged” any report; FactCheck.org, CNN, PolitiFact, and the Peterson Institute all examined the claim and found nothing, and a Trump-appointed former commissioner said plainly that rigging is not mechanically possible. The case that New York’s decline is purely a tax-flight story is contested — and the 2025 benchmark actually raisedNew York’s job count by 109,600. Honest accounting means carrying the facts that cut against the easy frame, not just the ones that confirm it.
- →TRUE: The 'Great Resignation' peaked at a 3.0% quits rate in late 2021 and is over — back to 2.0% now.
- →TRUE: The 2024 (-598K final) and 2025 (-911K preliminary) benchmark revisions are the largest on record — documenting Biden-era reporting weakness.
- →TRUE: New York's labor market trails the nation; 77%+ of 2025 growth was government-funded health care.
- →NOT SUPPORTED: That the BLS 'rigged' a report — multiple fact-checks found no evidence; the figures are computer-generated.
- →CONTESTED: That high taxes alone are driving New York's decline — and the 2025 benchmark RAISED NY's count by 109,600.
The labor market that produced the “Great Resignation” has cooled back to its pre-pandemic shape — the quits rate is at 2.0% and the story is now the “Great Stay.” But the data that measures the economy has a real reliability problem: two record-sized downward benchmark revisions, nearly a million jobs erased in the latest one, correcting numbers first reported under the Biden administration.
That is a case for fixing the measurement, not for declaring it “rigged.” When a weak July 2025 report led the President to fire the BLS Commissioner, multiple fact-checks — and a former commissioner he had appointed — found no evidence of manipulation, because the figures are computer-generated. A statistic the whole country relies on only works if both sides agree to trust it when it tells them something they don’t want to hear.
