By Christopher Lugerber
WASHINGTON (AP) — Monthly employment reports are already closely monitored on Wall Street and Washington, but are now new and important after President Donald Trump fired the officials who oversee it on Friday.
Trump claimed that June’s employment numbers were “equipped” to “look bad” him and other Republicans. However, he has not provided evidence, and William Beach, whom official Trump appointed in his first term to oversee the report, has denounced the firing of Erica Mantelfer, the director of the Bureau of Labor Statistics, appointed by former President Joe Biden. The shooting followed a job report on Friday that showed jobs were weak in July and nearly stagnated shortly after Trump rolled out tariffs that were wiped out in May and June.
Economists and Wall Street investors have long thought work numbers were reliable, as they often react sharply when stock prices and bond yields are released. Still, Friday’s revision was extraordinarily large – the biggest outside the recession in 50 years. Additionally, the surveys used to compile the report have faced challenges, particularly since Covid, due to declining response rates. This is because few companies complete the survey.
Nevertheless, it has not led most economists to doubt them.
“For me, the bottom line means that you don’t take a low collection rate as evidence that the numbers are not reliable,” said Omair Sharif, founder and chief economist at Inflation Insights, a consulting firm.
Many scholars, statisticians and economists have warned for some time that budget declines have put a strain on the government’s ability to collect economic data. There have been several government committees studying ways to improve survey response rates and more, but the Trump administration disbanded them earlier this year.
Heather Boohee, the top economic adviser at the Biden White House, noted that if Trump did not fire Mantelfer, he would focus more on last week’s data, pointing to a slower economy.
“We’re having this conversation about the configured issues to divert us from what our data is being shown,” Boushey said. “A negative directional correction of this magnitude could indicate something bad is happening for the labour market.”
Here’s what you need to know about job reports:
Economists and Wall Street trust data
Most economists say the Bureau of Labor Statistics is a non-political institution with people who are obsessed with getting numbers right. The only political appointee is a commissioner who will not see the data until it is confirmed two days before it is issued to the public.
BLS Commissioner Erica Grossen from 2013 to 2017 said in the report that he proposed a different language for “Liven It Up” but was shot down. She was told that if asked to describe the cup half empty or half, the BLS would say “it’s an 8-ounce cup with 4 ounces of liquid.”
The revised job data that attracted Trump’s rage actually matched other numbers than before the revision. For example, Payroll Processor ADP calculated its own Jobs reports using data from millions of clients, showing a sharp slowdown in adoption, close to BLS data revised in May and June.
Trump and his White House have a long track record of celebrating the number of jobs when they’re good.
These are the numbers Trump is attacking.
Trump focused on revisions to data for May and June, revised low on Friday, with employment growth in May falling from 144,000 to 19,000, and from 147,000 to just 14,000 in June. Monthly employment data will be revised over the next two months.
Trump also repeatedly attacked largely inaccurately from his campaign on the annual revision last August, reducing total U.S. employment by 818,000, or about 0.5%. The government also adjusts the number of jobs each year.
Trump announced an annual revision before the 2024 presidential election, “boosts” Vice President Kamala Harris’ “potential of victory,” but widely reported that two months before the election, the revision noted that it had reduced employment and weakened the economy during the Biden-Harris administration.
Here’s why the government corrects its data:
Monthly revisions occur as many companies responding to government investigations send data recently or correct numbers already submitted. The proportion of companies that send data later has risen over the past decade.
Each year, BLS makes additional revisions based on actual employment numbers derived from state unemployment insurance records. These figures cover 95% of US companies and are not derived from the survey, but are not available in real time.
These are the factors that lead to revisions
Understanding how many new jobs are added or lost each month is more complicated than you can hear. For example, if one person does the second job, should they focus on the number of jobs that have increased, or the number of people employed? (The government measures both. Unemployment rates are based on the number of people who have or do not employ them, but the number of jobs added or lost is counted individually).
Each month, the government surveys around 121,000 businesses and government agencies in more than 630,000 locations, including multiple locations in the same business, covering about a third of all workers.
Still, the government needs to provide estimates as well. What happens if the company goes out of business? You may not fill out a form showing lost work. And what about new business? They can take time to board the government radar.
BLS seeks to capture these trends by estimating the impact on employment. Of course, these estimates may be incorrect until revised by the annual revision.
In many cases, revisions grow around economic turning points. For example, if the economy is growing, the revision will be higher as there may be more startups than the government expects. If the economy is slowing or is in a recession, the revision could be on the downside.
Here’s why the May and June revisions were so large
Ernie Tedeski, an economic adviser to the Biden administration, points to the current dynamics of the labor market. Both employment and firing have declined sharply, and fewer Americans have quit their jobs to do other jobs. As a result, the majority of monthly job losses can probably occur in new or out-of-business companies.
And these are things government estimates using models, which can be made more volatile.
Grossen also points out that since the pandemic, there has been a surge in new startups after many Americans lost their jobs and sought more independence. But they may not have created as much work as the startups did before Covid.
The revision seems to be getting bigger
According to economists at Goldman Sachs, the revisions to the total employment in May and June, which reduced employment by a total of 258,000 since 1967, reduced employment by a total of 258,000.
Trump’s top economic adviser Kevin Hassett went on to NBC’s Meet the Press on Sunday, saying, “What we’ve seen in recent years is a massive revision of employment.”
Hassett condemned a sharp decline in response rates to government surveys during and after the pandemic.
However, calculations by Tedeschi show that while revisions skyrocketed after the pandemic, they are much smaller than in the 1960s and 1970s.
Other concerns about government data
Many economists and statisticians have been alarming for years about lower response rates and more. Ten years ago, about 60% of the companies surveyed by BLS responded. Right now, only about 40% of them do it.
This decline has been an international phenomenon, especially since Covid. The UK has suspended its official unemployment disclosure due to a decline in response.
And earlier this year, BLS said it has been cutting collection of inflation data due to the Trump administration’s employment freeze, raising concerns about the robustness of price data, just as economists are trying to measure the impact of tariffs on inflation.
According to a July report by the American Statistics Association, US government statistics agencies have seen a 16% decline in inflation adjustments since 2009 and since 2009.
“We’re at the inflection point,” the report said. “To meet current and future challenges, we need thoughtful and well-planned investments. In contrast, what we observed is collaborative, unplanned reductions without visible plans for the future.
Original issue: August 4, 2025, 7:34pm EDT
