“`html
President Donald Trump has removed the head of the U.S. Bureau of Labor Statistics (BLS) following a downward revision of recent job numbers exceeding 250,000.
The President stated that the figures were “rigged” to present his administration in a “bad” light.
While the latest revisions were significant, it is standard procedure for initial monthly figures to be adjusted, a practice observed under both Democratic and Republican administrations.
Former commissioners indicate that the BLS head, officially the commissioner, does not participate in data collection or compilation, only reviewing the final press release prior to publication.
“My reaction was, ‘That couldn’t happen,'” stated Katharine Abraham, BLS commissioner from 1993 to 2001, regarding President Trump’s claims of manipulated figures.
“The commissioner doesn’t have control over what the numbers are,” she added.
The BLS jobs report is based on two surveys: one gathering data from approximately 60,000 households and another from 121,000 public and private sector employers.
Job gain estimates are derived from the employer survey, often referred to as the establishment survey, which is generally considered more reliable due to its larger sample size.
Most responses come from large firms, often enrolled in programs for automatic employment information submission. BLS staff also conduct web surveys and telephone interviews.
“The initial estimates of payroll employment are a preliminary look at what occurred in each month,” the BLS told BBC Verify.
“It is the quick but lower-resolution snapshot of what went on in the job market for a particular month. Because the revised estimates are based on more complete data, they create a higher resolution picture – and occasionally the revised data produce a different picture,” it added.
The bureau updates figures in the two months following the initial monthly estimate as more responses are received. It also recalculates numbers annually to incorporate data from unemployment insurance tax records.
“There are all of these career people who also have the data and if the commissioner were to try to change the numbers they would all know and it would get out,” Prof Abraham says.
The figures for May and June have been revised down by 125,000 and 133,000 respectively.
The combined reduction of 258,000 for the two-month period marks the largest change on record, excluding the months following the onset of the COVID-19 pandemic in 2020.
However, adjustments occur monthly, and substantial changes are not unprecedented.
In this instance, many analysts anticipated revisions to the June figures, which had shown an unusual increase in school employment during a month when most schools were nearing summer closure.
Analysts also note that later responses disproportionately reflect smaller firms, which are more susceptible to economic shifts, such as tariffs.
The May figure was adjusted downward largely in response to the June revision and aligns with other data indicating an economic slowdown.
According to BLS data going back to 1979, the average monthly change to the jobs figures (either up or down) is 57,000, according to the BLS.
However, revisions tend to be more significant during periods of economic instability.
Beyond the most recent numbers and the 2020 COVID-19 period, there have been eight other instances since 2000 when the BLS revised monthly job numbers downward by more than 100,000, with most occurring around the 2008 financial crisis.
For example, the January 2009 figure, during President Barack Obama’s administration, saw a reduction of 143,000.
The BLS also reported that job gains for the entire year in 2009 were 902,000 lower than initially estimated – the largest full-year revision on record.
The jobs created in 2024 under President Joe Biden were revised down by 598,000, though that was a smaller change than the more than 800,000 initially estimated – an update which also caused political fallout.
Prof. Abraham states that updates are integral to the process. She was not surprised by the large revisions for May and June, given the increasing difficulty of collecting responses, the lack of investment in new methods, and the broader economic slowdown, partly driven by new tariffs.
“It’s always difficult when you’re at a point where things may be changing and then you add to that the fact that staffing has been constrained and the agencies haven’t had the resources to invest in following up with respondents the way they might have in the past,” she says.
Response rates have significantly decreased over the last decade, accelerating after the pandemic, raising concerns about the reliability of the data.
For example, the response rate for the establishment survey was less than 43% in March, compared with more than 60% a decade earlier.
Other countries, including Canada, Sweden, and the UK, have faced similar declines. Response rates to the labor force survey in the UK have fallen to approximately 20%.
In the U.S., the decline has spurred efforts to explore new data collection methods, including web-based surveys.
However, the significance of the issue remains debated.
A review by researchers at the Federal Reserve Bank of San Francisco in March of this year found that revisions in recent years were mostly in line with pre-pandemic patterns, which, it said, should be reassuring to those worried about reliability.
What do you want BBC Verify to investigate?
The Democrats facing arrest had fled to block a vote that could sway the balance of power in Washington.
The US president says India doesn’t care “how many people in Ukraine are being killed” – but India calls the tariff threat “unjustified”.
Several people were treated for smoke inhalation.
The reptile was retrieved by the officer and buckled into the back of a police car.
The comments come as Trump’s deadline for Russia to agree a ceasefire with Ukraine or face sanctions edges closer.
“`