Last Friday (August 1, 2025), the US Bureau of Labor Statistics published their latest – Employment Situation Summary – for July 2025. What followed was somewhat extraordinary. The data and the revisions to the previous two months data releases (which is standard practice) showed that the US labour market is in decline. It starkly runs counter to the official Trump narrative that the US is booming. While we don’t have enough evidence to really establish causality – it is likely (based on theoretical conjecture) that the highly volatile policy regime that Trump runs and his tariff flip flops is undermining the confidence in the economy. We need a few more months of data yet before we can be sure. But the BLS results certainly support the view that Trump’s economic policies are not working to advantage the American public. The extraordinary thing was that Trump then sacked the BLS head and signalled a further descent towards totalitarianism.
Payroll employment trends
The BLS noted that:
Total nonfarm payroll employment changed little in July (+73,000) and has shown little change since April … The unemployment rate, at 4.2 percent, also changed little in July. Employment continued to trend up in health care and in social assistance. Federal government continued to lose jobs.
There were some disturbing signs including:
1. “Among the unemployed, the number of new entrants increased by 275,000 in July to 985,000. New entrants are unemployed people who are looking for their first job.”
2. “In July, the number of long-term unemployed (those jobless for 27 weeks or more) increased by 179,000 to 1.8 million. The long-term unemployed accounted for 24.9 percent of all unemployed people.”
These trends signify a slowing economy.
The first graph shows the monthly change in payroll employment (in thousands, expressed as a 3-month moving average to take out the monthly noise). The red lines are the annual averages. Observations between March 2020 and March 2022 were excluded as outliers.
The trends since 2022 has been consistently downwards.
The next graph shows the same data in a different way – in this case the graph shows the average net monthly change in payroll employment (actual) for the calendar years from 2005 to 2023.
The red marker on the column is the current month’s result.
Average monthly change – 2019-2025 (000s)
| Year | Average Monthly Employment Change (000s) |
| 2019 | 166 |
| 2020 | -771 |
| 2021 | 603 |
| 2022 | 380 |
| 2023 | 216 |
| 2024 | 168 |
| 2025 (so far) | 85 |
The data is clearly recording a consistent downward trend.
The controversy
National Statistical Offices (NSOs) have faced major difficulties since the Covid pandemic began.
First, when the outbreak first became obvious, the NSOs were forced to abandon face-to-face interviews and office staff were required to work from home.
Many staff reported across many nations that they did not have the technological capacity at their homes of a quality that they used in their daily work in the main office.
The World Bank reported on August 11, 2021 – National statistical offices still face disruptions and challenges as they adapt to a “new normal” – that they were, in partnership with the United Nations launching a “global online survey … to monitor the effects of the pandemic on national statistical operations around the world”
In their final report – One Year Into the Pandemic : Monitoring the State of Statistical Operations Under COVID-19 – we learned of the “large disparities in the ability of national statistical offices in different countries to continue operating”.
Many NSOs had their government funding reduced, as a consequence of the usual fiscal fictions that my professions pushes out.
Importantly, “Reaching migrants, internally displaced populations and persons with disabilities during the pandemic is a challenge for may statistical offices, including in high-income countries”.
These problems were heightened in the low- and lower-middle-income countries, as one would expect.
Have these problems persisted?
There is evidence that it is now harder for NSOs to run surveys.
A recent article in the Japan Times (June 17, 2025) – Faced with dwindling response rates, pollsters turn to online surveys for data – suggests that the traditional face-to-face interviewing techniques deployed by NSOs continues to be a challenge even 5 years into the pandemic.
Reliance on telephone interviews has proved to be problematic.
For example:
Only a quarter of Britons say they answer such calls, with most wary of scams. Fewer than half of U.K. households have a landline, and those numbers can be difficult to find.
As a result, in the UK, for example, “Response rates have plunged” and:
By 2023, the data had become so patchy, the ONS was forced to suspend its unemployment reading as it desperately sought to repair a key release that helps inform government policy, influences interest rates and drives billions of dollars of investment decisions.
That controversy led to the resignation of the chief statistician in the UK.
Multiple problems have arisen that make it hard for NSOs to conduct their data gathering activies:
1. People wont answer their mobile phones if a strange number rings because of the rising incidence of scammers.
2. The rise of video surveillance on front doors.
3. The increase in secure apartments that make it impossible for surveyors to get to the front door.
NSOs have attempted to overcome these obstacles by shifting to online survey techniques.
This shift has introduced new problems:
1. “biases in the data, confusion over questions that a computer cannot address”.
2. “gaining trust amid a barrage of spam and fraud in people’s inboxes”.
3. “households may be just as likely to ignore an email with an online survey as a phone call”.
4. There is a tendency for online respondents to “give extreme answers on the web” whereas they “might think twice about giving any sort of extreme value to an interviewer”.
There is no simple answer.
In the US, the same problems have arisen:
… the rate for the Current Population Survey underpinning labor statistics has fallen by almost 20 percentage points in a decade to 68%. For the Job Openings and Labor Turnover Survey or JOLTS report, a business survey, response rates have tumbled to almost 30%.
The former boss of the US Bureau of Labor Statistics has said that the traditional Labour Force survey from which employment and unemployment data is gleaned “are dying, they’re decaying, they’re in very serious trouble”.
If the sample sizes fall below some level, then the data becomes useless from a statistical perspective.
Noting all the above, which is an ongoing problem, does not in any way suggest that the NSOs fabricate data.
The US has other problems that are Trump-made.
1. The “crackdown on immigration may be discouraging foreign-born workers from participating in the survey”.
2. Trump’s “job cuts across the federal workforce put greater pressure on funding and staffing.”
There has been a “catastrophic exodus of thousands of employees from the US Department of Labor” as DOGE went about slashing departmental budgets and job slots (Source).
I know several people personally in the Department who have lost their jobs since February 2025.
The Department has lost about “20% of its workforce” already and that is before the foreshadowed ‘reorganization’ has been implemented.
So, amidst all the issues I have summarised above, that have made the work of NSOs much more challenging, the US agency responsible for labour data collection and dissemination is facing further major job cuts that will, according to insders, “cause it to be absolutely dysfunctional”.
None of this is ‘accidental’.
This article from October 19, 1995 from the US Heritage Foundation – How to Close Down the Department of Labor – tells us that it has been on the Right’s agenda for decades to get rid of the Department.
Don’t forget that the Heritage Foundation is the architect of Trump’s Project 2025.
Right-wing politicians and their boosters have long been suscpicious of NSOs.
On August 15, 2022, the Australian Bureau of Statistics released a new report – COVID-19 Mortality in Australia, Deaths registered to 31 January 2022 – which provide the public with verifiable data and analysis about how severe the pandemic was at the time, particularly in terms of the way it was disproportionately impacting low-income families.
Right-wing commentators went ballistic and claimed the data was false – denying that Covid killed people – and asserted that the data was being used by the government to justify the lockdowns.
The concept of – Epistemic democracy – has long troubled the Right.
The idea is that our so-called ‘democratic institutions’ have to be able to “communicate, produce, and utilise knowledge, engage in forms of experimentation, aggregate judgements and solve social problems.”
This is one of the justifications for adequately funding NSOs.
If the quality of the public information is reduced or compromised then the capacity of society to make good decisions is reduced.
That has been the way that the Right has sought to undermine governments that promote generalised well-being.
By promoting massive disinformation campaigns they can undermine what John Rawls called “imperfect procedural justice” and tilt the decision-making institutions towards advancing the special interests of the Right, which usually coincide with the interests of Capital (the wealthy).
There are so many areas of public debate where misinformation dominates the information that is available – climate change, health care, etc.
In some cases, the officials stop publishing data – in Australia, for example, the conservative government sought to suppress labour market data because it was so bad.
Altering surveys to eliminate questions that produce ‘risky’ results for politicians is common.
But what happened on Friday is rather extraordinary.
When the BLS released the latest labour force and payroll data, it told us that:
Revisions for May and June were larger than normal. The change in total nonfarm payroll employment for May was revised down by 125,000, from +144,000 to +19,000, and the change for June was revised down by 133,000, from +147,000 to +14,000. With these revisions, employment in May and June combined is 258,000 lower than previously reported. (Monthly revisions result from additional reports received from businesses and government agencies since the last published estimates and from the recalculation of seasonal factors.)
The revisions are common in survey work like this because the NSO has to honour a regular temporal publishing calendar and for various reasons falls short of getting all the responses back in time.
Their solution – publish the partial results and then revise them when the full data is available.
And the NSOs always signal that these partial results are ‘preliminary’.
There are problems with this approach – it can mislead when the revisions are subsequently large – but the alternative is to do what the UK ONS office did a while ago when they suspended publication because of low response rates.
As noted above that heightened suspicions among the Right.
Anyway, regular researchers who use NSO data regularly understand the process of revision and are not alarmed by it.
But the US President clearly couldn’t square the latest BLS results with his ‘big booming economy’ narrative and instead of admitting the crazy tariffs and other policies that come and go on a daily basis were not helping, he decided to deny the data and sack the boss of the BLS.
I recall when I was an undergraduate student who regularly challenged the lecturers, I had an interchange about the veracity of one of the teacher’s claims about the labour market.
I quoted the actual data that had just been released and the response I got was that when the facts the theory conflict it is the facts that are wrong.
Holding those sorts of views are essential requirements for the patterned behaviour of denial and social psychologists refer to as – Groupthink.
Trump certainly didn’t hold back claiming that:
In my opinion, today’s Jobs Numbers were RIGGED in order to make the Republicans, and ME, look bad.
His lackey Labor Secretary chimed in with her support.
There is an interesting group in the US – The Friends of the Bureau of Labor Statistics – who regularly defend the agency and realise that it is “essential component of our national data infrastructure”.
On August 1, 2025, they released a – Statement on Commissioner McEntarfer’s Removal – and noted that the President had engaged in a “baseless, damaging claim” about the BLS purely “to blame someone for unwelcome economic news”.
Further:
… The process of obtaining the numbers is decentralized by design to avoid opportunities for interference. The BLS uses the same proven, transparent, reliable process to produce estimates every month. Every month, BLS revises the prior two months’ employment estimates to reflect slower-arriving, more-accurate information.
Descent into totalitarianism
The US is now heading from authoritarianism towards totalitarianism.
We have seen the way the science community has been constantly attacked by the Right.
Authoritarian regimes usually always seek to discredit the science community.
Facts are twisted to suit the ideological ends.
The long-standing campaigns by tobacco companies to deny the link between smoking and cancer is a good example.
Trump has demonstrated his willingness here.
Remember – Sharpiegate – when Trump’s team falsified the path of the official National Oceanic and Atmospheric Administration hurricane forecast for Dorian and then pressured the NOAA staff to retract their forecast.
Trump has also appointed a number of people into senior scientific positions that have no background at all.
Then withdrawn research funding and scrapping scientific advisory committees that are seen as an inconvenience (Source).
Now Trump sacks the BLS head to place doubt in peoples’ minds about the quality of the data.
Since the beginning of this year, the availability of US federal data sets has declined.
I have found that out from personal experience, particularly in datasets in the health area that I have used for years – now disappeared from access.
The Federation of American Scientists noted that “targeted, surgical removal of data sets, or elements of data sets, that are not aligned with the administration’s priorities” (Source).
Much of the US data covering the ‘diversity’ terrain has gone.
All of these trends are consistent with a shift towards totalitarianism.
That is how such a shift begins.
Conclusion
The ‘land of the free’ is now a failed state and it is scary that they are still the most martial.
Having written this and other things recently, I don’t expect I will ever get a visa to enter the US while the current regime is in power.
That is enough for today!
(c) Copyright 2025 William Mitchell. All Rights Reserved.
