That was the allegation made by President Donald Trump when he recently fired the head of the U.S. Bureau of Labor Statistics (BLS). This follows a historically large downward revision of recent job numbers, deepening worries that the employment data released during his presidency have been unreliable. The BLS revised down the job counts for May and June by more than 250,000 combined. This revision was the primary catalyst for alarm and backlash from the White House.
As we noted HERE, the BLS is constantly revising its employment numbers. It does this within the two month period after the initial estimates, as it chases down additional responses from employers. This ongoing process is meant to give a clearer and fuller view of the state of jobs and employment. “The initial estimates of payroll employment are a preliminary look at what occurred in each month,” a spokesperson from the BLS stated to BBC Verify.
As an aside, the downward revision accounted for much of the adjustment to May’s numbers. This unfortunate change was steered almost entirely by the amendments adopted in June. This re-benchmarking is consistent with other economic indicators that point to a slowdown in hiring and job gains in almost all private sector industries.
Katharine Abraham, who served as BLS commissioner from 1993-2001. She wanted to convey that revisions are an everyday, even progressive nature of the data collection process. She said she was not shocked by the large downward revisions for May and June. She laid the blame for these decisions on the increasing difficulty of getting truthful answers from employers.
“The commissioner doesn’t have control over what the numbers are,” Abraham remarked, highlighting the independence of the BLS in its statistical processes. She emphasized the importance of career professionals within the agency and protecting that, ensuring the integrity of the data. They’d be the first ones to catch any tomfoolery with the numbers.
The establishment survey conducted by the BLS covers approximately one-third of all non-farm payroll jobs and is regarded as more reliable than the household survey due to its larger sample size. Recent figures indicate that the establishment survey’s response rate has fallen below 43% as of March. This comes as a drop of more than 60% just 10 years ago. This decrease in response rates is leading many to worry about the quality of data that has been gathered throughout such a chaotic economic time.
In a normal year, the BLS tends to revise down monthly jobs numbers more than 100,000. This has happened in downward revisions on eight other occasions since 2000. Today’s combined reduction of 258,000 for the two-month period is by far the largest such downward revision in these records’ history. The sole exceptions are the changes to the standards adopted in 2020 during the early weeks of the COVID-19 pandemic. History shows that revisions sharply rise during times of economic distress, like the current economic climate.
Per BLS data going back to 1979, the average monthly revision of the job numbers is approximately – wait for it – 57,000. This change can be positive or negative. A deeper look into these recent revisions shows a disturbing pattern. This may be indicative of more systemic problems in the labor market as recession looms.
Those first estimates provide an early look at what’s happening in the job market for any given month. They provide a much lower-resolution snapshot, cautioned the BLS. Since the revised estimates are more informed due to the availability of more complete data, they provide a higher resolution picture and sometimes the revised data tell a different story.
Yet, despite such laudatory initiatives, all is not well. Is it even possible for the BLS to be politically neutral and independent while ensuring the most accurate and timely labor statistics are delivered? The agency’s recent ousting of its top leader has only sharpened focus on the controversial employment data that’s mirrored Trump’s presidency.