If you have been watching the news lately, you may have noticed that the job figures keep getting revised. One month the headlines say the labor market is booming, and then later we find out the gains were smaller than expected. It can feel confusing, but these revisions are actually a normal part of how the Bureau of Labor Statistics (BLS) tracks employment.
The BLS collects payroll data from about 120,000 employers every month. But not every business replies right away. That means the first report is based on partial information. As more businesses report, the numbers are revised two more times. And then, once a year, the BLS does a big adjustment called “benchmarking” to match the survey data with unemployment insurance tax filings, which are more accurate.
As former BLS commissioner Erica Groshen explained, “It is not a bug, it is a feature; it makes the data more accurate.”
Recommended:
Bad news for the US economy: job figures slashed
What is driving the big downward revisions
The recent revisions have been much larger than usual, and that has put a spotlight on the process. Economists say there are a few main reasons why this is happening:
- New businesses are not creating as many jobs as expected. The BLS uses a model to estimate job creation at new companies, but since the pandemic, many of these businesses are smaller and more short-lived. That means the model may be overstating job growth.
- Fewer companies are responding to surveys. The BLS has seen a big drop in response rates, down to about 43% in the past year from nearly 60% in 2019. That makes it harder to get a clear picture.
- Undocumented workers are not counted in benchmark data. Since unemployment insurance filings do not include undocumented workers, some of the jobs reported early on may “disappear” when the numbers are benchmarked.
Ron Hetrick, a labor economist, explained it this way: “It is probably going to get more accurate, because now we know these people probably are off of payrolls, so we probably know these numbers are a little too high.”
Recommended:
More change at the IRS: Trump removes Billy Long after just two months in the job
How big could the revisions be
This year, the benchmark revision could show that there were anywhere between 475,000 and 900,000 fewer jobs added between April 2024 and March 2025. To put that into perspective, that is like wiping out 50,000 to 75,000 jobs a month from the official totals.
Last year, a similar revision lowered the job count by nearly 600,000. That was the largest downward change since the financial crisis in 2009. These kinds of big swings usually happen during major economic transitions, like the Great Recession or the aftermath of the pandemic.
What this could mean for the economy
For you, as someone watching the economy, the key question is whether these revisions change the big picture. On one hand, even after last year’s adjustment, the job market was still strong. On the other hand, repeated large downward changes raise doubts about how reliable the first reports really are.
Economist Stephanie Roth explained, “This just adds to the narrative of concerns around the labor market and questions about the reliability of the data.”
The truth is that revisions do not mean the BLS is “getting it wrong.” They mean the first reports are early estimates, and the full picture takes time to come into focus. But in today’s political environment, where data itself has become a target, the revisions can fuel more arguments and uncertainty.
Related article:
Bad news for job security: Salesforce slashes 4,000 workers as AI handles support cases
What are the best jobs to do in retirement to make some extra money?
