Employment growth in the United States during the year ending in March may be far weaker than currently reflected in official data, highlighting that the labor market had already entered slow growth well before this summer's hiring slowdown.

Economists at Wells Fargo, Comerica Bank, and Pantheon Macroeconomics expect that preliminary benchmark revision data to be released by the Bureau of Labor Statistics on Tuesday will show that non-farm payroll growth for the year ending in March was likely weaker by nearly 800,000 jobs than current estimates, equivalent to an average monthly shortfall of about 67,000 jobs. Nomura Securities, Bank of America, and Royal Bank of Canada indicate that the downward revision could even approach one million jobs.

While the data being revised covers a period that has already passed, such a substantial revision would demonstrate that last year's labor market momentum was far weaker than previously believed, and would reinforce expectations for a series of Federal Reserve interest rate cuts. A second consecutive year of major employment data revisions could also anger President Donald Trump, who has already criticized the accuracy of Bureau of Labor Statistics data.

The Bureau of Labor Statistics conducts annual preliminary revisions of non-farm payroll numbers through March by comparing them against a more accurate but less timely data source - the Quarterly Census of Employment and Wages (QCEW) - which is based on state unemployment insurance tax records and covers nearly all jobs in the United States. This represents a further adjustment beyond the monthly non-farm data revisions to make the data more accurate.

"Compared to downward revisions to recent months' job growth, a major downward revision to employment growth through March 2025 would have less impact on monetary policy, but it provides context for overall economic performance," said Bill Adams, chief economist at Comerica. "All else being equal, downward revisions to job growth increase pressure on the Federal Reserve to ease policy."

This news would also bolster those who believe the Federal Reserve should have begun easing policy several months ago. Federal Reserve Governor Christopher Waller said he expects the benchmark revision to reduce job growth by an average of about 60,000 per month. While the Fed decided to keep rates unchanged at its last meeting in July, Waller voted in favor of a rate cut. The market widely expects the Federal Reserve to cut rates at next week's meeting.

**Political Implications**

Although the data revision won't change current views of the labor market, it will suggest that the recent slowdown in hiring actually began much earlier. The Trump administration could use Tuesday's data to claim that the deceleration in job growth began long before he took office. Final data will be released early next year.

"This essentially reflects job creation before Trump's term. So he can really argue that the data suggests the economy he inherited was actually much weaker than all of us previously thought," said Samuel Tombs, chief US economist at Pantheon Macroeconomics.

About a month ago, unusually large downward revisions to monthly employment data drew White House displeasure and prompted Trump to fire the Bureau of Labor Statistics commissioner. Trump's criticism targeted not only monthly revision figures but also last year's preliminary benchmark revisions, which represented the largest downward adjustment since 2009.

Despite Trump's criticism of data revisions, monthly and benchmark adjustments are standard practices for continuously updating estimates as the statistical bureau obtains more data. The relatively larger revision magnitudes in recent years are partly due to declining survey response rates.

**Business "Birth-Death" Model**

In recent years, monthly employment data has shown stronger job growth than QCEW data most of the time. Some economists believe this is partly due to the so-called "birth-death" model - which the Bureau of Labor Statistics uses for data adjustments to account for the net number of new and closing businesses. Calculations have become more difficult since the pandemic.

Others believe the data mismatch has different causes: immigration. The monthly non-farm report doesn't inquire about citizenship status, but the QCEW report uses unemployment insurance records - which undocumented immigrants cannot apply for.

Ultimately, economists and policymakers will use the preliminary benchmark data to assess the extent of labor market deceleration and await more complete 2025 data when the government makes final revisions in February next year.

"If last year's data is significantly revised, that helps me understand where the starting point is," said Carrie Freestone, economist at Royal Bank of Canada. "But I think what Fed officials will be most concerned about is the loss of momentum - that the labor market may have reached a turning point."