US Job Growth Revised Lower By 79,000 In Annual Benchmark Estimate

In our preview of today’s preliminary benchmark revision of US jobs – published by the BLS ‘conveniently’ just as Kevin Warsh started to speak – we said that according to Goldman calculations, for the first time in 3 years and just the second time since 2018, the BLS was going to revisedpayrolls modestly higher “based on the nine months of data released since the last benchmarked period, March 2025.”

Specifically, Goldman’s economists expected “a preliminary upward revision on the order of 50-450k which would translate to a 5-40k upward revision to monthly payroll growth over April 2025-March 2026. A final revision of this magnitude would result in the average pace of payroll growth over April 2025-March 2026 being revised up from about 25k/month currently to 30-65k/month.”

Alas, for one more year, it was not meant to be, and this morning the BLS announced that according to the preliminary estimate of the Current Employment Statistics (CES), the 2026 benchmark revision to total nonfarm employment for March 2026 was -79,000.  While just why of a positive revision, it was a far cry from last year’s record 911K negative job revision. For context, annual benchmark revisions over the last 10 years have had absolute average of 0.2% of total nonfarm employment.

Additionally, the revision for total private employment was -178,000, which means that government jobs were revised higher by 99K.

The 178,000 negative revision for private payrolls in the year through March reflected weakness in retail trade, education and health services, manufacturing and business services. Employment increased in transportation and warehousing, information, financial activities and construction. 

It is likely that the final final revision will actually tip into the positive. In accordance with usual practice, the final benchmark revision will be issued in February 2027 with the publication of the January 2027 Employment Situation news release. As we noted earlier, preliminary estimates for the benchmark revision tend to understate the final revision: the nextx chart shows that the preliminary estimate has been below the final revision in each of the last six years, by roughly 100k on average, which suggests that today’s -79K print will end up being in the +20K ballpark. This reflects that the QCEW itself has been revised up in every quarter since 2019 with the exception of 2020 H1, potentially reflecting ongoing issues with initial submissions to the administrative records that inform the QCEW

Before today’s revision, government payrolls data indicated employers added 211,000 jobs in the year through March on a non-seasonally adjusted basis, or an average of 17,600 per month, according to data compiled by Bloomberg. The preliminary benchmark revision suggests average job growth was likely closer to 11,000 a month.

Preliminary benchmark revisions have now lowered employment estimates in seven of the past eight years. Even so, the latest adjustment suggests that the labor market is roughly balanced – with employers slow to hire new workers but also slow to fire existing staff.

The BLS each year benchmarks the March payrolls level to a more accurate but less timely data source called the Quarterly Census of Employment and Wages that’s based on state unemployment insurance tax records and covers nearly all US jobs. While the new information improves the accuracy of its data, the process has gained additional attention in recent years.

Last year’s preliminary adjustment slashed employment estimates by the most on record, reigniting White House criticism of the BLS. About one month prior to the 2025 preliminary benchmark release, President Donald Trump fired the agency’s leader after a separate monthly report showed weak job growth. The Senate confirmed Trump’s pick to lead the BLS – Brett Matsumoto – on Aug. 7. Matsumoto, a PhD economist and BLS veteran, now helms an agency responsible for publishing some of the most market-moving statistics in the world.

What is behind the chronic negative revisions? First, there is the chronically wrong birth-death model, discussed extensively here in recent years. Yet just 14% of last year’s very large revision can be attributed to miscalibration of the birth-death model; the bulk instead falls into the residual category which would capture the reporting error arising from a systematic undercount of unauthorized workers.

As we discussed first a few years ago when we correctly previewed the massive negative revisions to 2023 and 2024 data, since the QCEW is based on unemployment insurance records, it likely excludes most unauthorized workers, who contributed to employment growth in the periods covered by those benchmark revisions. In most cases unauthorized workers do not qualify for unemployment insurance, so employers might see little reason to pay unemployment insurance tax on their behalf and might even see it as a needless risk in the cases of any immigrants they are employing who do not yet have work permits.

As such, the Trump admin’s aggressive purging of illegal aliens – and workers – has led to significant real-time overestimates of the labor market in the monthly series, which are then revised away every year once it becomes clear that there were far fewer illegal aliens in the workforce.