As the January Non-Farm Payroll (NFP) report approaches, Barclays and Citigroup exhibit significant divergence in their forecasts for the month's job additions, yet they share a high degree of consensus on a more critical issue—US employment data for 2025 is systematically overestimated. An upcoming annual benchmark revision is poised to expose this discrepancy in one fell swoop, a risk that the market has significantly underpriced.
Barclays presents the most direct assessment, suggesting that based on the latest Quarterly Census of Employment and Wages (QCEW) data, the level of non-farm payrolls for March 2025 could be revised downward by approximately 1 million. This implies that the average monthly job growth over the past year has been overstated by 80,000 to 90,000.
Citigroup, meanwhile, issues a warning from a data "quality" perspective. Even if the January NFP appears strong on the surface, this strength could likely stem from residual seasonal factors and model biases rather than indicating genuine labor market stabilization. Signals that truly reflect a downward trend are more likely to become concentrated and apparent during the spring and summer months.
In other words, while the market continues to speculate on whether the January NFP will be "good or bad," the factor with the real potential to reshape the employment narrative is a systemic reassessment of the 2025 employment level.
On the surface, the two investment banks offer contrasting answers to the question of whether January's NFP will be strong or weak.
Barclays forecasts a meager increase of just 50,000 jobs for January, with the private sector adding a similar number—around 50,000—and government employment remaining essentially flat. The unemployment rate is expected to edge down slightly to 4.3%, influenced by rounding and shifts in the labor supply-demand gap.
Citigroup, in contrast, anticipates a much stronger gain of 135,000 jobs, with the private sector contributing roughly 140,000 new positions—a figure significantly above the market consensus of approximately 70,000. They project the unemployment rate will hold steady at 4.4%.
It is noteworthy, however, that both banks explicitly state that the January employment data itself carries limited informational value. Barclays emphasizes that fourth-quarter 2025 data was significantly distorted by "delayed resignation plans" in the government sector, making short-term volatility an unreliable indicator of the underlying trend. Citigroup points out that January is historically one of the most favorably seasonally adjusted months of the year, often exhibiting a pattern of "deceptively strong readings at the start of the year, followed by a subsequent pullback."
Thus, while their numerical forecasts differ, this divergence does not undermine their shared assessment of the broader trend.
Despite their differing views on January, Barclays and Citigroup are in complete agreement on one fundamental issue.
The core problem with US employment lies not in "how many jobs were added this month," but in "how many jobs have been cumulatively overcounted over the past year."
Barclays's central evidence comes from the QCEW—a "near-complete sample" employment dataset based on mandatory business tax filings, widely regarded as the statistical measure closest to the true level of employment. Their key finding is stark: from March 2024 to March 2025, the establishment survey (CES) indicated a cumulative job increase of approximately 1.8 million, whereas the QCEW measure showed an increase of only about 800,000—a discrepancy approaching 1 million jobs.
This implies that the official NFP data has systematically overestimated job growth over the past year, with an average monthly overstatement of around 80,000 to 90,000 positions.
More critically, the magnitude of this deviation is significantly larger than the 2024 benchmark revision (-598,000) and the 2023 revision (-187,000), indicating that post-pandemic structural changes in business birth-and-death dynamics are causing persistent disturbances to the statistical models.
Citigroup does not deny that January's job numbers might appear strong, but its reasoning is inherently counter-intuitive. Citigroup highlights three specific sets of warning signals.
Hiring has not recovered: the JOLTS hiring rate continues to decline to around 3.2%; cyclicality-sensitive sectors such as temporary help, retail, and transportation are still contracting. Signals from the consumer side are deteriorating: The Conference Board's measure of respondents finding jobs "hard to get" increased counter-seasonally in January. Employment data for the latter part of 2025 already shows clear signs of distortion: for instance, private sector job growth for October 2025 was a mere 1,000, but subsequent months' figures were significantly inflated by seasonal adjustment.
Citigroup's conclusion is unequivocal.
If the January data is strong, there are multiple plausible explanations; however, if it is weak, it almost certainly points to underlying fundamental problems. The true pressures within the job market are more likely to become fully apparent in the spring and summer, once the advantage of favorable seasonal adjustments diminishes.
The primary reason lies in a misaligned pricing framework. The current market focuses intensely on whether the monthly NFP figure beats or misses expectations and whether the unemployment rate holds below 4.5%. It largely overlooks three critical implications.
A revision to the employment *level* would alter the historical trajectory: it's not a matter of "growth being slower," but rather "there wasn't as much growth over the past year" to begin with. The interpretive framework for wages, consumption, and GDP would need to be rewritten. The Federal Reserve's perceived "employment safety cushion" might prove to be non-existent.
Once a downward revision of 700,000 to 1 million jobs for 2025 is confirmed.
The current unemployment rate would need to be re-evaluated in terms of the actual strength of labor demand it represents. The labor market's proximity to a potential inflection point would be much closer than the market currently anticipates.
Synthesizing the views from Barclays and Citigroup leads to a highly consistent conclusion.
The January NFP is noise; the annual benchmark revision is the true trend confirmation.
If employment is revised down by nearly 1 million jobs.
The perceived "resilience" of the 2025 labor market would be exposed as largely a statistical illusion. The US job market in 2026 would not be "stabilizing once again," but rather continuing to decelerate from a significantly lower starting point.
This is a risk for which the market remains inadequately priced.
Comments