Labor Markets Briefs

Payroll Revisions Could Cut 800,000 from Jobs Gains, or Not

August 23, 2024

The BLS released its preliminary benchmark estimates Wednesday that showed a potential downward revision to payrolls of 818,000 over the 12 months from April 2023 to March 2024. This translates to an average 68,000 fewer payroll additions per month over the period. The announcement is preliminary with further potential changes when the  annual adjustment posted in February 2025.

Trusted Insights for What’s Ahead^®

• BLS payrolls currently show 2.9 million jobs added over the period between April 2024 and March 2024 (or 242,000 on average per month), a robust pace. However, the preliminary revision suggests that job gains were smaller at 2.1 million (or 173,000 on average per month).
• While representing a material change to payrolls, the preliminarily revised level would better align with other measures like household employment and JOLTS hiring rates.
• Even with the likely revision, payroll growth was healthy in 2023 and in Q1 2024 and Fed officials had posited that payrolls have been “overstated.”
• The downgrades do not alter the story of a gradual but sustainable labor market normalization, but do add to the growing evidence of softening signals in recent months.
• The final payroll revision in February 2025 may differ from this preliminary estimate, while data issues related to measuring immigration and new business creation continue to affect employment figures.

Largest Preliminary Benchmark Revision Since 2009

While the BLS announced payroll revision is the largest preliminary change since 2009 (Figure 1), the potential revisions do not change the story of healthy employment and labor force growth since the start of 2023. The overall narrative of strong employment and labor force growth remains unchanged, and better aligns payroll figures with other labor market measures. We’ve seen a divergence between monthly payroll and household employment gains over the past year, while JOLTS hiring has declined more quickly than payrolls.

Source: Bureau of Labor Statistics and The Conference Board.

Mild but growing signs of labor market softening have occurred only in the last several months. In addition, other factors like estimates of immigration and new business creation also raise questions about how the multiple employment series will be revised going forward.

Preliminary Revisions – By Sector

The BLS expects revisions to nonfarm payrolls may substantially affect 2023-24 job gains in several sectors (Figure 2). Composing half the revision, employment in professional and business services could be lowered by 358,000. This volatile sector, which includes temporary help services, has been a significant driver of revisions over the past 5 years, reflected by its 50% average contribution to final revisions over the 2019-2023 period.

Retail trade and leisure & hospitality—two primary drivers of post-pandemic employment growth—could see employment lowered by 129,000 and 150,000, respectively. The combination of these two sectors now shows cumulative preliminary revisions since 2019 of nearly 950,000. Employment in information and technology could also see an outsized revision that would lower employment by 68,000 or 2.3% of the sectors’ employment—the largest percentage revision.

Three sectors might see additional jobs added due to the revision: education & health, transportation & warehousing, and other services. Healthcare and transportation & warehousing have continued to be leading growth sectors in recent months, but the revisions would compound the concentration of recent gains as many sectors have plateaued.

Source: Bureau of Labor Statistics and The Conference Board.

Note: 2019-2023 contributions are based on the average of each sector's annual contribution to final revisions, not a share of the overall 5-year change.

Diverging Employment Series – Confounding factors: Immigration and New Business Creation

The estimated revision to payrolls reflects new data released by the Quarterly Census of Employment and Wages (QCEW) for 1Q 2024. The QCEW is less timely but more complete than monthly payrolls, relying on state unemployment insurance records for the more than 95% of US workers covered. That series is used to revise payroll employment annually, which comes from the Current Employment Statistics (CES) survey based on a much smaller set of 60,000 establishments.

Over the last year, the different employment series have diverged. The QCEW and CES data have historically moved together, and revisions help maintain a consistent gap between the measures (Figure 3). However, the pandemic recovery has altered the consistency of these measures. Over the Apr 2023-Mar 2024 period under revision, CES shows an employment gain of 1.9%, while the new release of QCEW for the first quarter shows a 1.3% increase. The household Current Population Survey (CPS), used to produce unemployment and participation, shows job gains have moderated more quickly than CES or QCEW, with just 0.4% growth or 642,000 jobs added over that 12 months.

Source: Bureau of Labor Statistics and The Conference Board.

Challenges in capturing the 2023 immigration surge are partially responsible for this divergence. The QCEW likely best measures legal immigrant workers. Yet, by relying primarily on quarterly unemployment insurance records submitted to states by private companies and governments, it does not capture the full population of unauthorized workers. CES payrolls and the CPS household survey are likely covering some portion of these workers, but neither do so comprehensively. When surveyed firms report monthly payrolls, it is traditionally thought that most respondents exclude unauthorized workers. Alternatively, the CPS survey has likely captured some of these households, particularly within ratios of foreign-born workers. However the CPS bridges the gap between survey results and the total US population by leveraging Census population counts from the year prior, which may be substantially undercounting immigrants.

BLS payrolls may also be overestimating the employment generated from net business creation, contributing to the large revision estimate. BLS estimates suggest that new firms continued contributing to payrolls near historical highs, while growth in the QCEW establishment count started to slow over 2023. Still the QCEW count remained above 2019 levels.

These factors push in opposite directions. It may be true that correcting payrolls down to the QCEW series growth rate helps factor out new business issues. But all three series may continue to undercount immigration’s impact.

Benchmark Changes before Final Release

The QCEW, which is also subject to revision, may ultimately give back some jobs in the final revision that are taken in the preliminary estimate. The QCEW will undergo several more revisions—some after the benchmark is set in February 2025—that may moderate the preliminary 818,000 downward estimate. In every quarter since 3Q 2020, QCEW employment has been revised upward (Figure 4). Those revisions have ranged between 120,000 and 300,000 since the pandemic, outside of 1Q 2022 where nearly 775,000 jobs were added through revisions on top of initial employment estimates. If upward revisions to QCEW continue and raise its employment level closer to CES payrolls, the final payroll benchmark revision could be smaller than BLS’ preliminary estimate.

Source: Bureau of Labor Statistics and The Conference Board.

Conclusion

Labor market data have produced signals in recent months that have Fed officials on alert for continued softening. The final payrolls revision that will be announced in February 2025 may still differ from this preliminary estimate, and measurement issues related to immigration and new business creation may continue to distort population and employment estimates. The revisions would signal cooler but still healthy labor market growth throughout 2023 and early 2024. The final release will distribute the adjustments over each month and shed more light on the path of the economy.

Our broader outlook has not shifted. Fed Chair Powell acknowledged payrolls could be “overstated” at the June meeting. Over that same period of revision, consumer spending remained strong, GDP growth surprised to the upside, and inflation continued to moderate. So far, there have been few layoffs or a substantial uptick in unemployment claims that could lead to broader deterioration. However, this potential lowering of payroll gains bolsters the concern in recent months of slow hiring being able to absorb workers who continue to re-enter the labor force.