October Payrolls: As Strong as the Headline Suggests?

Estimates of payroll employment from previous months were revised higher. September’s employment gain was revised up by 84,000 positions to 159.13 million. August’s estimate increased by 53,000 to 159.07 million.

The composition of the increase in October payroll employment was heavily tilted toward the service sector. Goods-producing establishments added 15,000 jobs or 9.1% of the total. Employment in service-providing establishments increased by 144,000 or 87.3% of the total. The remaining 3.6% came from the government sector, where employment was estimated to have risen by 6,000 positions.

Within the service-providing sector, the largest increases in numbers came from professional and business services (+70,000), private education and health services (+53,000) and trade, transportation and utilities (+22,000).

General merchandise retailers and couriers & messengers – sectors in which employment is subject to significant seasonal variation – logged increases of 18,300 and 17,400, respectively. Such gains are almost certainly attributable to elevated staffing needs for the holiday season. Yet, the reported increases are seasonally adjusted. Thus, the non-seasonally adjusted gains in these two categories were higher – 64,800 and 57,000, respectively. General merchandise retailers and delivery companies are apparently optimistic about the upcoming holiday season.

On balance, while the job gains are stronger than those reported in recent months, the high concentration in services suggests, in my mind, that they probably do not signal a sustainable pickup in employment or economic activity. While there were solid gains in a few high paying sectors, such as professional, scientific and technical services (+32,900 jobs), other notable gains were recorded in sectors which usually have higher percentages of temporary and/or lower paying jobs. Besides general merchandise retailers and messengers & couriers, other examples of such subsectors include temporary help services, ambulatory health care services (including home health care), hospitals and individual & family social assistance services.

October’s employment gains are on par with some of the highest reported increases in recent months, but labor market growth remains well below levels seen in recent years, as shown in the chart below:

Monthly gains in non-farm payrolls have slowed steadily since the early months of the pandemic. The economy “lost” 22 million jobs in March and April 2020 and then recovered about half of those – 11.4 million jobs – from May to September of that year. The employment recovery continued after September 2020 but at a steadily declining pace. By the beginning of 2023, the labor market dynamics had normalized, with monthly job gains returning closer to their longer-term historical averages. The slowdown in monthly job gains continued in 2024 and 2025, with 2025 job gains hovering near zero. However, the 2025 estimates were affected by a large annual revision, which raises questions about just how weak job growth was that year. So far in 2026, with two more months to go monthly gains are higher than those reported in 2025, but only modestly so.

Assessments of employment trends have become more difficult recently because of larger than average annual revisions to the data. Each year, the BLS benchmarks its employment estimates from the establishment survey to employee counts from state unemployment insurance programs. In February 2026, the BLS announced that its revision to its March 2025 estimates reduced total nonfarm employment by 898,000 positions or 0.6 percent, about three times higher than average annual revisions. From the March baseline, each month was then revised according to a formula. The BLS considers the size of the benchmark revision as a gauge of the overall accuracy of the establishment survey. By that measure, its 2025 estimates of non-farm payrolls were meaningfully less accurate than average revisions in prior years.

Similarly, for the Household survey, population estimates are updated every January and compared against prior estimates. This year’s adjustment, implemented in February (one month later than usual because of the Federal government shutdown) would have reduced 2025 household job gains by about 1.4 million. (BLS does not revise prior monthly estimates of household employment, so the entire adjustment was recorded in January.)

In order to address some of its current methodology’s shortcomings or update its methodology to reflect changes in demographic trends, the Bureau has made changes, such as incorporating updated demographic information from the 2020 census, including net immigration. These methodology changes are intended to improve the accuracy of the Bureau’s initial estimates to reduce the magnitude of future population control adjustments.

Nevertheless, accuracy of the monthly employment data remains a concern. The BLS has said that responses to its monthly employer surveys have become much less timely; but it asserts that by the end of the three month period that it allows for revisions, participation rates are around 95%, equal to the historical averages when surveys were submitted on a more timely basis. As already noted, the October report included large revisions to the previous estimates for August (+53,000) and September (+86,000), so it would be wise to place less emphasis on the initial readout.

. Note: 2026 data is year-to-date through October.

With two months left to go in 2026, it appears that the full year figures for establishment (non-farm) payrolls will be better than 2025, but still among the smallest increases of the past 25 years. Yet, if the 2025 figures, after adjusting for the annual revision, are truly indicative of activity that year, then non-farm payrolls have been strengthening modestly this year and especially in recent months, as indicated by the rolling 12-month figures shown in the chart below.

Source: U.S. Bureau of Labor Statistics data

Among the measures to which I am paying more attention now is the labor force participation rate (LFPR). As the chart below show, the LFPR declined steadily from 2007 to 2015. Most of that decline has been attributed to the retirement of baby boomers. Yet, from 2015 to 2019, until the onset of the pandemic, the participation rate rose modestly, even though baby boomers were still retiring. The pandemic caused a sharp decline in the LFPR in 2020 and 2021, followed by with a partial recovery from 2022 to 2024. In 2025 and 2026, however, the participation rate has declined by 0.7 percentage points from 62.6% to 61.9%. (The figures in the chart are full year averages, with a 10-month average for 2026.)

The 2025 and 2026 declines in the LFPR appear to be out of step with current economic trends. We know, for example, that a significant number of households, especially those at the lower levels of the income spectrum, are struggling to make ends meet due in large part to higher interest rates and inflation. Households continue to be pressured by price increases in major goods and services, such as housing, food, energy and interest on loans. The toll is evident in declining consumer confidence and rising consumer loan delinquencies. With these pressures, more people should be looking for work. The LFPR did inch up by 0.1% to 61.9 in October, but it is still down over the past two years.

According to its methodology, people who don’t have a job and have not looked for a job within the past four weeks are not counted in the labor force. Certain BLS-defined categories of people not in the labor force include: (1) those who currently want a job, but are discouraged, (2) those who believe that no work is available, (3) those who have not been seeking a job because of school or family responsibilities, (4) those in poor health and (5) those with transportation problems.

The BLS, in its monthly employment report, calculates alternative measures of labor underutilization. U-3 is the official unemployment rate, which is simply the total number of unemployed divided by the total number included in the labor force. U-4 includes discouraged workers (added to both the numerator and denominator, since discouraged workers are not included in the labor force). U-5 includes categories 2-5 above, which together comprise the cohort that is “marginally attached” to the workforce (but not officially included in the labor force because they have not looked for work within the past four weeks). So the U-5 calculation includes this cohort in both the numerator and denominator. U-6, the broadest measure of labor force underutilization, adds in those who are working part-time for “economic reasons”, mostly because they cannot find a full-time job. (This cohort is added to the numerator, but not the denominator because it is included in the labor force.)

If more households are struggling to pay their rent, make their mortgage, auto loans and credit card payments, put food on the table and gasoline in their cars, we should expect the unemployed within those households would either be actively looking for work or be classified as discouraged, marginally attached or working part-time for economic reasons. However, BLS data shows that the recent measures of labor force underutilization are squarely within their historic averages.

In fact, the readings on unemployment claims suggest that labor market conditions are improving. Weekly claims have recently been running at around 200,000, well below their historical average of 220,000; while continued claims (i.e. those who are continuing to receive benefits after the first week) have been running at around 1.71 million, below their long-term average of 1.85 million.

Consequently, according to the BLS reports on employment and unemployment claims, there appear to be no signs of stress in the U.S. labor market.

There are two factors that can be contributing to this dichotomy between official labor market statistics and on the ground readings of consumer sentiment and rising loan delinquencies: First, the retiring of baby boomers still has a meaningful impact on the size of the labor force. (That influence is peaking, however, so that the downward pressure on LFPR from baby boomer retirements should subside by the end of the decade.) In addition, there has been a notable increase in early retirement offers, as evidenced by recent programs offered by Microsoft, Google/Alphabet, UPS and Verizon, as well as Federal government employee buyouts in the Defense sector and the IRS, among others. I am not aware of any tallies of the total impact of early retirement and voluntary buyout offers on a national basis, but Microsoft CoPilot says that these buyouts have expanded sharply across multiple industries this year. Voluntary separation packages have become a useful corporate tool to manage headcount, as they are a softer alternative to layoffs.

There are meaningful differences between what the BLS employment data and other sources, such as consumer confidence and anecdotal evidence, are telling us about the state of the U.S. consumer and by association, the U.S. economy. For that reason, until these differences are reconciled, it would be prudent to monitor these economic reports closely rather than just focusing on the headline figures.

November 12, 2026

Stephen P. Percoco
Lark Research
839 Dewitt Street
Linden, New Jersey 07036
(908) 975-0250
admin@larkresearch.com

© 2015-2026 by Stephen P. Percoco, Lark Research.   All rights reserved.

This blog post (as with all posts on this website) represents the opinion of Lark Research based upon its own independent research and supporting information obtained from various sources. Although Lark Research believes these sources to be reliable, it has not independently confirmed their accuracy. Consequently, this blog post may contain errors and omissions. Furthermore, this blog post is a summary of a recent report published on this subject and that report provides a more complete discussion and assessment of the risks and opportunities of any investment securities discussed herein. No representation or warranty is expressed or implied by the publication of this blog post. This blog post is for informational purposes only and shall not be construed as investment advice that meets the specific needs of any investor. Investors should, in consultation with their financial advisers, determine the suitability of the post’s recommendations, if any, to their own specific circumstances. Lark Research is not registered as an investment adviser with the Securities and Exchange Commission, pursuant to exemptions provided in the Investment Company Act of 1940. This blog post remains the property of Lark Research and may not be reproduced, copied or similarly disseminated, in whole or in part, without its prior written consent.