Careers

Hiring Looks Strong, but Unemployment Spells Are Getting Longer

October payrolls beat forecasts, yet average time out of work hit its highest level since 2021, and recent graduates still face a harder market than before the pandemic.

11 min read
Empty railway platform at sunset, with tracks running into the distance under overhead wires and a dirty glass-walled waiting shelter on the right.
Image by RoumanetD from Pixabay

Payrolls rose 165,000 in October, but the average unemployed worker has been looking for 27.1 weeks. The gap reflects a low hiring rate, thin job quality for graduates, and limits in what the data can measure. Employer surveys offer clues about adapting, though proven strategies are scarce.

The October jobs report beat forecasts, but the average unemployed American has now been looking for work for 27.1 weeks, according to figures reported by CNBC. SHRM describes that as the highest reading since September 2021. Strong hiring headlines and long searches can coexist because they measure different things: payroll growth counts net jobs added, the hires rate tracks how often employers bring people on, and unemployment duration tracks how long jobless people have been out of work.

The distinction matters most for two groups: workers who lose jobs in a market where layoffs are low but hiring is slow, and recent college graduates entering it. Reuters noted after the report that the recent pickup has not yet begun to shorten the time people who lose a job spend out of work, and that those workers are out of work longer than at any point in five years. No source in the available evidence measures how long a completed job search takes, so the picture has to be assembled from several imperfect measures.

Strong payrolls, but a slow hiring rate

The Bureau of Labor Statistics (BLS) reported that nonfarm payrolls rose 165,000 in October and the unemployment rate stood at 4.1%. Forecasts were far lower: CNBC cited a Dow Jones consensus of 59,000, and Reuters cited a poll median of 65,000. September's gain was also revised up to 60,000 from an initial 29,000, according to Reuters.

Employment grew in professional and business services (70,000 jobs, per the BLS), health care, and transportation and warehousing. CNBC reported that the labor force expanded by 421,000 after shrinking, a swing analysts connect to demographics and reduced immigration. SHRM noted that the unemployment rate has stayed between 4.1% and 4.3% since March.

Estimates of how many jobs are needed each month to keep conditions stable are modest. RSM puts the break-even pace at about 35,000 jobs a month, while Reuters reporting carried by RTHK puts it at no less than 40,000. By either estimate, October cleared the bar easily. The question for job seekers is whether that translates into more hiring of people who are currently unemployed.

What the hiring flows say

The BLS Job Openings and Labor Turnover Survey (JOLTS) for September, released Nov. 3 and preliminary, counted 7.5 million openings, 5.4 million hires and 1.6 million layoffs and discharges. The hires rate was 3.4%, and the layoff rate was 1.0%. Economist Guy Berger reported that the hires rate was 3.3% a year earlier and the openings rate 4.5%, up from 4.3%.

TheStreet, using Bloomberg figures, put the 2019 average hires rate at 3.87%, so September remained below the prepandemic norm. The same report cited 1.06 openings per unemployed worker. Berger wrote that a 3.4% hires rate was last seen in the 2010s, when unemployment was just over 7%, and he characterized the current recovery as "small, incremental and slow." That is one analyst's reading, not a dataset, but the underlying JOLTS rates are official.

Not every analyst reads the data that way. RSM described the economy as mid-cycle with improving hiring. The two views are not strictly contradictory: payroll gains support the stronger interpretation, while a hires rate below its 2019 level supports the cautious one.

What the duration numbers show, and what they miss

CNBC reported that the average (mean) duration of unemployment jumped 2.3 weeks in October to 27.1 weeks on a seasonally adjusted basis. SHRM called that the highest since September 2021, while CNBC framed it as the longest outside the pandemic period since 2016. The two descriptions are consistent but handle the pandemic differently, so they should not be blended.

Long-term unemployment shows the same direction. The BLS reported about 2.0 million people jobless for 27 weeks or more in October, 27.6% of all unemployed people. CNBC said that is the highest share since late 2021. RSM, a secondary source, put the October median duration at 11.9 weeks.

Unemployment: fewer people, longer spells (September, not seasonally adjusted)
Measure Sep 2025 Sep 2026
Unemployed (thousands) 7,324 6,859
Mean duration (weeks) 24.7 25.4
Median duration (weeks) 10.7 12.1

The table, drawn from BLS Table A-35, shows a smaller pool of unemployed people than a year earlier, with longer spells at both the mean and the median. That pattern matters because a single month of data can mislead. The seasonally adjusted mean moved from 24.8 weeks in September to 27.1 in October, a swing large enough that part of it may be sampling noise in a survey. The year-over-year change in unadjusted figures points the same way, which strengthens the case that the lengthening is real.

Historical perspective

Current levels are well short of the worst postrecession stretch. The BLS reported that long-term unemployment exceeded 40% of the unemployed from December 2009 to November 2012. In October 2016, by contrast, the BLS recorded a median duration of 10.2 weeks and a long-term share of 25%. Today's 27.6% share sits above that 2016 reading but far below the post-2008 peak.

The measure also has a structural limit. Because it captures spells still underway, and excludes people who search while employed, it is not a direct reading of how long a typical search takes. One secondary source in the research suggests in-progress medians tend to understate completed spells, but that is an unverified inference and is treated here as such.

Why searches are lengthening: a low-hire, low-layoff market

CNBC reported that layoffs and discharges fell to their lowest volume in more than a year. Few layoffs keep the unemployment rate low, and fewer new entrants to the pool means fewer short spells to pull the average down. At the same time, a hires rate of 3.4% means relatively few people are being moved out of the pool. Taken together, the data fit a smaller but stickier group of jobless workers.

That mechanism is an interpretation, not a tested finding. No study in the available evidence isolates how much of the rise in duration comes from low hiring, and how much from the mechanical effect of fewer fresh layoffs. The reading is consistent with official data on layoffs and hires, but it should be held with moderate confidence.

There is also a counterargument. Duration tends to lag the rest of the labor market, so it could fall later if the hiring pickup continues. Reuters' wording was careful on this point: the pickup had not yet shortened spells. That leaves the direction open rather than settled.

Strain beneath the headline

SHRM, citing BLS data, reported that the broad U-6 rate, which includes discouraged workers and involuntary part-timers, rose to 7.9% from 7.6%. Involuntary part-time work rose by 339,000 to 4.84 million. These figures suggest that some workers are settling for fewer hours than they want.

Financial pressure also builds with time. CNBC noted that most unemployment benefits end after about 26 weeks, and that a rising long-term share can partly explain why jobless claims stay low. NPR reported in October that nearly four in ten people exhaust their benefits before finding work. That figure rests on NPR alone, and the underlying dataset was not located, so it is best read as indicative.

Recent graduates: lower unemployment, thinner job quality

The New York Fed defines recent graduates as people ages 22 to 27 with a bachelor's degree. In a May 2026 media advisory, the bank reported that their unemployment rate was 5.6% in March 2026, up from 3.6% in March 2019. For the third quarter of 2026, the bank's college labor market page says the rate fell to 5.2%, its lowest in more than a year. Reports of the third-quarter figure differ slightly by measure, so only the New York Fed's own number is used here.

Underemployment, defined by the New York Fed as working in a job that typically does not require a degree, moved the other way. The bank reported that it ticked up in the third quarter, and the Maryland Daily Record, reporting on New York Fed data released Nov. 3, said it is at its highest since October 2020. Both trends can be true at once: fewer graduates are jobless, but more are in jobs that do not use their degree.

An NYC Comptroller analysis also found that unemployment among graduates ages 22 to 27 now exceeds that of older workers without a degree, reversing a pattern that held until late 2022. The report's date was not visible in the retrieved text. A 5.2% rate is not extreme in absolute terms, so the concern lies in the gap relative to older workers and to 2019, plus the underemployment.

The employer side

Employer surveys from the National Association of Colleges and Employers (NACE) show a volatile outlook. In fall 2025, NACE projected hiring for the Class of 2026 would rise 1.6% on average, while the median projection was a 2.4% decline. A spring update projected a 5.6% increase at the median. The spring survey ran from Feb. 12 to Mar. 17, 2026, drew 185 respondents and was sponsored by Jobscan, so it is a small, voluntary sample, and the switch between mean and median complicates comparison.

NACE also said growth is uneven across industries, and about 45% of employers rated the market "fair." NACE headlines report that more than two in five 2026 graduates had an offer before graduating, and that employers made fewer offers but earlier. That timing detail is relevant to anyone planning a search around the academic calendar.

Outcomes vary by major. Forbes, citing New York Fed research on 2024 Census data across 73 majors, reported that education majors have the lowest unemployment, with special education at 0.7% and elementary education at 1.2%. Forbes noted that low unemployment can coexist with high underemployment, so a low jobless rate alone does not show a good match between degree and job.

Is AI lengthening searches?

Axios reported on Nov. 3 on a Goldman Sachs analysis of more than 15,000 workers laid off from 2016 to 2025. Workers most likely displaced by AI in 2025 took roughly a month longer to find work, and were about 10 percentage points more likely to be unemployed a year later. Among workers under 35 who were most likely displaced by AI, the delay in finding a full-time job was about three weeks.

Axios also described a shift in where these workers land. Among displaced workers from AI-exposed jobs who found work in 2025, about three-quarters took occupations with similar or greater AI exposure, versus just over half in 2023. In other words, moving to a less exposed field became less common.

The limits are substantial. AI displacement in the analysis is estimated, not counted, inferred from a worker's occupation, industry AI adoption and circumstances. It comes from one bank's research, the report itself was not located, and it has not been independently replicated. Oxford Economics' Michael Pearce, quoted by Reuters, added that the most AI-exposed sectors, information and finance, have been shedding jobs on trend, while professional and business services added them. The size of AI's role in overall duration remains unresolved, and the brief did not find evidence testing alternatives such as remote work.

What the evidence supports for adapting

Evidence on which job-search tactics work is thin in the research available here. What exists is employer-side data on what companies say they want and when they recruit, which is informative but not proof that any particular strategy shortens a search. This is context, not personal advice.

  • Timing: NACE headlines say employers made fewer offers but earlier, so searches that start late in the academic year may meet a narrower field.
  • Demonstrated skills: NACE reports that 70% of employers use skills-based hiring, up from 65%.
  • AI literacy: NACE reports that demand for AI skills in entry-level jobs has nearly tripled since fall 2025.
  • Sector exposure: BLS data show job gains in professional and business services, health care and transportation and warehousing, while Oxford Economics' Pearce points to job shedding in information and finance.

Financial planning is a separate consideration. With most benefits ending after about 26 weeks, according to CNBC, the length of a search interacts with the timing of income support. Whether job seekers who adapt in these ways actually find work faster is not something the available data can show.

What to watch

The strongest test comes in the next two data releases. The next JOLTS report is due Dec. 1 and the next jobs report Dec. 4. If the hires rate rises from 3.4% and the long-term share starts to fall from 27.6%, the case that duration is simply lagging the recovery will gain support. If payrolls stay strong while spells keep lengthening, the case for a market that adds jobs without moving the jobless back into work will strengthen.

Reuters also raised the possibility of Fed rate hikes tied to inflation from a Middle East conflict, which could weigh on hiring. That is speculative at this stage. For now, a strong headline and a long search describe the same labor market from two angles.

Frequently asked questions

How long does a job search take right now?

No source in the research measures completed searches. The BLS measures how long unemployed people have been looking so far: the average was 27.1 weeks in October, per CNBC, and the median was 11.9 weeks, per RSM. Both describe searches still in progress.

Why is unemployment duration rising when jobs are being added?

Payroll growth counts net jobs added, not how often the unemployed are hired. The September hires rate was 3.4%, below its 2019 average of 3.87% according to TheStreet, while layoffs were low. Low layoffs and slow hiring together fit a smaller but stickier pool of jobless workers, though that link is an interpretation.

What counts as long-term unemployment?

The BLS defines it as being jobless for 27 weeks or more. In October, about 2.0 million people fit that definition, or 27.6% of all unemployed people. In October 2016 the share was 25%.

Are recent graduates still struggling to find jobs?

The New York Fed says unemployment for recent graduates ages 22 to 27 fell to 5.2% in the third quarter of 2026, its lowest in more than a year. It was 3.6% in March 2019. Underemployment, working in a job that typically does not require a degree, ticked up and was reported at its highest since October 2020.

Is AI making job searches longer?

A Goldman Sachs analysis reported by Axios found that workers most likely displaced by AI in 2025 took roughly a month longer to find work. The AI label is estimated rather than observed, and the finding has not been independently replicated, so AI's overall role remains uncertain.

Sources (24)

  1. Employment Situation News Release - 2026 M10 Results , U.S. Bureau of Labor Statistics (2026-11-06)
  2. A-35. Unemployed total and full-time workers by duration of unemployment , U.S. Bureau of Labor Statistics (2026-10-02)
  3. Job Openings and Labor Turnover Survey News Release - 2026 M09 Results , U.S. Bureau of Labor Statistics (2026-11-03)
  4. The Economics Daily: Duration of unemployment, October 2016 , U.S. Bureau of Labor Statistics (2016-10-01)
  5. A long-term unemployment trend is hiding in the October job report , CNBC (2026-11-06)
  6. U.S. economy added 165,000 jobs in October, way more than expected; unemployment rate slips to 4.1% , CNBC (2026-11-06)
  7. US Job Growth Surged in October , SHRM (2026-11-06)
  8. Four things to note from Friday's US jobs report , Reuters via Regional Media News (2026-11-06)
  9. Strong October jobs report sends yields higher , Reuters via Yahoo Finance (2026-11-06)
  10. More new jobs added in US as unemployment eases , RTHK (2026-11-06)
  11. October jobs report shows strong hiring on the back of robust growth , RSM US (2026-11-06)
  12. JOLTS Recap (September) , MacroMostly (Guy Berger) (2026-11-03)
  13. JOLTS Report Strong Across the Board , TheStreet Pro (2026-11-03)
  14. It's taking longer to find a new job after losing one , NPR (2026-10-11)
  15. Workers laid off because of AI face longer job searches, Goldman Sachs finds , Axios (2026-11-03)
  16. The Labor Market for Recent College Graduates , Federal Reserve Bank of New York (2026-11-03)
  17. New York Fed to Release Research on the Role of Remote Work in Youth Unemployment , Federal Reserve Bank of New York (2026-05-28)
  18. Economy added 165K jobs in October, showing surprising strength , Maryland Daily Record (2026-11-06)
  19. The Job Market Is Tough For Grads. Which Majors Have The Lowest Unemployment Rates? , Forbes
  20. What Is Going on with NYC Jobs? , NYC Comptroller
  21. Hiring Flat for the College Class of 2026 , NACE (2025-11-12)
  22. Outlook Brightens for College Class of 2026 Entry-Level Hiring , NACE (2026-04-27)
  23. 2026 Job Outlook - Spring Update , NACE (2026-04-27)
  24. Trends & Predictions - Job Market , NACE

Share this article

Back to list