Quits hit an 18-month high, but job-hunting risk depends on a job offer
September's jump in quits and a 27.1% long-term unemployment share point to a market that rewards people with options and punishes those without.
Quits reached 3.365 million in September, the highest level since March 2025, while 27.1% of unemployed Americans had been jobless for 27 weeks or more. The data suggest that leaving for another job is a different bet from leaving with nothing lined up.
Quits rose to 3.365 million in September, the highest level since March 2025, according to a news analysis of the Bureau of Labor Statistics (BLS) Job Openings and Labor Turnover Survey (JOLTS) released Nov. 3. In the BLS jobs report for the same month, 27.1% of unemployed Americans had been out of work for 27 weeks or more, up from 23.6% a year earlier. Read side by side, the two figures seem to say opposite things about whether leaving a job is a confident move or a reckless one.
The evidence supports a narrower answer than either headline. The quits jump is a single month, and BLS itself describes the quits rate as little changed. Long-term unemployment is high compared with a year ago, but it has hovered in a band since spring. The two numbers describe different groups: employed workers who may have options, and jobless workers who are struggling to find work.
What the quits number does and does not show
The September JOLTS reading was a sharp move. A news analysis of BLS data reports that quits rose from a revised 3.044 million in August to 3.365 million, and the quits rate rose from 1.9% to 2.1%. TheStreet Pro, citing the same release, describes the increase as the largest since May 2023 and well above the 3.096 million estimate.
The official release is more restrained. BLS reported that quits increased to 3.4 million but that the rate was little changed at 2.1%, and that quits were up 375,000 over the year. That wording matters: a statistical agency is signaling that the monthly move should not be over-read.
A higher count, but not an unusually high rate
The rate puts the jump in perspective. BLS reported 3.3 million quits and a 2.1% rate in March 2025, a figure that has since been revised. In March 2018 the rate was 2.3% at the same 3.3 million quits, according to a BLS archive. Employment is larger now, so a given count translates into a lower rate. On that measure, September looks like a return to early-2025 levels rather than a surge in confidence.
The month before the jump was also unusually weak. Secondary sources, including a university career-services blog, say August's 1.9% rate matched the lowest since 2020, though that figure should be checked against BLS tables. A rebound from a low base can look dramatic without signaling a lasting change.
Where the quits came from
TheStreet Pro reports that restaurants and hotels accounted for most of the gain in quits. That detail comes from a single source and has not been verified in the BLS industry tables. If it holds, it matters because those industries have naturally high turnover, so the jump may say little about the options facing an average office or professional worker.
Another limit is built into the survey. JOLTS counts workers who left for a better job and workers who left with nothing lined up in the same total. The 3.365 million figure therefore cannot be read as a count of people stepping into unemployment risk, nor as a count of people trading up.
Other September JOLTS figures sketch a market that is active but not tight. Layoffs and discharges were 1.641 million, the lowest since March 2025, according to Trading Economics. Hires were 5.402 million, a 3.4% hiring rate, and openings were 7.523 million, a 4.5% rate. Openings remain well below the March 2022 peak of 12.301 million, TheStreet Pro reports.
One analyst at Investing.com calculated about 205 quits per 100 layoffs in September, against 172 in August and 165 a year earlier. That is a derived calculation rather than an official statistic, but it illustrates the point: workers were leaving by choice at a faster pace relative to those being pushed out.
Long-term unemployment: elevated, but not accelerating
The September jobs report showed payrolls up 29,000 and an unemployment rate of 4.2%, with 7.1 million people unemployed, according to BLS. Of those, 27.1% had been jobless for 27 weeks or more. BLS Table A-12 puts the seasonally adjusted number at roughly 1.94 million, compared with 1,815,000 in September 2025, when the share was 23.6%.
So the claim that long-term unemployment is worsening is true year over year. Month to month, the picture is flatter. The share was 25.3% in April, rose to 27.5% in May, and has moved between 25.5% and 27.3% since, according to BLS releases and Table A-12. The recent pattern looks like a plateau at a high level rather than a continuing climb.
| Label | Long-term unemployed as % of all unemployed, seasonally adjusted |
|---|---|
| Sep 2025 | 23.6 |
| Mar 2026 | 25.4 |
| Apr 2026 | 25.3 |
| May 2026 | 27.5 |
| Jun 2026 | 27.3 |
| Jul 2026 | 25.5 |
| Aug 2026 | 27 |
| Sep 2026 | 27.1 |
Texas A&M's Private Enterprise Research Center (PERC), analyzing BLS data, reported that the share was 25.4% in March 2026, against 21.3% in March 2025. It put average unemployment duration at 25.3 weeks, compared with 21.7 weeks in 2019, and the 2019 average share at 21.2%. By those yardsticks, jobless spells are lasting longer than before the pandemic.
High, but not a repeat of the early 2010s
The historical comparison is sobering in the other direction. In July 2013, BLS counted 4.2 million long-term unemployed, 37.0% of all unemployed people. Today's share is high for an unemployment rate of 4.2% but well below that earlier episode. The typical jobless spell is also much shorter than the average: an RSM forecaster quoted by Kiplinger put median unemployment duration at about 11.5 weeks.
Part of the rise is arithmetic. When few people are newly losing jobs, those who have been unemployed for a long time make up a larger slice of a small pool. That reading is an interpretation of the sources rather than a finding by BLS. It also explains how the share can climb while layoffs are at their lowest level since March 2025.
Why a split market can produce both readings
Federal Reserve researchers have described a labor market that treats the employed and the unemployed differently. A May 2026 Dallas Fed analysis found reduced hiring and rising firing for the unemployed, alongside a comparatively robust and insulated market for workers who already have jobs. The research brief summarizes the Dallas Fed's description as "hire-less, fire-more" for people without work.
A secondary summary of San Francisco Fed work (Economic Letter 26-20, Aug. 5, 2026) says the job-finding rate for the unemployed is typically 25% to 30% per month, and concludes that finding work has become harder for people actively looking. The primary letter was not available for this article, so those figures should be treated as secondhand.
Broader numbers fit that description. CNBC reports that payroll growth from January 2025 through September 2026 averaged just 35,000 a month, while unemployment rose only 0.2 percentage point. Labor force participation fell to 61.8%, a 0.8-point slide. A shrinking labor force is part of why the unemployment rate has barely moved even as hiring has been weak.
Anxiety that suppresses quitting
An Atlanta Fed research note dated Oct. 6, 2026 describes the market as "low-hire-low-fire-low-quit." Its estimates link worker anxiety to staying put: a 1-point rise in expectations of not finding a job lowers the probability of switching jobs by 0.021 points, against a baseline switching probability of 1.6%. The effect per point is small, but it points to worry about re-employment as one reason workers stay.
That context makes September's quits rebound harder to interpret. It could reflect returning confidence among some workers, normal volatility in a low-turnover market, or the high-turnover industries noted above. The sources retrieved cannot distinguish among those explanations.
What switching jobs pays now
The Atlanta Fed's Wage Growth Tracker, a 3-month moving median of pay changes, still shows a premium for changing employers. In August 2026 the overall tracker was 4.1%, with stayers at 3.6% and switchers at 5.0%, up from 4.4% in July, according to the Atlanta Fed's post on X. In January, stayers were at 3.5% and switchers at 4.7%, per the Atlanta Fed.
A vendor blog, Joveo, reports September figures of 3.9% overall, 4.7% for switchers and 3.4% for stayers. That is a weak source and the numbers need confirmation on the Atlanta Fed's site. If they hold, the gap between switchers and stayers has stayed at roughly 1 to 1.4 points through 2026, with July's gap of 0.8 points the narrowest.
Other data suggest a thinner premium
Fortune reported in June, citing Bank of America data, that the switching premium has eroded since 2022, and that Gen X and Boomer switchers saw flat or declining wages in the first quarter of 2026. Fortune also cited ADP data showing the premium hit a record low of 1.9 percentage points in early 2026. The article's year-by-year comparison is ambiguous, and the underlying Bank of America and ADP reports were not retrieved.
The sources disagree partly because they measure differently. The Atlanta Fed tracker is a median drawn from matched Current Population Survey records, while the others draw on payroll or bank-deposit data. Any of them can be right about its own sample while telling a different story.
All of these premiums share a limit: they are measured only among people who found a new job. A worker who quit and never landed one does not appear in the switcher figure, so the premium says nothing about the outcomes that matter most for the risk question. For context on pay overall, Trading Economics reports average hourly earnings up 3.0% year over year in September.
Where the risk sits: an offer in hand versus nothing lined up
Putting the pieces together produces an inference, not a direct finding. For someone who leaves one job for another, the data show an economy with low layoffs, an average switching pay gain of about 1 point in the Atlanta Fed series, and an employed-worker market the Dallas Fed calls comparatively insulated. For someone who leaves with nothing lined up, the same data show slow hiring for the unemployed, longer jobless spells than before 2020, and a long-term unemployment share of 27.1%.
The headline unemployment rate of 4.2% understates that second risk. It counts the share of the labor force without work, and it has barely moved partly because participation has fallen. It does not describe how long a newly jobless person can expect to search.
- Leaving for another job: layoffs are at their lowest since March 2025, and switchers still out-earn stayers in the Atlanta Fed tracker, though other measures show a smaller premium.
- Leaving with nothing lined up: job-finding is described as slower for the unemployed, and 27.1% of the unemployed have been out of work for 27 weeks or more.
- Staying put: stayers' pay growth was 3.4% to 3.6% in the 2026 months the tracker shows for August and September, below switchers' 4.7% to 5.0%.
What to watch next
The outlook is unsettled. Layoffs are low, but payroll growth has averaged only 35,000 a month since January 2025. One source warns that the low-fire half of the market is fragile and could tip toward more layoffs, while others expect steady hiring; that is forecasting rather than evidence. Kiplinger also reports that the Federal Open Market Committee is focused on inflation, and one forecaster says a modest jobs number would keep a December rate hike in play.
The September quits jump is best read as a data point, not a verdict. A single month, a statistically unchanged rate and a possible concentration in restaurants and hotels argue against treating it as an all-clear. The more durable signal sits in the gap between workers with jobs and workers without: the same market that offers movers a modest pay bump has made the search slower for people who start it from unemployment. The October report and the December JOLTS release will show whether that gap is narrowing.
Frequently asked questions
Are quits really at an 18-month high?
In level terms, yes. Quits of 3.365 million in September were the highest since March 2025, according to a news analysis of BLS data. But BLS describes the quits rate of 2.1% as little changed, and the rate was 2.1% in March 2025 as first published.
Is long-term unemployment getting worse?
Compared with a year ago, yes: 27.1% of the unemployed had been out of work 27 weeks or more in September 2026, against 23.6% in September 2025, per BLS. Since spring the share has stayed between 25.3% and 27.5%, which looks more like a plateau.
Do people who change jobs still get bigger raises?
In the Atlanta Fed Wage Growth Tracker, yes. Switchers were at 5.0% in August 2026 against 3.6% for stayers. Fortune, citing ADP and Bank of America data, reports a smaller or eroding premium, and all of these measures cover only people who found a new job.
How risky is quitting without another job lined up?
The retrieved evidence points to more risk than the 4.2% unemployment rate suggests, because the Dallas Fed describes reduced hiring for the unemployed and the long-term unemployment share is elevated. This is an inference from aggregate data, and it cannot predict an individual's search, which depends on field, location and finances.
Does the quits number include people who left with nothing lined up?
Yes. JOLTS counts voluntary separations reported by employers and does not show whether the worker had another job waiting, so the total mixes workers leaving for better jobs with those leaving without one.
Sources
- Job Openings and Labor Turnover - September 2026 , U.S. Bureau of Labor Statistics (2026-11-03)
- Employment Situation News Release - 2026 M09 Results , U.S. Bureau of Labor Statistics (2026-10-02)
- Table A-12. Unemployed people by duration of unemployment - 2026 M09 Results , U.S. Bureau of Labor Statistics (2026-10-02)
- Employment Situation News Release - 2026 M04 Results , U.S. Bureau of Labor Statistics (2026-05-08)
- Employment Situation News Release - 2026 M07 Results , U.S. Bureau of Labor Statistics (2026-08-07)
- Job Openings and Labor Turnover (March 2025 release) , U.S. Bureau of Labor Statistics
- Job Openings and Labor Turnover (March 2018 release) , U.S. Bureau of Labor Statistics (2018-05-08)
- BLS – July payroll employment increases; unemployment rate edges down , Harvard TagTeam (beSpacific, quoting BLS) (2013-08-02)
- US Job Openings Hold at 7.5 Million, but Worker Turnover Is Shifting , Investing.com (2026-11-03)
- JOLTS Report Strong Across the Board , TheStreet Pro
- United States Job Openings , Trading Economics
- United States Non Farm Payrolls , Trading Economics
- Rising Labor Market Anxiety and Declining Quits , Federal Reserve Bank of Atlanta (2026-10-06)
- Climbing the employment ladder tough when bottom rung is broken , Federal Reserve Bank of Dallas (2026-05-26)
- How Long Individuals Remain Without Work in a Changing Labor Market , Texas A&M Private Enterprise Research Center (2026-04-01)
- The Not-so-Fluid, Low-Hire, Low-Fire Economy , Conversable Economist (2026-08-05)
- Atlanta Fed on X: Wage Growth Tracker edged up to 4.1% in August , Federal Reserve Bank of Atlanta
- Wage Growth Tracker 3.6 Percent , Federal Reserve Bank of Atlanta (2026-02-23)
- Wage Growth in 2026: What the Employment Cost Index and Wage Data Mean for Job Advertising , Joveo
- As loyal Boomers win and job-switching Gen Zers lose, the labor market of 2026 reveals a decade of bad career advice , Fortune (2026-06-01)
- The October jobs report is out Friday. Here's everything to expect , CNBC (2026-11-05)
- What to Expect From the October Jobs Report , Kiplinger
- What to Look Out for in Economic Data This Week (November 2-6) , Kiplinger
- October Jobs Report Seen Showing Resilient Labor Market , Morningstar (2026-11-05)
- What Does "Low-Hire, Low-Fire" Mean for New Grads? , University of Richmond Career Services (2026-10-21)