UK unemployment falls to 4.6%: why payrolls tell a weaker story
Unemployment fell to 4.6%, but payrolls and vacancies weakened. The retreat by small employers gives me more reason for caution than the headline gives for optimism.
5 min read
UK unemployment has fallen to 4.6%, but I would hesitate to call it a recovery. Over the same three months, payroll employment fell by 32,000. Vacancies fell by 22,000 in August to October, with the smallest businesses accounting for a drop equal to that entire decline.
The encouraging headline and the weaker hiring figures are both official statistics. The useful question is which tells us more about the jobs people can actually get. For now, I put more weight on payrolls and vacancies.
The jobs market behind the unemployment headline
The ONS November labour market release gives this picture:
| Measure and period | Latest estimate | Change |
|---|---|---|
| Unemployment rate (16+), July to September | 4.6% | Down 0.3 percentage points on the quarter |
| Employment rate (16 to 64), July to September | 75.4% | Up 0.3 percentage points on the quarter |
| Payrolled employees, October (provisional) | 30.2 million | Down 34,000 on the month |
| Vacancies, August to October | 690,000 | Down 22,000 on the quarter |
Lower unemployment and higher employment would usually be good reasons for optimism. Here, an awkward detail weakens that reading: part of the improvement came from a change in the households being surveyed.
The Labour Force Survey replaces a group of respondents each quarter. This time, the departing group had lower employment and higher unemployment than its replacement. ONS says the change contributed to the better headline figures. The incoming group was broadly in line with other recent groups, so this casts more doubt on the size of the quarterly fall than on the 4.6% level itself.
That leaves room for a real improvement. It also means the whole fall cannot confidently be credited to employers creating more work. I would want another measure to support the recovery story before accepting it.
Payroll employment is drifting down
Payroll figures are based on employers' tax records and exclude the self-employed. Across July to September, the comparable period for the unemployment figure, they were down 32,000 on the quarter and 90,000 on the year.
The November PAYE bulletin then records a provisional fall of 34,000 in October, leaving payroll employment 95,000 below a year earlier.
Payroll employment has slipped from its 2024 peak
UK payrolled employees, seasonally adjusted (millions)
Line chart of UK payrolled employees, seasonally adjusted, from January 2023 to October 2026. The number of payrolled employees was 29.96 million in January 2023, peaked at 30.46 million in October 2024 and was a provisional 30.25 million in October 2026, about 215,000 below the peak.
UK payrolled employees
Millions of employees; axis starts at 29.9 million
- Payrolled employees
- Provisional
October 2026 is provisional. Data vintage: 17 November 2026.
Source: ONS / HMRC, November 2026 PAYE release
View data
| Period | Payrolled employees |
|---|---|
| January 2023 | 29,964,056 |
| February 2023 | 30,019,637 |
| March 2023 | 30,078,201 |
| April 2023 | 30,107,532 |
| May 2023 | 30,128,648 |
| June 2023 | 30,176,633 |
| July 2023 | 30,179,959 |
| August 2023 | 30,202,388 |
| September 2023 | 30,239,447 |
| October 2023 | 30,279,015 |
| November 2023 | 30,313,313 |
| December 2023 | 30,360,308 |
| January 2024 | 30,373,405 |
| February 2024 | 30,380,053 |
| March 2024 | 30,390,683 |
| April 2024 | 30,414,051 |
| May 2024 | 30,436,219 |
| June 2024 | 30,444,717 |
| July 2024 | 30,451,388 |
| August 2024 | 30,430,896 |
| September 2024 | 30,430,343 |
| October 2024 | 30,462,742 |
| November 2024 | 30,444,701 |
| December 2024 | 30,424,155 |
| January 2025 | 30,421,432 |
| February 2025 | 30,404,270 |
| March 2025 | 30,391,532 |
| April 2025 | 30,386,668 |
| May 2025 | 30,366,606 |
| June 2025 | 30,353,701 |
| July 2025 | 30,358,411 |
| August 2025 | 30,376,682 |
| September 2025 | 30,348,341 |
| October 2025 | 30,342,582 |
| November 2025 | 30,319,260 |
| December 2025 | 30,325,415 |
| January 2026 | 30,332,124 |
| February 2026 | 30,322,150 |
| March 2026 | 30,320,027 |
| April 2026 | 30,310,472 |
| May 2026 | 30,309,507 |
| June 2026 | 30,287,496 |
| July 2026 | 30,268,448 |
| August 2026 | 30,261,917 |
| September 2026 | 30,281,669 |
| October 2026 (provisional) | 30,247,443 |
The longer view matters more than one month's movement. Payroll employment in October was about 215,000 below its October 2024 high in the period shown, a decline of roughly 0.7%. The latest monthly fall was about 0.1%. On the current data, payroll employment fell in 19 of the 24 months after that peak.
Those figures describe a gradual weakening. They do not justify talk of a sudden collapse. Equally, a small decline repeated across a long period can leave fewer opportunities for someone trying to move jobs or enter the workforce.
The October estimate is provisional and may be revised as more employer payroll submissions arrive. A revision could change the latest month's story; it would take a broader improvement to reverse the direction visible in the chart.
Small firms are pulling back on recruitment
The vacancy figures strengthen the case for caution. At 690,000, vacancies were down 22,000 on the quarter. Businesses employing one to nine people accounted for a fall of 22,000, leaving an estimated 73,000 openings in that group.
On ONS's rounded estimates, that is about a 23% fall for the smallest employers, from an implied 95,000 in the previous quarter. Across the other employer size bands combined, the net change was roughly flat. There were increases and decreases within those bands, so this was not a uniform retreat by every kind of business.
This concentration matters. A business with a handful of staff can face a large commitment when it adds one more person. If the owner is uncertain about orders or costs, leaving a vacancy unfilled may be easier than hiring and later having to cut back. That is an explanation of the incentive, not a claim about every firm's decision.
ONS reports that feedback from smaller employers points to labour costs and other economic conditions affecting recruitment. The survey cannot assign every missing vacancy to one cause, but it gives the payroll weakness a plausible business context.
For a jobseeker, a vacancy that never appears is still an opportunity lost. A cooling labour market can be felt through longer searches and fewer alternatives before it produces a dramatic unemployment headline.
The strongest reason to be less pessimistic
There is a reasonable case against leaning too heavily on payrolls. The survey covers more kinds of work, including self-employment, and payroll data cannot tell us where every departing employee went.
Under the ONS definition of unemployment, someone must be without work, seeking it and available to start. A person who leaves a payroll to become self-employed still has a job. Someone who retires and stops looking is economically inactive. Neither change has to raise unemployment.
Those possibilities explain why the measures can diverge. They do not establish that a welcome shift into self-employment explains this particular decline. Nor would I infer a surge in people giving up work: the working-age inactivity rate actually edged down to 20.8% in the survey.
My argument is narrower. When a survey improvement is partly affected by its changing sample, and both payrolls and vacancies weaken, the evidence for a broad hiring recovery is unconvincing. A sustained improvement across those measures would change my view.
Easier hiring can come with weaker demand
For businesses and investors, the effect depends on where they sit. A recruitment agency needs employers to fill roles. A retailer may welcome easier recruitment while worrying about its customers' incomes. The same labour-market release can therefore help on one side of a company's accounts and hurt on the other.
Wage growth adds another distinction. In Great Britain, average regular pay, excluding bonuses, rose 3.5% annually, down from 3.7% in the previous three-month period. After adjusting for inflation including owner-occupiers' housing costs, the rise was only 0.2%.
That is a thin improvement in purchasing power. A business should be careful about assuming that lower pay pressure automatically arrives alongside stronger consumer spending. I would look at sales volumes, staff turnover and vacancies in its own market before drawing a conclusion from the national unemployment rate.
It is also too much to ask this release to settle the Bank of England's next decision. Pay, inflation and demand can pull policy in different directions. For savers, even an eventual policy change need not pass through immediately, as I explain in why savings rates do not always follow central-bank rates.
The next useful sign would be employers retaining more people and offering more jobs, alongside a sustained improvement in the survey. Until that happens, 4.6% is an encouraging number with an unconvincing recovery story behind it.
Comments
No comments yet - start the conversation.