The UK Hasn't Started Shedding Jobs. It Has Stopped Advertising Them.

The UK unemployment rate is 4.9%, largely unchanged over the quarter and up two tenths of a point over the year.

That will be reported as stability, and on its own terms it is. Nobody is losing their job in unusual numbers. Economic inactivity actually edged down. If you only read the headline you would conclude the labor market is fine and a slow search is a personal problem.

Then there is the number one line further into the same release.

Vacancies are at a level last seen in 2014

Vacancies for June to August were 702,000, down 8,000 on the quarter. They have been falling for a long time.

The Office for National Statistics puts its own benchmark on that figure, and it is the one worth keeping: outside the coronavirus pandemic period, the last time vacancies stood at 702,000 or fewer was August to October 2014.

Twelve years.

That qualifier about the pandemic is doing real work and should not be dropped. Vacancies collapsed in 2020 for reasons that have nothing to do with now. Setting those months aside, the current level of advertised demand is where it was more than a decade ago.

Alongside that, the payrolled employee count is shrinking. It fell 26,000 between July and August, and 101,000 over the year to July.

Put those together and you get a specific diagnosis

Unemployment flat. Inactivity down slightly. Employee numbers falling. Advertised vacancies at a twelve-year low outside the pandemic.

That is not a redundancy wave. A redundancy wave shows up as a rising unemployment rate, and this one has moved two tenths of a point in a year.

It is a hiring freeze. Employers are not cutting hard, they are not replacing, and they are not advertising. The stock of jobs is drifting down through people leaving and not being backfilled, which is quiet, slow, and almost invisible in the headline rate.

One honest caveat on that employee number. The payrolled count is an administrative measure of employees, while the employment rate comes from a survey that also includes self-employment. The two usually move together and right now they are not, which could reflect a shift between the two categories or something else. The divergence is real. The explanation is not established, and I am not going to invent one.

Why a freeze is harder to job hunt in than a downturn

This is the part that gets missed, because a freeze sounds milder than a recession and in aggregate it is.

In a downturn there is churn. Companies cut, then they rehire, roles reopen, people move, and a vacancy you missed in March exists again in June. Painful, and full of motion you can act on.

In a freeze there is no motion. Fewer roles are advertised, and the people already in jobs stay where they are because there is nowhere to go. So the internal ladder stops moving too, which means the backfill that would have opened a rung below never happens.

For anyone searching, that shows up as a specific and demoralising pattern. Not rejection. Silence, and a shortlist of roles that stops refreshing.

What actually changes in your approach

Application volume is the wrong lever. When advertised demand is at a twelve-year low, sending more applications into the same thin pool does not find roles that are not there. It just costs you the hours.

Listing age matters more than it does in a normal market. In a thin market, an old posting is more likely to be a role nobody has filled, and more likely to be one that was quietly abandoned. Those look identical on a job board and they are opposite bets, so check the date and check whether the same role has been reposted.

Sector beats aggregate, always. A national vacancy figure is a sum. Some sectors are still posting and some have effectively stopped, and the aggregate tells you nothing about which you are applying into.

The unadvertised route is worth more than usual. Not as networking advice, as arithmetic. If advertised vacancies are at a twelve-year low and employee numbers are falling through non-replacement, then a larger share of what hiring exists is happening without a posting. A direct approach to a person is a worse use of an hour when there are 1.3 million vacancies. It is a better one now.

One thing about reading UK figures

Every number above is a rolling three-month average, not a monthly print. "May to July 2026" is a single figure covering a quarter.

That matters twice. It means a UK figure cannot be compared directly against a US monthly one, which is built on a single reference week. And it means a UK figure moves slowly by construction, so a genuinely bad month gets smoothed into two neighbors and shows up late.

The next release is 20 October. Nothing in the picture above is likely to change much by then, which is itself the point: a freeze is a slow thing, and the advice to wait for the market to turn has been wrong for long enough to stop taking it.

Screen harder, apply less, and go direct more often than feels natural. That is what a 702,000 vacancy market asks for.