Nobody is being sacked. The door is just opening more slowly, and that is harder to notice
Corrected 15 September 2026. This piece was published with the model disclosure "claude-opus-4-8" under its byline. It was written by claude-opus-5. The wrong string was copied from an earlier edition rather than checked. The disclosure is now correct. The error was ours.
The clearest fact in this morning's tax records is not about who lost a job. It is about who did not get one.
In the year to July 2026, the number of people leaving payrolled employment rose 4.5 per cent. The number entering rose 0.7 per cent. In absolute terms, 596,306 people left in July and 576,987 arrived.
Around 570,000 to 600,000 people move out of payrolled work in Britain in a typical month, and a similar number move in. That churn is normal and it always has been. What has changed is small, and it compounds.
Why this is difficult to feel
A redundancy is an event. It has a date, a letter, a conversation, sometimes a collection. It is legible to the person it happens to and to everyone around them, and it produces a number that a government can be asked about.
A job that is not advertised produces nothing. There is no moment. Nobody is told. The person who would have been hired does not know they were not hired, because the vacancy never existed for them to apply to. They experience it as a slightly longer search, then a slightly longer one, and eventually as a story about themselves.
This is the central asymmetry in how people read their own working lives. Events get attributed to the world. Non-events get attributed to the self. A person made redundant in a plant closure knows what happened to them. A person who sends out ninety applications over eight months and hears back from four does not usually conclude that the inflow rate fell by a few percentage points. They conclude something about their CV, their degree, their interview manner, themselves.
What the shape of the data does not tell you
The lead is careful, and I will be too. We do not know whether the 25 to 34 fall is fewer jobs or fewer people, and the piece alongside this one explains why that cannot yet be settled.
But the flows are not ambiguous in the same way. Outflows exceeding inflows in ten of the last twelve months is a statement about hiring, and hiring is a decision somebody makes.
What follows from that is modest and worth saying anyway. If you are in the middle of a long search at the moment, the arithmetic of the thing you are inside has changed, whatever its cause turns out to be. That is not reassurance and it is not meant as any. It is the difference between a fact about a labour market and a fact about you, and people are consistently bad at telling those apart when the news arrives as silence rather than as a letter.
The one group visibly doing well
Payrolled employees aged 65 and over rose 4.9 per cent over the year, the largest proportional increase of any age band, and their median pay rose 5.9 per cent, the fastest of any group.
There are several possible readings of that and I am not going to pick one from a single table. It is worth noticing that in a year when the payroll shrank, the part of it that grew fastest was people past the traditional retirement age, and the part that shrank most was people in their late twenties and early thirties.
What does a labour market that stops hiring feel like to the person inside it?