A mid-decade census will be held in 2028, at an estimated cost of £140m, in response to what ministers described as “an unprecedented divergence between projection and observation” in rural areas.
Statisticians have welcomed it without reservation. The current estimates for eleven council areas rest on a base that predates the population change entirely, and every downstream calculation inherits that error.
The timing problem
Field work runs through 2028. First outputs are expected in mid-2029, with small-area data following into 2030.
The replacement national planning framework is scheduled for adoption in early 2029.
It will therefore be adopted approximately six months before the data commissioned to correct it becomes available, using the projections everyone agrees are wrong. Officials have confirmed the framework will be “reviewed in light of” the census outputs once they arrive.
“We are spending £140m to find out what we already broadly know, six months after the decision it was supposed to inform. I still think we should do it. I would simply have started in 2025.”
A demographer
The field force
One operational difficulty has been raised early and has a familiar shape.
A census requires several thousand field staff, deployed heavily in exactly the areas where population change has been sharpest. Those are the areas with no available accommodation, which is why the health board cannot recruit, why the school cannot house four teachers, and why the planning department is running at 61% vacancies.
The census office has begun discussions about temporary accommodation, and has approached, among others, the operators of a 40,000-pitch campervan aire.
The census
- Field work: 2028
- First outputs: mid-2029
- Small-area data: into 2030
- Planning framework adoption: early 2029
- Cost: ~£140m
- Field staff required: several thousand, mostly where there is nowhere to stay
