Does UK construction cut training first, and pay for it ten years later?.
A belief, not yet a finding. Here is the argument, the five questions that would prove or break it, and the public data each one is being tested against.
I look for people with ten to twenty years' experience, and they are not there in the numbers they should be. My explanation is this. Every time the industry hits a downturn, it cuts training first. The people who would have started that year never start. They do not come back when the work does, and about ten years later the gap resurfaces as a shortage of experienced people, the ones every contractor is trying to hire.
If that is right, the 1990s recession and the years from 2008 to 2011 each left a hole, the people who would be in their mid thirties to forties now are thinner on the ground than they should be, and the dips in 2020 and 2022 mean the next hole opens in the mid 2030s.
That is a belief from the search side, not a proven fact, and it is not CITB's position or the government's. So it is being tested against public data, question by question, as part of research now under way in an academic environment. An honest “the data cannot say” counts as an answer, and if the data says I am wrong, this page will say so.
Five questions, one for each link in the argument
1. Does training intake fall faster than employment in a downturn?
It tests: The industry cuts training first.
Against: Apprenticeship starts in construction (DfE, from 2010/11) against construction employment (ONS JOBS02 and EMP14), through 2008 to 2011 and 2020. Ratios, not levels: both fall in a recession, so the test is whether starts fall faster and recover slower.
If yes: The mechanism is real. If no: Training falls in line with jobs, and the shortage has another cause.
2. Is there a dip in workers aged roughly 33 to 40 today?
It tests: The people who were never trained after 2008 are still missing.
Against: CITB's evidence on the workforce's age profile, and the Labour Force Survey, which is the only direct series of age by industry.
If yes: The gap from 2008 to 2012 never refilled. If no: Other entrants or migration filled it, and the ten-year lag does not hold.
3. How many people does the industry actually need?
It tests: Nobody has put the headline numbers on one basis.
Against: CITB, the UK Trade Skills Index and the other published estimates, reconciled to the same definition of the workforce and the same period.
If yes: One number that can be quoted with its definition. If no: Also an answer: the sources measure different things and cannot be reconciled.
4. Which occupations carry the biggest gap for their size?
It tests: The shortage is concentrated, not general.
Against: CITB's occupational tables and ONS vacancy figures.
If yes: It shows where employers and government should point their effort. If no: The shortage is general, which is a different argument.
5. When does the next gap open, and in which trades?
It tests: The pattern predicts.
Against: The training dips of 2020 and 2022, carried forward on the same lag.
If yes: A dated forecast of the next shortage, which nobody has published. If no: The pattern is too weak to forecast from.
The shortage is quoted as one number. It is several.
- 41,200 extra workers a year, about 206,000 by 2030. CITB's Construction Workforce Outlook for 2026 to 2030: net additional workers across a workforce of 2,606,380 in 2025.
- 47,860 a year, the year before. CITB's 2025 edition, for 2025 to 2029. Each edition rebases, so the two are not like for like, but the headline has come down.
- 937,000 new recruits by 2032. The UK Trade Skills Index 2023: a gross ten-year figure for trades, including replacing everyone who leaves. Reported as “nearly a million”.
- 900,000 tradespeople. Research for Kingfisher by Cebr used it for the size of the trades workforce itself. The same number, meaning the opposite thing.
Net or gross, five years or ten, trades or the whole industry: the honest range runs from about 200,000 to nearly a million depending on who is counting and what. Until the numbers are on one basis, none of them should be quoted on its own.
One more figure cuts against the easy story. Construction apprenticeship starts in England were 24,590 in 2024/25, up from 24,230 the year before, the first rise since 2021/22, as reported from the Department for Education's figures. Employers tell me they turn away far more applicants than they take on. If young people want in and cannot get in, “nobody wants to work in construction” is not the whole answer.
Does a missing cohort matter if older workers stay on longer and younger ones come through faster? If the people at either end can cover the middle, the gap is real but the shortage is not. That question is being tested alongside mine, and it deserves the same honest answer.
The research is ongoing within an academic environment, and the answers come in over the winter. Each one goes on this page as it lands, with its chart and the definition it was measured on, whichever way it falls. The full picture goes into a paper in spring 2027 for employers, industry bodies and government.
If you hold data or experience that bears on any of the five questions, intake and leavers by year, the age profile of your own workforce, the occupations you cannot fill, I would like to hear from you: scott@lechley.com. Nothing you send is published with your name on it without your agreement.
More on who is hiring at the top, and where the demand is going: Insights.