UK infrastructure is the one part of construction still growing, and the gap between it and everything else widened through the first half of 2026. Output is forecast to rise 3.2% this year and again in 2027, while private housing falls by around 10%. That divergence is the most important fact in this market, and almost nobody is reading it correctly.
The reason is the sentiment data. The construction PMI has sat in deep contraction all year, and the civil engineering sub-index hit 22.1 in June — its worst reading since the first Covid lockdown. Taken alone it reads like collapse. The hard data points the other way: ONS output up 1.6% in the three months to May, a third consecutive rise; £19.8bn of contract awards in H1 with Q2 running 77% ahead of the same quarter last year; CECA members reporting their strongest order books in three quarters, a net 60% expecting workloads to rise, and over half planning to grow headcount. Firms do not hire into a market they think is dying.
This report explains the contradiction. What the PMI is measuring is the wait — consented, funded work that has not yet reached site. The bottleneck is conversion speed, not demand, and we set out the three things that break it before the year end: AMP8 mobilising in volume, the DCO reforms that came into force on 24 July, and main works starting on Lower Thames Crossing in November.
For anyone hiring, the conclusion is uncomfortable. The constraint on delivery is no longer money — CITB puts the requirement at 41,200 additional workers a year to 2030, and over half of CECA members are already short of skilled operatives. Water, energy transmission, nuclear and major roads will be drawing on the same project, commercial and planning population from Q4, and AMP8 mobilisation lands within weeks of Lower Thames Crossing. If you have approved headcount for the autumn, the searches want running now rather than into the middle of that.