City Hall tested its own plan. Most schemes failed.
The Urban Forecast
Of 989 development scenarios modelled for the draft London Plan, 23 per cent were viable. The number comes from City Hall's own testing, not from a developer lobby.
That provenance is the point. Viability numbers in London usually arrive from whichever side stands to gain, and are discounted accordingly. This one is the Mayor's own evidence base for the Mayor's own plan.
The qualifier matters as much as the figure. The 23 per cent is before market-realistic assumptions are applied. It is the optimistic reading, not the pessimistic one.
Nick Cuff has set out what that does to policy in practice. If the plan's own testing shows most schemes failing, then the policy requiring viability transparency stops being a tidy procedural requirement and becomes a mechanism for re-litigating the plan itself, application by application. Every scheme arrives at committee with a defensible argument that it is one of the 77 per cent.
Set that beside what happened in September. Lambeth approved Vauxhall Square at 25 per cent affordable on a four to three vote, with no viability appraisal submitted because the Mayor's fast-track route does not require one. Southwark refused Berkeley's attempt to cut Borough Triangle to 10 per cent, in a case where the applicant and the council's own assessor disagreed by £207m on the size of the deficit while agreeing there was one.
Same policy context. Same fortnight. Opposite outcomes. The variable was not viability, because viability is now contested in every case. It was where the offer sat relative to a threshold.
Our read: a plan whose own modelling says three quarters of schemes do not work is not a plan with a viability annexe problem. It is a plan with a viability problem.