Part 1 of a propviews investigation into the assumptions underpinning the latest London Plan consultation
There are two ways you can read the draft London plan published last week. The first is as an ambitious pitch to the new Burnham administration that social housing needs a vast injection of cash. London is ready and waiting with the sites, need and capacity to get things going.
The second way you can read it is as a continuation and recalibration of the land value capture thesis. The Islington school of thought failed the first time because they tied themselves to brownfield only and then got caught out with various macro shocks. By enlarging the Capital’s game board with huge tracts of greenbelt land, they give themselves another shot at proving the thesis.
For this new draft, published on 16 July 2026, contains a target of 558,451 new homes over the next decade. Of those, 56,000 — approximately ten percent — are to come from Green Belt release across eight outer London boroughs: Barking and Dagenham, Barnet, Croydon, Enfield, Havering, Hillingdon, Redbridge and Enfield.
The greenbelt gamble
The greenbelt gamble – and it is a gamble, rests on the belief that it is through virgin land that you can embed affordable housing and gain control of place making. This is a tweaking the old Parkhurst Islington defence that if you flatten land values, you can drive up affordability.
The assumption embedded in the Plan is that you can achieve a higher percentage affordable and so lift the Plan numbers as a whole which have up until now been let down by pesky brownfield complexity.
The viability case for Green Belt release rests on a single assumption that appears in paragraph 4.146 of the viability study (the LPVS) which underpins the document.
“A nil premium above EUV has been assumed within the greenbelt viability testing as a starting point as any premium above EUV can only be realised through compliance with the NPPF’s Golden Rules and the policies of the Draft London Plan.”
Nil premium. No uplift above existing use value. The benchmark land value for the Green Belt sites underpinning 56,000 homes is set at the agricultural value of the land — and not a pound more.
What EUV+ Is — And What Nil Premium Means
The Existing Use Value plus premium methodology — EUV+ — is the now standard approach to establishing benchmark land value in planning viability assessments. The logic is straightforward: a landowner will not sell for less than the value of their existing use, plus a premium sufficient to incentivise them to transact rather than stay.
As readers of propviews will know, the GLA over corrected after Parkhurst by tightening EUV to an inch of its life and extinguishing alternative use value. The result was the slow death of the land housing supply market and a retrenchment to public land and contractor plus deals. It has created regulatory capture and decimated SMEs.
The new draft continues the trend. Rigid EUV+ is cast across the land market as the basis for modelling the plan assumptions. I am yet to find a single private site that has been sold for a premium of 20% however, I won’t relitigate that point here.
For all site typologies in the LPVS, the GLA applies this 20% premium above EUV to test if the plan works. Light industrial sites, commercial premises, surface car parks — all receive 20% above their existing use value as the minimum benchmark.
For Green Belt land — agricultural fields, grassland, amenity open space — the GLA has applied zero percent above EUV. The BLV equals the agricultural value of the land. Nothing more.
Annex 5 of the viability study sets out the precise figures. For pure greenfield Green Belt land — the baseline typology — the EUV is £37,000 per hectare, equivalent to approximately £15,000 per acre. The premium is zero. The BLV equals the EUV.
On the largest strategic typology — GB_SS_RES_1, a 14-hectare site delivering 1,438 homes — the BLV is £518,000 for the entire site. That is £360 per plot.
The second strategic typology — GB_SS_RES_2, a 15-hectare site delivering 2,224 homes — has a BLV of £555,000. That is £249 per plot.
To put those figures in context: an industrial open storage site in Value Band B — a redundant light industrial unit in outer south London — is valued at £6,742,857 per hectare plus a 20% premium. The Green Belt greenfield site is valued at £37,000 per hectare with zero premium. The industrial site attracts a benchmark land value 219 times higher than the Green Belt. And both are expected to deliver housing.
The ratio of BLV to completed sale value on the Green Belt strategic sites is approximately 0.07%. The land is being valued, for plan-making purposes, at less than one tenth of one percent of the value of the homes that will be built on it.
The Justification — And Why It Fails
The GLA’s reasoning for the nil premium is stated in paragraph 4.146: “any premium above EUV can only be realised through compliance with the NPPF’s Golden Rules and the policies of the Draft London Plan.”
The argument is that Green Belt landowners should not receive a premium above agricultural value because they have not yet complied with the Golden Rules — the NPPF requirements that unlock Green Belt development, including 50% affordable housing and infrastructure delivery. Until they comply, they get agricultural EUV. Once they comply, the residual land value generated by the scheme provides the uplift.
This is theoretically coherent as a land value capture argument. But it has two fatal flaws that the viability study itself acknowledges.
Flaw one: the residual land values are not generated by the development — they are generated by public subsidy.
Paragraph 4.146 continues: “Viability testing of greenbelt development typologies indicates that residual land values reflecting emerging policy requirements can generate significant uplifts above EUV; however, the residual land values themselves are largely generated by public subsidy interventions.”
The GLA is saying: yes, there will be uplifts above agricultural EUV — but only because the government is pumping in grant funding, low-cost loans, CIL in-kind contributions, and public sector-led delivery mechanisms. Remove the subsidy and the residual collapses back to near-agricultural EUV. The landowner premium is not created by the development. It is created by the taxpayer.
Flaw two: the day one values don’t support policy requirements even at nil BLV.
Paragraph 1.36 of the executive summary states: “The viability analysis highlights overall that current (‘Day one’) residential development values alone cannot viably support NPPF/Draft London Plan policy requirements, as well as meeting the costs of enabling infrastructure and transport improvements.”
Day one values do not work even when the BLV is nil. The scheme requires both a nil BLV and substantial public subsidy to clear viability. And in Value Band F — the lowest value outer London locations where much of London’s Green Belt sits — paragraph 11.19 confirms: “there are no viable scenarios even with grant funding and CIL in-kind.”
No viable scenarios. With grant. With nil BLV. In the locations where Green Belt land is most abundant.
The Hope Value Problem
The GLA’s nil premium argument assumes landowners will wait for compliance with the Golden Rules before expecting any premium above agricultural EUV. There is a word for what landowners actually do when their land is designated for development in a statutory plan. That word is not “wait.”
The NPPF and the PPG are explicit that the EUV should disregard hope value. They are equally explicit that the landowner premium must “provide a reasonable incentive for the landowner to sell.” A nil premium on Green Belt land provides no incentive whatsoever. It is the minimum the landowner receives if they do nothing and keep farming. The plan cannot acquire this land voluntarily at nil premium.
The Compulsory Purchase Answer
The GLA’s answer to the landowner problem is compulsory purchase. Paragraphs 1.63-1.65 make this explicit: public sector-led delivery models, Mayoral Development Orders, Local Development Order masterplan consents, and strategic land assembly are “critical to unlocking housing delivery on strategic sites.”
CPO is a legitimate tool. But it is not a fast one, a cheap one, or a guaranteed one. A CPO process for a 14-hectare strategic Green Belt site involves:
– Designation in an adopted Local Plan — which requires the borough to cooperate, which already Croydon has indicated it will not
– Public inquiry into the CPO
– Legal challenge by the landowner on the compensation amount
– Resolution of compensation disputes through the Upper Tribunal (Lands Chamber)
– Actual acquisition, funding and site preparation
At the most optimistic, this might take 5 years from designation to acquisition. The draft London Plan runs to 2037. The 56,000 Green Belt homes need to be delivered within that window. The CPO process alone, in the most cooperative boroughs with the most willing landowners, cannot be completed at the scale required within the plan period..
The Political Headache
Enfield has already withdrawn from the New Towns programme. Croydon’s leader has publicly announced opposition to Green Belt release. The May 2026 local elections have installed a Reform administrations in Havering. The eight boroughs identified for Green Belt release are not uniformly enthusiastic about delivering it.
The planning inspector examining the draft London Plan will ask a simple question: is the Green Belt target deliverable? The answer in the viability study — no viable scenarios in Value Band F with grant; viability only through public sector-led delivery and nil premium BLV — will be tested against the political reality that at least three of the eight target boroughs have already signalled opposition before the consultation has formally opened.
What the Numbers Say
The viability study is not fraudulent. It is internally consistent. It says: if you apply nil premium BLV, assume substantial grant funding, assume CIL in-kind contributions, assume public sector-led delivery models, assume transport investment that synthesises value over time, and assume that somehow the land can be acquired at agricultural EUV — then the scheme is viable in Value Band D and marginally viable in Value Band E.
Every one of those assumptions is heroic. Together they are not a viability case. They are a wish list dressed in Argus Developer output.
The nil premium BLV is the foundational assumption on which the entire Green Belt viability case rests. It is lower than the premium applied to a redundant car park in inner London. It is lower than the premium applied to a derelict industrial unit in Outer London. It is the value at which no rational landowner sells voluntarily. And it is only defensible in a plan-making context because the NPPG allows plan-makers to disregard hope value — not because the market will.
London has come far too late to the party on Green Belt release. The Broad Locations for Growth are the right policy direction. But a plan that assumes 56,000 homes on the basis of nil premium acquisition, with no funded CPO programme, no committed transport investment, luke warm to uncooperative boroughs, and no viable scenarios in the lowest value Green Belt areas even with grant — is not a housing delivery plan. It is a housing trajectory that flatters the target and defers the problem.
In part 2, propviews will explore the GLA sensitivity modelling in depth and test whether it survives reality.
