I thought I understood the draft NPPF definition of affordable housing as 80% of market values. I thought at first this was ridiculous and assumed it was existing market values.
However reading the draft NPPG revisions alongside it together with the recent Parkhurst Road caselaw I think I was wrong.
However I think we still have a compcated return to circular valuation in cases where affordable housing is not grant funded.
The important thing about the new method (or rather the Islaington.GLA Three Dragons approach applied nationally) is that land value os an output rather than an input after applying all policies. So developers can never claim a scheme is unviable if a developer has paid too much for land.
It can be laid out as a formula – with a few reasonable assumption
Market Value =((EUV PMH*1.2)+(Build costs MH)*1.2) + ((EUV(1-PMH)*1.2)+(Build costs AH)*1.06)-IN
Where PMH = proportion of site market housing
AH = affordable housing
In=infrastructure S106 costs +CIL
Where I assume that build costs includes all costs to a developer (including promotion) other than land.
The guidance doesn’t give a standard return to landowners (unlike developers) however in a market economy the return on capital will equalise with return on land which is exactly the same assumption of caselaw (Shifield principle).
if you fix the number of units and infrastructure costs then there are two dependent variables, the proportion of market housing and market value. The infrastructure effectively draws a horizontal line across the graph which sets the intercept point – it fully comes off the price of the land.
However as the proportion of market houses rises the price falls. Remember the policies including affordable housing are inputs not outputs.
Indeed it is possible to rearrange the formula into its components price of market housing and price of affordable housing.
In the simplest case then of a 100% funded affordable housing scheme the market price becomes a premium on the affordable housing value depending on the Homes England formula.
For most schemes with a mix the the exact premium will depend site by site on infrastructure costs plus the relative cost of affordable and market building, but it seems to me in plan making the assumption is that such a ‘affordable housing price * X’ approach would be used.
How is land value captured at this point? Prices for land to developers is squeezed, however housebuilders can then still sell at market values. If land is released but developers still sell at market values then we have a dual market situation where housebuilders can buy cheap and arbitrage by selling expensively, or where landowners wont release unless they do joint ventures with developers to equalise returns. . Indeed land owners could make far more money by developing it themselves.
This is not really a tenable situation compared to the continental system whereby the local state acquires the site at EUV+ and captures the uplift., or SVT in some form.
What if I am wrong on the formula and the market housing price is an input – well then we are back in ‘circularity’ territory.
This puts us in interesting territory. If plan policies are not viable the solution might be in many cases to increase the proportion of affordable housing to force down prices.