A simple affordable housing tariff? It’s complicated.

The London Housing Commission has produced an interesting document reviewing the causes of  London’s housing crisis and attempting to set out a clear programme for how the next mayor, the 33 London boroughs and central government should work together to tackle it. They speculate that the shrinking share of affordable housing provided through planning gain has been the result of the increased emphasis on development viability in planning negotiations. Their solution to the issue of ‘negotiable viability’ was to suggest a

 …tariff in order to set a fixed affordable housing contribution for developments in London. Applying a tariff is not necessarily easy or uncontroversial, on account of differing land values across the capital, which means that in some places a tariff will underprovide for affordable housing in some areas, and in others bring into question the viability of the housing development. A great deal of care needs to be taken therefore to apply a tariff that is fixed, workable and fair, and to ensure that the delivery of affordable housing and market housing is not compromised in the process…On balance, if these conditions are met, the commission is in favor of applying a planning tariff in London, so long as it is achieved after in-depth consultation.

This is something that I’ve been thinking about every now and then for a while. How do you set the tariff?

If we see the tariff as a tax, we might want one that was progressive and simple and which minimised any negative effect on housing supply. One advantage of the current system is that, in principle, it attempts to customise the tariff to the individual project through project-specific viability appraisals.  In practice, the uncertainty in the inputs required for calculation facilitates opportunistic behaviour by some developers, creates uncertainty and introduces an element of chance.

Also, implicit in any tariff would be some strong assumptions about the allocation of the land value uplifts that a planning permission produces. Broadly, it would be expected that the higher the property values in an area, the higher will be the land values.  Land value capture is what is really going on here and the higher the land value, the more there is to capture for affordable housing.  It should be possible to link an affordable housing tariff to property values – which can be reasonably easily observed.

I could envisage an approach where it might be possible to create an affordable housing ‘charging schedule’. For example,

Development value (£ psm) Plot ratio Affordable housing requirement
£1000 – £2000 1.00 0%
£2000 -£2999 1.00 10%
£3000 – £3999 1.00 20%
£4000 – £4999 1.00 30%
>£5000 1.00 40%


Alternatively it would be fairly straightforward to create a simple algorithm that would provide indicative affordable housing requirements based on a small number of easily obtained inputs. This would really be just a simplified viability appraisal that stops all the games and focuses on typical or average estimates.

The calculations could be based upon a simple formula.

Cash residual surplus available for affordable housing


Total development revenues psm of site area


Total development costs psm of site area + Return to land owner + Return to developer

The estimated cash residual surplus for affordable housing could be then converted to an equivalent affordable housing requirement.

It’s maybe best to work through an example. Let’s take an area of central London where

  • Residential values are estimated to be £5000 psm.
  • For a plot ratio of 3.00, this generates revenue of £15,000 psm of site area.
  • Development costs (including construction, professional fees, taxes, agents’ fees, CIL, additional s106 costs etc) are estimated to be £3000 psm. This generates normal development costs £9,000 psm of site area.
  • Let’s assume that the land owner should get 20% of the development value as a return = £3000 psm of site area. This is 20% of the development value scheme assuming no affordable housing. (Obviously this is a crucial assumption.)
  • Let’s assume that the developer should get 10% of the development value as a normal return = £1500 psm of site area. Again, this is 10% of the development value scheme assuming no affordable housing. (This is another key assumption)

The leaves a surplus of £1500 psm of site area available for affordable housing. We can convert this into a percentage of affordable housing space.

  • Let’s assume that Registered Providers will pay on average 60% of the Market Value of affordable housing.

By my calculations, the surplus of £1500 is equivalent to a 25% affordable housing tariff assuming an average 40% discount on market levels.

Obviously, in order to make these type of calculations, I had to make some important assumptions – particularly about profit and return to the landowner. 20% of the GDV assuming no affordable housing is fairly arbitrary.  You also have to live with broad averages.  If you want a simple system, then to some extent simple models are needed.  There are other weaknesses.

  • This approach would only work with fairly short-term projects (maximum of three years). Long term, phased projects with large abnormal costs are likely to be best handled by bespoke viability appraisals.
  • If we go for bands, there are cliff-edge problems. Clearly, at the margin developers would try to demonstrate that they should be in a lower band.
  • In very high value areas, such as Westminster, very high proportions of affordable housing become viable given that the return to the land is ‘suppressed’ at 20% of GDV. However, affordable housing requirements could be capped.
  • Some sites will have abnormal development costs that need to be taken into account. This should be possible to incorporate into an algorithm but reduces simplicity.
  • It would also be necessary to ensure that the Alternative Use Value was taken into account.

So…it seems to me that justifying a ‘simple’ tariff system is quite a wicked problem. A single tariff would be regressive and could also prevent desirable development in marginal sites. However, a tapered tariff requires generalisations about typical costs and revenues and judgements about how land value uplifts should be distributed. I think that a simple algorithm is feasible and would eliminate a lot of the games that are currently being played with viability appraisals.  For simple tariffs or for customised viability appraisals, the key is to establish clear decision rules about appropriate returns to land owners and developers.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s