Put Very Simply – The Flaw in the Government’s OAN Method – and How Ireland Shows Us How to Simply Fix It

We have covered OAN in great depth and very technically on this site – I fear from the recent government’s response to the consultation which battered every point aside without explanation that it was over some people heads.

So lets put it very simply. The government’s method takes a demographic baseline from ONS household projections.  It then recognises that this is too low as it does not translate from households to homes, second homes, frictional vacancies etc. are not assessed.  So it takes the figure ‘experts’ have recommended as the national need (300,000, now out of date more like 340,000) and rounds it up with a formula based on how unaffordable housing is in the area.

The problem with this is it mixes two things up.  One is the correction to the baseline (households to homes) the other is correcting for suppressed household formation where house building in the past hasn’t kept up, so reducing the demographic baseline as they are fewer existing households to spawn new ones.

This creates a problem.  Lets say unaffordability  increases everywhere but more in the south rather than north – as has happened. then under the government’s method the North would get less housing and the South more, indeed under their method  there would be a net transfer of housing from South to North.  This would be exacerbated by the governments decision to make the figures ‘backward looking’ in terms of past employment trends rather than forward looking in terms of employment potential.  This hits hard on the North.

There was a good reason for this as recommended by LPEG.  Not everywhere can have ‘above average’ employment growth.  there is a risk of double counting in terms of over ambitious LEP based projections.

Also the affordability based weighting, can have perverse effects.  For example can you justify Berks having a 30% uplift and Central Beds 100%?  Which seems down to a previous cheap area having a housing boom as it was one of the few cheaper areas within a commute of London.  Or can you really justify Oxfordshire having a reduction (which seems solely down to a low demographic baseline)?

I think the correct approach is:

Step One – Correct the demographic baseline with a proper household to homes conversion

Step Two – Correct for the suppressed household formation for the recession, but in a way that does not create a low housebuilding-low household formation feedback loop.

This compromise approach has been suggested by Savills as opposed to the huge global fudge factor in the original Nat Litch method for LPEG.  What you do is correct for the recessionary suppression of household formation but weight this element in terms of unaffordability.  What this does is normalise for this element nationally breaking the feedback loop.

This then creates a ‘corrected demographic baseline’.

Now you have to project this forward.  You could do a trend projection.  But I rather like the approach adopted in Ireland’s new NPF which uses an econometric model showing the growth potential of second tier cities, Dublin expands but second tier cities expand more.  Why not do this with the growth ambitions of Northern powerhouse etc. using Cambridge Econometrics or Oxford Economics Regional Models.

Step Three:  Project Forward using consist Regional Economic Models and balance internal migration accordingly.

This should include a feedback loop recognising that areas that bulld homes more will grown more and attract more work based in migrants.   It should also assume the gradual building out of the backlog of home over 15 years or so.

There you have it.

The problem is that it will still mean large increases in London and the South East (Green Belt) areas.  The government’s approach was to have a ‘cap’ on the proportionate increase   Which would have created a systematic underprovision of housing.  I suggest tehre should be a cap, but that shortfalls from the cap should be redistributed to all LPAs (capped and uncapped) based on a GIS based method on how unconstrained they are, with redistribution’s only occurring within commuting distance of major cities.   This ‘fund’ would not automatically go on OAN, rather LPAs under the DTC would be required to demonstrate how they would contribute to meeting it jointly.  Something that would encourage the new wave of joint statutory strategic plans.