Britain’s housing crisis is deepening

professor glen bramleyBritain’s housing crisis appears to be deepening, and attracting widespread comment in the media. Professor Glen Bramley, Director of IHURER, explains what policies could promote greater housing supply.

Britain has a serious housing shortage. My colleague Colin Jones blogged on this topic a few weeks ago. This week I had the opportunity to speak at a seminar in Bristol, sponsored by the South West Observatory, on planning for new housing, sharing a platform with among others the Minister for Planning and the newly elected Mayor of Bristol.

Bristol and the South West of England generally are in the front line of this hot issue. As someone who previously lived in Bristol for 21 years and has undertaken a number of studies of housing markets and housing needs in this region, I feel moved to offer some comments. You can find my presentation here.

In my opening contribution to this Policy Blog I wrote about ‘evidence based policy’. The commitment of the present Coalition Government to an evidence-based approach to policy is, shall we say, a bit variable. However, I would make the observation that the planning system is structured and premised on the principle of basing plans on evidence, not least in the contentious area of new housing supply. The recently-revised National Planning Policy Framework (NPPF) for England attracted a lot of anxious commentary about opening the floodgates to ill-conceived developments through its ‘presumption in favour of sustainable development’. But the revised document is very clear that the remedy lies in the hands of local authorities. They need to have in place a Local Plan which provides enough land for new housing development which meets the need and demand for new housing, backed up by sound evidence of that need and demand.

The Government appears to recognise that there is a problem of housing supply and that better planning is needed to tackle this. Unfortunately their first major step on entering government, foreshadowed before the 2010 election, made the problem a whole lot worse. They scrapped the regional tier of planning and associated ‘top down’ targets and handed the decisions on how much land to provide to the local authorities and (in some circumstances) local communities. While doubtless electorally popular in some quarters, this hasty move to ‘Localism’ in planning confronts them with a major obstacle to increasing housing supply. Public sentiment across most of England is quite strongly opposed to new housing development in the local area- NIMBYism is widespread and respectable.

The 2010 British Social Attitudes Survey showed that opponents of housing development outnumbered supporters in the ratio 3:2 and this pattern applied in most areas, particularly those parts of southern England where the need for more housing is most pressing. The survey provided some glimmers of hope that more people might be persuaded to support development if they were confident that it would bring in its train certain improvements to local infrastructure, and if the type of housing was more suitable (e.g. affordable for new households). Putting a reasonable interpretation on these data, and linking with other evidence including 2010 voting patterns, I developed predictions about what would happen to local plans for housing across all the local authorities in England. The overall conclusion was still pessimistic, in suggesting that the number of authorities cutting their planned housing numbers would be almost double the number increasing them, and that the reductions would be concentrated in the South of England.

Recently, the consultancy Tetlow King have updated their survey of changes in planned housing numbers to late 2012. Of the three-quarters of authorities which have decided, the pattern of change is quite similar to that which I predicted. Although there are a few more cases of increases in the south than I had estimated, the number of reductions is even greater in certain key regions, notably the South West and also the West Midlands.

My analysis of existing planning stances towards housing and likely/emerging changes shows that these patterns are generally perverse, in the sense that policies are more restrictive and changes are more negative in areas which already have very poor affordability and higher levels of need. They are also perverse in the sense of being more restrictive in areas which have (on evidence of trends over the last couple of decades) the greater potential for economic growth. These conclusions also echo the arguments of the Centre for Cities in their contribution to the conference.

The Bristol sub-region is actually a striking example of this. It has some of the worst affordability indicators outside the London area, and it had some of the highest increase in jobs in the period 1997-2007. Yet in 2010 these local authorities reduced their already inadequate plans for new housing by about a third. They reversed previous decisions to support several significant urban extensions to the Bristol conurbation, primarily on grounds of wanting to preserve existing ‘Green Belt’ boundaries. One of the authorities involved is currently subject to a High Court challenge on this.

Public sentiment across most of England is quite strongly opposed to new housing development in the local area
Public sentiment across most of England is quite strongly opposed to new housing development in the local area

It was interesting to share the platform with the minister on this occasion. He pointed out, quite rightly, that I had not tried standing for election on this issue of building more housing. It was noted that the dominant group voting in local elections were comfortable, middle aged, middle class home owners who were not directly affected by the adverse effects of the failure to build housing. There was no disagreement about the outcomes we were trying to achieve. He observed that he found himself as minister in a position where he could pull various levers but they did not necessarily have much effect on the ground. I was too polite to point out that part of the reason for that was that his boss Mr Pickles and predecessors Grant Shapps and Greg Clark had deliberately sawn through the rods and wires connecting national policy and local plans when they scrapped regional planning and housing targets and embraced localism in 2010.He ended on the point of making what amounted to a moral appeal to local authorities and communities to acknowledge the needs of younger people to get a chance to access the housing market. A number of us present felt that this was tantamount to admitting that he recognised that the current policy set-up was likely to fail.

Leaving aside the politics, there are also issues here about the nature of evidence. Traditionally planning relied on a mainly demographic approach, centred around household projections. This is a useful starting point but has its limitations, because ultimately trend-based projections tend to build in the suppression of household numbers by inadequate supply, a form of ‘circularity’. The Barker Review of Housing Supply in 2004 introduced the concept of treating affordability as a key outcome to be targeted through planning. Local housing need and market analyses pay increasing attention to affordability, and there is plenty of evidence on this. I would argue that attention should also be paid to indicators of unmet housing needs, such as overcrowding and concealed households, which national surveys show to be deteriorating, alongside measures of homelessness and waiting lists (see our recent Homelessness Monitor for Crisis. I would also argue that assessments of housing requirements should pay attention to prospective growth in employment, both in order to support economic growth and in order to promote more sustainable patterns of commuting.

There is also an evidence challenge which goes beyond the range of types of measures. Because planning is about future needs, the most appropriate evidence would arguably be forecasts of future demand, supply and unmet need. Indicators of current needs and problems provide a useful starting point, but are not conclusive; present problems may resolve over the market cycle, or they may progressively intensify. Forecasts are more demanding, particularly in a complex and potentially volatile market like housing. It is quite common to use projections of future need/demand, generally based around the household projections and assumed trends or levels for key indicators like affordability or the tenure shares. But almost by definition, these approaches cannot deal with discontinuities, changes in trends, major cyclical disturbances or regime changes. Robust forecasting models for housing at the appropriate sub-regional scale are still in their infancy, unlike the situation in certain other sectors like transport.

For our sins we have tried to develop sub-regional models to forecast housing markets, affordability and housing need. We do not claim that our models are the last word on the subject; quite the reverse. The models are quite complex and require quite a lot of data input (ideally key numbers going back over quite a long time period). Some may criticise them for being ‘black boxes’, although I would refute that in respect of the models I have developed, which sit in a quite transparent form in a familiar spreadsheet setting. We have recently had the interesting experience of establishing a working model, whose architecture is based on a sub-regional model developed for Gloucestershire, in a different country (New Zealand). These models are particularly useful for showing the interdependence between the situation in one area and what is happening in other surrounding areas and at national level.

The people in the front line of deciding whether local plans are ‘sound’ in terms of their interpretation of the evidence on need are the planning inspectors (‘PINS’). I am not sure what skills and training they have in the area of modelling and forecasting, but my impression is that they are more comfortable in the traditional territory of household projections than in the newer world of economic models.

Those opposed to planning for more new housing have taken perverse comfort from the effects of the Credit Crunch, Global Financial Crisis and resulting recession. This ‘proves’ that the previous housing boom was due to lax bank lending rather than an over-tight housing supply. With housebuilding running at half of its previous level, there is no immediate pressure to release more land. With developers sitting on record numbers of outstanding planning permissions, it is easy to blame them. The fact that this is turning into the longest recession/depression on record does not help, in enabling these debates and diversions to continue indefinitely. However, we are beginning to see some revival in mortgage lending and housing market transactions, and house prices are rising at least in London and parts of the south.

While regional planning may have fallen out of favour, there is close to a professional consensus that the right area to plan housing provision for is that of the ‘housing market area’, and these geographies are generally sub-regional in scale and larger than individual local authorities. This does not entirely square with the official mantra of localism. It is very clear from my modelling work that local housing markets are very open, and that the effects of new housing supply on house prices and affordability are not only quite spread out over time but also diffused over a wider geographical area. This means that an individual local authority has a limited incentive to release more land for new housing, even if they are motivated ‘morally’ to ‘do the right thing’. If the other local authorities in the surrounding areas do not act in a similar way, the benefits in terms of affordability will be small. I have written in various papers about this as a sort of planning version of ‘the prisoners’ dilemma’.

The Government recognises this to a degree in that the Localism Bill imposes a ‘duty to co-operate’. The way that this is interpreted, by Local Authorities, PINS and Ministers, should be watched with close interest.

I concluded my remarks to the Bristol gathering with a few modest policy suggestions, relevant to the general mission of trying to promote greater housing supply.

  • Smarter incentives – there are now various incentives in the system (‘New Homes Bonus’) but I would argue that these should be bigger, but more targeted on places which needed an increase and were willing to make a quantum increase (above a threshold); such a super-bonus should be  conditional on sub-regional cooperation.
  • Urban extensions are very often the most sustainable way of providing large increments to housing supply ( because they can tie in with existing infrastructure and transport routes, and reduce the length of travel and car dependency), but this often requires redrawing of Green Belt boundaries. Traditional ‘polo mint’ traditional Green Belts designed in the 1940s are not fit for purpose in the 21st century. There is more biodiversity in many surburban gardens than in typical agricultural ‘monoculture’ fields. I am all for permanent protection of the best green landscapes, but that is not what a lot of Green Belts are.
  • Maintain section 106 planning agreement mechanisms for affordable housing we need a lot of affordable housing and this is the main way to subsidise it, when government has little or no money to offer for this purpose.
  • Land development agencies – getting developers to build is like pushing string; you can give them permission, but they do not have to take this up and can determine the rate of buildout.I would establish in key growth areas  a public/private agency to bring land forward and auction it to builders on license, requiring them to deliver prescribed minimum numbers each year.  I was interested to hear the new elected mayor of Bristol George Ferguson talking about his plans for ‘Property Board’ which may, through pooling public and private land resources, play a role of this kind.
  • PINS need to apply more rigorous approach to assessment of housing requirements, as argued above.