Martin Wolf adds to the continuing debate on the Green belt in the FT 5th Feb ‘There should be a presumption of development in green belts, unless the cost of new infrastructure exceeds the benefits.’ Usefully he does not take the crude position from pamphlets of the IEA and Policy Exchange that the Green Belt should be abolished or ‘rolled back’ irrespective of location, or worse the counter extreme from Eric Pickles that it should be considered inviolate. The debate has moved on in recent months with contributions from a amongst other the London Society, consultants Nathanial Litchfield and the Adam Smith institute asking if there are specific limited areas in the Green Belt that should be deleted and developed for housing and if so where?
Martin adds a useful potential test but seems unaware that the Green Belt was first proposed in part because the costs of infrastructure at the edges of metropolitan areas already exceeded the benefits. Take London just after WWII. The planned underground extension north of Edgware to Elstree and Bushey Heath was abandoned, and you can still see sheds and trackbed. This was because Frank Pick of London Underground calculated the trains would be full before they reached inner London and so the gains from revenue from the northern line extension would never compensate from the loss of revenue on the existing network (1). Certainly new infrastructure can fix it, but for a major city this means the ‘threshold costs’ are often prohibitory as you have to double and triple up on networks (2). This was used by the pioneers of the Green Belt and Regional Planning such as Sir Patrick Abercrombie to justify a Green Belt around London with New Towns/Garden Cities outside London taking the overspill. The trouble is this necessary counter to the restrictions of the Green Belt was abandoned, leaving the Green Belt as a purely negative policy of restriction rather than as a positive tool of economic policy (Even David Cameron has acknowledged this in a speech, in 2012, though he has done everything to exacerbate the asymmetry through abolishing regional planning).
Contrary to public perception the planning and legal justification for Green Belt has nothing to do with its landscape amenity. Its absence isn’t even a ‘material planning consideration’ for loss of Green Belt in the jargon. It is a strategic policy designed to prevent urban sprawl and coalescence. The test should be whether this sprawl has costs to society that exceed the benefits.
Certainly we are building too few homes, the planning system has contributed to this, and this can add to excessive credit growth and financial instability. The liberalisation of the planning system through the NPPF led first a slump to housebuilding to the lowest level since the 1920s as many planned growth areas were deleted by councils through ‘localism’; and then a raucous counter attack by housebuilders who through ‘planning by appeal’ have built up huge stocks of land where it least makes sense – where it is most expensive to service and least accessible to jobs, in commuter villages beyond the Green Belt. Some villages now planned to increase in size by 30%, 50% or more. Some rural areas are in open revolt, like in the Prime Ministers own constituency, led by Peter Jay.
Yet the major urban areas in England continue to grow in populations at a far greater rate than what detailed and independently verified studies have shown can be realistically built inside their Green Belt’s inner rings. In recent months independent inspectors at Birmingham and London have ruled that 100,000s of houses need to ‘overspill’. Studies are planned or in progress as to where this might go, but without any mechanism to agree a distribution and the planning minister Brandon Lewis this week has said that must agree it amongst themselves (translation fight it out like rats in a sack and not come to an agreement, simply delaying the hard choices that must be made till well after the General Election).
The Green Belt grew enormously in the 70s and 80s as counties quite deliberately adopted policies to shift housing to their neighbours. Certainly intelligent review of those proportionately limited areas where it makes sense should and is being conducted in many areas. Sadly this is not at all driven by any kind of analysis of infrastructure costs and benefits. Loss of green field (and 70% of green field is not Green Belt) is inevitable. England needs over 3 million homes over the next 15 years and even the CPRE’s figures suggest Brownfield sites could take 1-1.5 million, even if they were all in the right place (those in the South East can take about 1 years of the nation’s needs). Sites around commuter stations in the London Green Belt seem attractive, but census evidence shows an 80-90% transport split in favour of the car over public transport even in the best connected by rail commuter towns like St Albans. The risk then of an uncoordinated roll back of the Green Belt without well thought out new transport networks will simply clog up roads in the most congested areas.
Martin is right that a Henry George type tax on uplift in land values can help fund infrastructure and avoid land being left vacant, and there are globally many successful examples. The disbenefits of sprawl on the environment (cost of carbon) and on cost of infrastructure almost never factored into the scribblings of ‘market urbanists’ in the United States or at the LSE or University of Reading. By contrast ‘Smart Urbanists’ study smart growth – where and how to grow cities that makes environmental and economic sense, and unmitigated sprawl is rarely the answer. My own research has shown that certainly within the M25 there are very few areas where Green Belt loss would not cause major problems. I estimate this at around 72,400 dwellings, not much bigger than Brentwood and only making a tiny dent in London’s housing needs.
It is clear than that the vast majority of overspill will need to be built outside the M25, some carefully planned and confident strategic Green Belt releases will make sense, but there is a trade off the closer to London the more they will clog already crowded commuter networks, whilst further away (As Milton Keynes and Cambridge show) cities can gain critical mass to develop their own employment and if large enough new Garden Cities could justify their own public transit networks. If any pundit suggests a solution that meets one policy objective (housing) but breaches others (carbon reduction targets) they don’t have a workable solution.
There is a method in economics to make sense of this. As cities grow their economies of scale (agglomeration) rise with population, but the diseconomies of scale (infrastructure) rise with size. So at some point at some density a city reaches its optimum size, and taxing land value uplifts cannot compensate for these losses. At which point the New Economic Geography (which won Krugman a Nobel prize) predicts it makes optimum sense in growth terms to develop a large new city well away from the old one. I call this the broken city model., it explains why the growth of some cities such as Jakarta, Moscow, and Mexico City comes to a halt as they become so large it becomes prohibitively expensive to fund the infrastructure to support their growth.
Large cities need to grow, but the economic lesson is not always that they should get bigger. What is needed is proper studies of what infrastructure investments make economic sense and where these can support the housing we so badly needs. How to do this is sadly out of fashion – larger than local planning.
- Hall, Peter(1984). “Chapter 2: Challenges and Responses”. In Sutcliffe, Anthony. Metropolis 1890–1940. Mansell. ISBN 978-0-7201-1616-8. Note: Though also the creation of the Green Belt also has an influence in the final abandonment of the ‘Northern Heights’ plan in 1950.
- TMaliaz, Boleslaw (1972) Threshold Analysis as a Tool in Urban and Regional Planning. Eleventh European Congress of the Regional Science Association.
Sent to the Ft to consider for publication.