Wednesday, 17 August 2011

Location matters: putting people first in planning

Posted by Tim Leunig, SERC, LSE and CentreForum

I recently visited Bordon, a small town in North East Hampshire. It was one of Labour’s proposed eco-towns. The Conservatives opposed this in opposition, but are now keen. The proposed 5500 houses would double the town’s size.

I talked to a leading light of Bordon Area Action Group. He was a NIMBY, opposing all new housing. He favoured sending immigrants home (adding that this was not BAAG policy). He had many objections to housing – including that more people walking on local countryside footpaths would damage invertebrates’ habitats.

I am not mocking him. He was knowledgeable, and his views were heartfelt. He is correct on many things. He feels that the Tories nationally lied, and that the council – which covers a big area – have slanted consultations to get the answers they want. He likes his town the way it is.

I think we need much more housing in the South East. But not in Bordon. This has nothing to do with lizards, and everything to do with people. Put simply, people don’t want to live in Bordon if they can live elsewhere nearby.

We can see this in local house prices. The cheapest 3 bedroom house in Bordon is £140,000. The equivalent in the neighbouring towns of Alton, Liss, Liphook and Farnham are £188,000, £195,000, £215,000 and £224,000 respectively. People generally prefer to live in other towns in the area.

The other towns also have railway stations, and are on dual carriageway roads. They are better connected, and better placed to prosper in the future. It makes sense to build there, rather than in Bordon.

Both the previous and current government chose Bordon because the army is moving out, releasing brownfield land that can be redeveloped. Everyone says that building on brownfield land is a “Good Thing”.

This is not true unless the brownfield land is where people want to live. We should also consider land restoration, and building elsewhere.

I found that a house costs £80,000 more in Farnham than in Bordon. If we can restore the Borden army site for less than £80,000 per house space, we should do that, and build in Farnham.

Davis Langton [pdf] suggest that restoration costs about £1.75m per hectare, or around £60,000 per house space. These costs are very site specific. They assume, for example, 1.9 tonnes of waste per square metre of the site. This colossal figure is for dense inner city sites: the figure for a former army camp such as Bordon will be far lower.

Many of the costs have to be paid anyway, whether the land is used for housing, or returned to green space. If that is so, we might as well build the houses where people want to live.

Whether Bordon is the right place for development requires accurate, site specific estimates of the cost of remediating the land, for both development and greenfield uses. But the figures here make it overwhelmingly likely that cost-benefit analysis would not support redeveloping the Bordon army site for housing. Land auctions, an idea I've developed here would offer a sensible way through the decision-making process.

On thing is for sure, developing the Bordon army base for housing simply because the War Department decided it was a good place for army training in 1863 is not good spatial economics or good planning.

* Tim Leunig is a SERC Affiliate. He is a Reader in Economic History at LSE, and Chief Economist at CentreForum think thank.


andrewlainton said...

This is a weak argument as Whitehill currently has poor road and no rail links, but the ecotown proposal will pay for the upgrading of the A 325 and the repoening of the rail link - as proposed by ATOC

To my mind your case perfectly shows the fallacy of the crude land auction model for site allocation. It takes into account current and not future land values following construction of infrastructure.

It is better to do a cba on the basis of capturing current low land values and then taxing the uplift - the classic ebezor howard model. May do such a calculation/cash flow analysis for this site on my blog showing how the land auction model leads to a misallocation of resources

Tim Leunig said...

Community Land Auctions allow the council to capture future land values. If the cost of putting in a rail link raises the land value by more than the cost of doing so, then land auctions give the council an incentive to install the rail link and capture the value.

In the case of Bordon, however, it is highly unlikely that the town will get a rail link if the houses are built. This is both the reality that the government has run out of money, and because of the underlying economics.

Passengers numbers are not sufficient for a new line to make sense. (Clearly you can pay consultants to say the reverse, but since neither side is paying me, I can look at the evidence).

Nearby Alton currently generates 2000 journeys a day (1000 in, 1000 out). Current Bordon residents are less prone to using trains - they have chosen to live in trainless Bordon, after all. So even with the expansion it is unlikely that Bordon will generate more journeys that Alton. If the proposed service is half hourly, 6am-11pm, then arithmetically 29 people will board a typical train leaving Bordon. The remaining 499 seats will be empty. It is not worth building a new train line for so few people.

Furthermore, there are no additional slots to Waterloo. So every Bordon train is one fewer train to Basingstoke, Alton, Portsmouth, Bournemouth and so on. Replacing well-used train services with less-well-used services would not be an easy sell. The alternative would be to run the Bordon train as a shuttle to Bentley, connecting with the Alton train. That would make the service much less popular with commuters, and therefore reduce the number of people using it.