What is an efficient land use
Critics of suburban sprawl often claim that the land being encroached upon would have been more valuable as prime agricultural land. It is a interesting point, the amount of good agricultural land we have is limited, as is the amount of land available around cities. By a twist of fate, land good for agriculture -- flat, near a constant water supply -- is often also good for cities -- accessible and near a constant water supply. I'd argue therefore, that your attitude to sprawl across agricultural land depends in part on what model you view the building of houses: as production, or as consumption.
A model as production views a house as a producer of shelter for its inhabitants. One where the good produced is consumed entirely by its workers, who often own the means of production as well. Compared to an agricultural use, shelter is far and away superior in economic terms. Rent on a half dozen properties on an acre of land will earn a dozen times that of any farm produce: given the choice, farmers have chosen to subdivide every time.
Is this bad? Or rather, would this land have been better used as agriculture and another, less fertile piece of land, used as housing? A parallel can be seen here with the concept of trade. Here, our good land, P, is superior in two respects to another piece of land, Q. P is better than Q as a piece of farmland and also better than Q as housing -- you can assume, because it was subdivided first, that it is closer, reducing travel expenditure and infrastructure requirements. Suppose the economic returns were as follows:
P-h = 100, P-f = 10, Q-h = 80, Q-f = 2
You can see that even though P might be five times more productive as farmland, and only marginally better as housing, the total economic return is greater with P as housing, and Q as farmland. Using this model, conversion of agricultural land to houses makes sense. Assuming people have taken into account all the costs and benefits 1, housing on prime agricultural land is a better use of the land than the same houses on poorer land.
However, a model of consumption takes a different view. It views housing as the consumption of a product, consuming resources and generating waste. In this model, housing is an essential resource, not a generator of an economic good. Hence, land should be consumed at the most economic rate, leaving other land uses to generate the economic profits. In this model Q is a better place for housing than P, because it consumes farmland of little economic value. Better still, the houses built should be at higher densities to reduce the consumption requirements.
It is the model of consumption that most planners follow, leading to policies for growth boundaries, higher densities and public transport over automobiles. But it is also the model of the average suburban resident, who see higher densities as infringing on their right to consume the available land with large houses and spacious backyards.
This attitude has led to two rather strange complaints. Planners rail against suburban sprawl on economics grounds when the economics are still very much in sprawl's favour 2. Meanwhile, residents in the suburbs complain about higher densities being imposed by planners when planners do no such thing, and in fact, are merely facilitators for developers trying to increase the economic returns on that land.3 Both groups should probably reassess their position.
 They haven't. It doesn't make it wrong however, just not as good as it might seem today.
 And it will continue until it isn't. Planners would do better to consider the implicit subsidies they provide sprawl and consider how to manage it from that viewpoint.
 There are arguments against higher desnsities and proper planning of what these densities would be. But these reasons are generally secondary to rampant NIMBYism that has the indirect effect of raising house prices across the board (ie. turning their house into a protected industry).
19th August, 2005 21:52:20