Saturday, 1 March 2014

Do economists not read Christaller?


Andrew Lilico a very good and highly respected economist has written a piece about geography. I know he thinks he's writing about economics but he's wrong - and that's why the article is poor.

Here's the gist of the argument:

We could simply refuse to grant permission to build more houses or more factories or more offices in the higher population-density regions. Over the past two decades, for all the complaints about planning restrictions, the surpluses of dwellings over households in London and the South East has risen – we've built properties faster than the population of households has risen. We could, instead, build fewer houses. That wouldn't immediately prevent the population in these regions from growing, because existing properties could be used with more dwellers.  But it would result in property prices in London and the South East rising even faster, relative to the rest of the population, than they have done recently. The consequence would be, eventually, that it would become economically unattractive to live and work in the high-population density regions. That would drive more population and more business activity into lower-population-density regions.

I guess that, for an economist, it all makes sense - indeed Andrew Lilico points out that this is simply the price mechanism used to drive a regional policy. And this is where Walter Christaller comes in - where the matter becomes a question of geography rather than economics. Perhaps it's Walter's active involvement in the Nazi Party that puts economists off him or maybe his love of central planning.

Christaller developed a thing called central place theory:

Walter Christaller, a German geographer, originally proposed the Central Place Theory (CPT) in 1933 (trans. 1966). Christaller was studying the urban settlements in Southern Germany and advanced this theory as a means of understanding how urban settlements evolve and are spaced out in relation to each other. The question Christaller posed in his landmark book was "Are there rules that determine the size, number and distribution of towns?" He attempted to answer this question through a theory of central places that incorporated nodes and links in an idealistic situation.

The point is that successful (large) places are at the nexus of other settlements. That success is less a function of (as many earlier geographers argued) the location of resources but a consequence of trade, exchange and interaction. Andrew Lilico assumes that the economic success of London is contingent on financial investment rather than human interaction - it is the building of houses, offices and factories that precedes economic activity rather than these things (as Christaller's theory predicts) being a consequence of human activity.

Just as importantly, Andrew Lilico also assumes that immigration is to fulfil a specific demand for labour rather than movement to a place where the opportunity for higher wages exists (typically from rural to urban). In modern economies where, in the main, resource location is irrelevant - thus we get a development of CPT (from another German geographer called Losch):

Losch started from the "bottom" of the model by considering one "equivalent customer" or one unit of consumption and build up from there. In the Losch model, the ten smallest market areas, each with a different k-value are plotted with each network surrounding a central place. These networks were then laid over each other and positioned to produce the largest number of places for each k-value. This model produced wedges of city-rich and city-poor areas spread out around a major central place.

The real problem with Lilico's idea is not only that urban economic activity largely precedes the building of houses, offices and factories (especially in the modern economy) but also that, if you prevent investment in London, you cannot assume that this investment will automatically transfer to Newcastle or Belfast - it's as likely to go to Amsterdam, Indianapolis or Jakarta.

The 'empty places' Lilico refers to in Scotland, in Northern Ireland and in in the North of England are empty because they lack economic value. And they lack that value because of geography.


No comments: