The starting point for the discussion of farmland-use is the standard von Thunen model of land allocation on a homogeneous plain (Capello, 2007; Edwards, 2007). Von Thunen developed his model of the spatial distribution of agriculture more than 200 years ago from observations on the distribution of activities on farmland surrounding towns in southern Germany. His ideas have become the basis for much of the literature on the spatial distribution of activity, but until recently there have been few attempts to develop analytical support for his observation.
To take an example, suppose a city, which serves as the market point for all agricultural production, is surrounded by a homogeneous plain − how will the land be allocated? Von Thunen showed that, with positive transportation costs and different net returns per hectare for agricultural output, the most likely outcome is a series of concentric rings, with bands of production within them (Figure A A.1.). Commodities with higher unit returns and high unit transport costs will occupy the inner rings and commodities with low returns and low transport costs will be further out. The interesting aspect of von Thunen’s analysis is the clean edge between land uses. At some point, increasing transport costs overwhelm higher net returns per unit of land from one commodity over another and land switches to the next lower net return commodity, because its transport costs are low enough to just offset its lower selling price. The result is a sharp edge between rings. Production ceases at the extensive margin where transport costs fully exhaust the net return for the commodity produced in the final ring. The analysis shows that a change in either relative net returns per area or in transport costs will alter the transition zones, but in ways that preserve precise edges.
While von Thunen himself is silent on the issue, a similar process must determine the boundary between the city and the highest return from an agricultural commodity. At some distance from the place of work, the cost of a house, including the cost of travel to work, must increase to the level where it no longer makes sense to build houses. At this point agriculture begins, assuming that housing is a more valuable land use than agriculture (Capello, 2007). Specific and localised types of farming, such as vineyards near Dijon, glass-houses near Amsterdam and thoroughbred farms near Lexington, are able to effectively compete with housing for land, but this is not the case for high-volume agricultural commodities.
As von Thunen noted, if conditions change, the boundaries should also change. An increase in urban population or an increase in urban income that drives up demand for housing should lead to an outward movement of the urban land zone and a rippling effect through other agricultural zone boundaries. Conversely, an increase in the price of food or a decrease in transportation costs for various farm commodities has a more ambiguous effect. While there is some likelihood that the zones of agricultural production might shift, there is less likelihood that the boundary delimiting the urban fringe would move significantly.