A local housing market model with spatial interaction and land-use planning controls

There have been few documented attempts to construct and estimate econometric models of local housing market dynamics and linkages. Previous housing market analyses have been undertaken primarily at the national and regional levels of aggregation. This paper reports the results of the estimation of a system of equations using a panel dataset. The work extends previous housing market analyses in a number of respects. First, the analysis is undertaken at the local level. Local authority (former district council) boundaries are used as a proxy for local housing market areas. Second, the modelling approach takes explicit account of spatial interaction between local land and housing markets. The system constructed and estimated includes equations for private household migration between local housing markets, house price growth and house building output. The supply equation (house building output) is driven partly by measures of land made available through the planning system for house building. This includes measures of the relative importance of urban infill and brownfield sites in the supply of land. Our estimations have shown that these variables are statistically significant. The equations are estimated using two stage least squares on panel data. The spatial coverage of the data is the central belt of Scotland. The data span the period 1986-1997.