It is self evident that the pattern and scale of housing investment shapes the spatial structure and quality of housing within a city. A great deal of work in urban geography, land economics and economics has drawn attention to the ways in which new construction, especially at the suburban edge, develops and generates feedback effects (movement, succession, etc.) in the rest of the housing stock. However, we have a much less clear understanding of the dynamics of existing neighbourhoods and, in particular, the factors which generate disinvestment and decline or reinvestment and upgrading (see Grigsby, Baratz and Maclennan, 1986).