Deriving Supply-side Variables to Extend Geodemographic Classification

The traditional proprietary geodemographic information systems that are on the market today use well-established methodologies. Demographic indicators are selected as a proxy for affluence and are then often linked to customer databases to derive a measure of the level of consumption expected from the different area typologies. However, these systems ignore fundamental relationships in the retail market by focusing upon demand characteristics in a ‘vacuum’ and ignore the supply side and consumer-supplier interaction. This paper argues that there may be considerable advantages to including supply-side indicators within geodemographic systems. Whilst the term ‘supply’ in this context might imply the number of consumer services already in an area, equally important for understanding demand are variables such as the supply of jobs and houses. We suggest that profiling an area in terms of its labour market characteristics gives a better insight into the income chain while the supply of houses could be argued to be a crucial factor in household formation that in turn will impact upon demographic structure. Using the regional example of Yorkshire and Humberside in northern England, we indicate how a suite of supply-side variables relating to the labour market can be assembled and used alongside a suite of demand variables to generate a new area classification. Spatial interaction models are calibrated to derive some of the variables that take into account zonal self-containment and catchment size.