SERIE RESEARCH mEmORIMDn

This paper aims at providing an operational framework for housing market research by means of disaggregate choice analysis. First, a concise survey of spatial choice and interaction models is presented, starting from a systematic typological approach. Then the advantages of disaggregate spatial choice models (especially probit models) are pointed out. Next, the paper presents extensively empirical results obtained from the application of a disaggregate residential choice model based on a probit analysis to the Dutch housing market. The analysis is based on a multi-step procedure, viz. the assessment of the prior propensity to move, followed by the assessment of the (conditional) probability for an actual move into a specific dwelling type. Brief attention is also paid here to spatial differences in moving behaviour. 1. INTRODUCTION The housing market is a multi-faceted and heterogeneous market characterized among other things by: complex search and choice processes of (potential and actual) movers, diversity in information content regarding supply and demand on various sub-markets, drastic shifts in looational behaviour due to structural economie changes, strong impacts of social and geographical spillover effects (e.g., segmenta-tion, environmental quality), occurrence of disequilibria on various sub-markets (due to inertia in public decision-making, socioeconomic and demographie changes, strong land use competition, etc), and high degree of public and institutional interference. In the past, a wide variety of housing market models has been developed and applied in order to investigate the abovementioned The present paper deals mainly with one aspect of the housing market: it aims at providing an operational framework for disaggregate residential choice models for housing market analysis. By way of introduction, first a brief typology for spatial choice models will be