Comparing the causal mechanisms underlying housing networks over time and space

This paper concerns the ontology and epistemology of comparative research, which are the foundations of a research strategy. It builds upon a Realist ontology (Bhaskar, 1975, 1979; Sayer, 1984,1985, 2000; Lawson, 1997, 1999) of necessary and contingent relations forming causal mechanisms, in order to extend existing frameworks for explaining difference and change in housing networks. Housing networks are perceived as interactive, open systems of agents bound by underlying necessary relations, subject to ever-present contingent relations, their own individual agency and path-dependent behaviour. Unique clusters of necessary and contingent relations underlie interacting realms of the housing network, which generate tendencies for certain tenure divisions, property rights, systems of credit provision, housing production and urban form. Specifically, the paper concretises these concepts and defines causal ‘clusters’ through the use of illustrative examples. It is argued that explanatory research requires the comparison and contrast of these clusters over time and space, rather than only key housing events or experiences of them. Laborious, concrete, historical research is required to reach sophisticated, multi-causal explanations for the differences between and changes within housing networks. Such a process should be informed by an explicit ontological and epistemological ‘package’. A Realist-retroductive research strategy aims to identify and contrast the unique causal clusters of social relations underlying housing outcomes, which exist in the context of specific local contingencies. Amongst others, significant clustering of social relations surrounding the supply of land for residential development, system of credit provision for credit for housing production and welfare relations influencing the consumption of housing services. These relations have a combined influence upon investment, production, consumption and exchange of housing services. Uniquely defined, these clusters of relations promote divergent housing tenures and urban forms. To test and revise this postulated model, two case studies are proposed in the Netherlands and Australia.