Australia, like most developed countries, has an ageing population. South Australia (SA) has a higher percentage of older people than any other Australian State, with 14.5% of the population aged 65 or over (Australian Bureau of Statistics (2001a) and Australian Bureau of Statistics (2001b)), and this figure is expected to increase to 24% by 2051 (ABS, 1999a). Women, particularly those living alone, represent a higher proportion of the older population and this proportion increases with age due to longer life expectancy than men (Australian Bureau of Statistics (1999b) and Australian Bureau of Statistics (2002)).
Over the past two decades, much research has sought to examine how health outcomes vary according to where one lives (Ellaway, Macintyre, & Kearns, 2001; Macintyre, Ellaway, & Cummins, 2002; Sooman & Macintyre, 1995). This research examined how characteristics associated with place of residence might affect the health of inhabitants, irrespective of characteristics of the individuals themselves. Quantitative understanding of such 'area effects' on health have predominantly been addressed using the analytical strategy of multi-level modelling (Diez Roux, 2003a). This body of research has been pivotal in investigating how residential environments might affect individual risk for cardiovascular disease, for example (Diez Roux, 2003b; Diez Roux, Merkin, Hannan, Jacobs, & Kiefe, 2003). Some commentators have, however, pointed out, “characteristics of places are typically distilled in this type of analysis to a few limited variables..place is frequently considered a black box (of variable sizes and shapes) in which unidentified “non-individual” processes take place” (Tunstall, Shaw, & Dorling, 2004, p. 6). While such approaches have advanced our understanding of how place of residence can affect health outcomes, they do not (nor do they aim to) address the complexities which underlie individual relationships with neighbourhood environments and how these might impact on health. Exploration of the connection between place and health through accounts and perceptions of those living in the communities studied (Campbell & Gillies, 2001; Cattell, 2001; Schulz & Lempert, 2004; Sixsmith & Boneham, 2003) is adding to the empirical evidence linking the social and built environment with health.
Emerging from such research is the notion that social capital—mutually beneficial social relationships between citizens characterised by interpersonal trust and norms of reciprocity (Kawachi, 1999; Putnam, 1995)—might be an important factor for health. It has been suggested that the extent of social capital in an area might explain differences in overall levels of health (Mohan & Mohan, 2002). Research by Kawachi (1997) and Kawachi and Kennedy (1999) suggested that living in areas characterised by higher 'stocks' of social capital is associated with lower mortality rates, and that disinvestment in social capital or breakdowns in social cohesion can have both indirect and direct effects on health. The exact mechanisms through which social capital influences population health, and indeed how the notion itself should be conceptualised, remains the subject of considerable scrutiny (Hawe & Sheill, 2000; Mohan & Mohan, ...