‘‘Sense of community belonging’’ in health surveys: What social capital is it measuring? Richard M. Carpiano a,n , Perry W. Hystad b a Department of Sociology, University of British Columbia, 6303 NW Marine Drive, Vancouver, BC, Canada V6T1Z1 b School of Population and Public Health, University of British Columbia, Canada article info Article history: Received 20 August 2010 Received in revised form 9 December 2010 Accepted 9 December 2010 Available online 11 January 2011 Keywords: Community belonging Social capital Social networks Neighborhoods Self-reported health abstract Canadian national health surveys regularly ask respondents to rate their sense of belonging to their local community. Health studies commonly use this question as a social capital indicator, but what social capital domains community belonging is measuring remains unclear. Analyzing Canadian General Social Survey data, we evaluate the validity of this measure with respect to network-based social capital and health. Results indicate that sense of community belonging is associated positively with several network-based social capital measures. Neighborhood network-based social capital most substantially reduced associations between sense of community belonging and health measures, but results differed by urban and rural settings. These findings indicate the need for public health surveys to include specific measures of respondents’ networks. & 2011 Elsevier Ltd. All rights reserved. 1. Introduction The health implications of community social connections, although having long been a research focus (Leighton, 1959), has witnessed renewed interest over the past two decades. This resurgence can be credited in significant part to the concept of ‘‘social capital’’ as a popular focus within public and population health (Moore et al., 2005, 2006). Social capital, the resources embedded within social networks, considers the actual or potential material, informational, and affective resources that individuals have access to via social networks and that may be used as a means to pursue individual or collective action (Bourdieu, 1986; Coleman, 1988; Portes, 1998). As noted by Moore et al. (2006), health research on social capital has not typically focused on these network-based ele- ments, and has instead often relied upon attitudinal measures of perceived interpersonal trust and reciprocity to serve as proxy indicators of real-life social relationships and social capital. Nevertheless, an increasing number of health researchers have utilized this network approach, focusing on (a) general social capital, that is, social capital that either exists within individuals’ social networks in general (e.g. Moore et al., 2009a,b) or that emanates via their membership in groups and organizations (e.g., Rose, 2000; Veenstra, 2005; Harpham, 2008)—and (b) geographi- cally bounded social capital that exists within particular locations (e.g., neighborhoods or larger local geographic areas) (e.g. Bhattacharya, 2005; Boneham and Sixsmith, 2006; Carpiano, 2007, 2008; Friedman et al., 2007). Network-based approaches to social capital, which have a long theoretical and empirical tradition in the social sciences, are crucial in studying social determinants of health. From a ‘‘fundamental causes’’ theoretical perspective, which focuses on access to resources as a key mechanism for maintaining health, a network-based approach to social capital enables a more precise examination of people’s differential access to resources that may promote or harm health, thereby, helping us to better understand how social structures and policies shape health inequalities (see Carpiano et al., 2008). Coinciding with this renewed interest in community social ties, public health surveys in Canada have, since at least 2000 (Shields, 2008), included a single question that asks a respondent to describe the level of her/his ‘‘sense of belonging to her/his local community.’’ To date, a number of published studies using these nationally based survey data have reported that this single item is positively and significantly associated with a variety of health outcomes, including better general health (Ross, 2002; Wister and Wanless, 2007; Shields, 2008) and mental health (Shields, 2008; Romans et al., 2010), as well as increased odds of undertaking a change in a variety of health behaviors (Hystad and Carpiano, 2010). From the perspective of theory, these studies are motivated by prior health research on social relationships and social capital and their findings are consistent with other research that utilizes more specific measures of actual social network ties (House et al., 1988). Some authors have even explicitly stated that they are using this sense of community belonging item as either a measure of social capital (Wister and Wanless, 2007; Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/healthplace Health & Place 1353-8292/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.healthplace.2010.12.018 n Corresponding author. Tel.: + 1 604 822 3845; fax: + 1 604 822 6161. E-mail address: carpiano@interchange.ubc.ca (R.M. Carpiano). Health & Place 17 (2011) 606–617