Elsevier

Social Science & Medicine

Volume 85, May 2013, Pages 32-37
Social Science & Medicine

Individual-level social capital and self-rated health in Japan: An application of the Resource Generator

https://doi.org/10.1016/j.socscimed.2013.02.027Get rights and content

Abstract

Despite accumulating evidence of associations between social capital and health in public health research, a criticism of the field has been that researchers have exclusively focused on concepts of social cohesion to the exclusion of individual-level approaches. In the present study, we evaluated the association between social capital measured by the Resource Generator (an individual-level assessment of access to social capital) and self-rated health among Japanese population in a cross-sectional study. A postal survey of 4000 randomly selected residents in Okayama City (western Japan) was conducted in February 2009. We divided the overall scores from the Resource Generator Japan scale into quartiles. Odds ratios (ORs) and 95% confidence intervals (CIs) for self-rated health were calculated separately by sex. Individuals with the highest quartile of scores had significantly lower odds of poor health compared to the lowest group after covariate adjustment among both men and women (men; OR: 0.45, 95% CI: 0.24–0.86, women; OR: 0.44, 95% CI: 0.25–0.79, respectively) and there were also significant dose–response relationships. In the sub-domains of Resource Generator Japan scale, a differential pattern was observed by sex. Women showed a clear dose–response relationship with health across all four sub-scales (domestic resources, expert advice, personal skills, and problem solving resources). In contrast, only the domain of expert advice exhibited a strong association with men's health. Among both men and women individual-level social capital measured by the Resource Generator was related to reduced odds of poor health even after taking into account individual confounders. Although we cannot exclude reverse causation due to the cross-sectional design, our study adds to the accumulating evidence of the potential utility of the Resource Generator for evaluating the relationship between individual-level access to social capital and health.

Highlights

► Research on social capital has been focused mainly on measuring social cohesion. ► We adapted the Resource Generator – a network-based approach – to Japanese context. ► Individual social capital was associated with a protective effect on poor health. ► We found gender differences in the associations of sub-scales of Resource Generator. ► Our findings provide further evidence of the utility of the Resource Generator.

Introduction

The public health literature on “social capital” has overwhelmingly focused on social cohesion – i.e., community norms of trust or reciprocity – as a determinant of population health (Kawachi, Kim, Coutts, & Subramanian, 2004). With this approach, researchers have sought to demonstrate the health benefits (and health penalties) accruing to individuals as a consequence of their membership of close-knit groups. Conceptualized in this manner, social capital is treated as an ecological or group characteristic, and multi-level analysis has been typically used to identify the contextual influence of social capital on health. Indicators of social capital in this manner include moral resources such as trust and reciprocity that the individual can access through membership of the collective.

On the other hand, the methodology of individual social capital based on social network approach has been hampered by offering diverse measures (e.g., social network structure, presence of specific alters, size and diversity of networks, and exchange of social resources) that tap into different dimensions of social capital (Flap, 1999; Lin, 2001; Van der Gaag & Webber, 2008). Of these, we now focus on social resources as the indicators of individual social capital in the present study. To date, three main measurement instruments for individual social capital have been described and applied in public health research: Name Generator (McCallister et al., 1978; Marsden, 1987), Position Generator (Lin et al., 1986, 2001), and Resource Generator (Snijders, 1999; Van der Gaag & Snijders, 2005).

Lin's Position Generator is perhaps the most frequently used survey instrument for the measurement of individual social capital (Van der Gaag, Snijders, & Flap, 2008). The advantage of the Position Generator (which it shares with the Resource Generator) is that the instrument does not require mapping the whole social network and therefore saves time. The Position Generator inquires about each respondent's connection to others who hold “valued” occupations such as doctor, banker, lawyer, and college admissions officer (Van der Gaag & Webber, 2008). In other words, the Position Generator is specifically designed for the measurement of access to social resources useful in instrumental actions, i.e., “getting ahead” in life. The Position Generator is then used to construct network resource indexes such as heterogeneity (akin to the concept of bridging social capital), “upper reachability” (akin to the concept of linking social capital) (Lin et al., 2001). The limitation of the Position Generator is the list of prestigious occupations with embedded “value” is likely to be: a) culturally contingent (for example, knowing a lawyer is not as critical in Japan compared to the United States which tends to be more litigious); and b) situation-specific (e.g., it is more important to know a plumber than a surgeon if your toilet happens to break down on a weekend). Moreover, the Position Generator is not explicit about the types of resources that are accessed through social connections; it is simply assumed that being acquainted with others who hold valued occupations leads to the ability to access information, prestige or other instrumental resources (Van der Gaag & Webber, 2008).

In contrast to the Position Generator, the Resource Generator adopts a checklist approach to inquire about the specific resources that respondents have access to through their networks (Van der Gaag & Webber, 2008). Therefore its main advantage over the Position Generator seems to be that it inquires about specific domains of domestic resources – viz. expert advice, personal skills, and problem solving resources – which may have relevance for health promotion. A limitation of the Resource Generator is the kind of resources/favors one commonly ask of friends tends to vary by culture.

A previous study reported that presence of common mental disorder (12 item General Health Questionnaire scores ≥5) was inversely associated with social capital as measured by the Resource Generator in a UK sample (Webber & Huxley, 2007). On the other hand, a six-month prospective cohort study among 193 individuals with depression in UK reported that the Resource Generator had no independent effect on the course of depression measured with the Hospital Anxiety and Depression scale (Webber, Huxley, & Harris, 2011). Based on these two previous studies, the evidence linking social capital measured with the Resource Generator to health outcomes remains sparse and inconsistent. On the other hand, a growing amount of evidence has suggested significant associations between social capital using the Position Generator and self-rated health (Carpiano et al., 2011; Moore et al., 2011; Song et al., 2009; Verhaeghe, Pattyn, Bracke, Verhaeghe, & Van De Putte, 2012), or mental health (Song et al., 2009), although the mechanism linking network social capital and health is still unclear. In the present study, we cannot examine the prospective association between social capital and health due to the cross-sectional design. However, as self-rated health has been established as a robust predictor of mortality (Idler et al., 1997), we expected that it would provide an additional test of the association between social capital and health. In addition, one could argue that self-rated health might be temporarily influenced by health status such as mental distress. Therefore, we hypothesized that individual-level social capital as measured with Resource Generator could be associated with self-rated health among a Japanese population, assuming mental distress as a mediator between social capital and self-rated health.

Furthermore, a recent study in Sweden suggested gender differences in the associations between social capital and self-rated health due to cultural beliefs that men and women are expected to behave differently (Eriksson, Ng, Weinehall, & Emmelin, 2011). In Japanese society, women spend more time in the home environment compared to men, while men spend more time in their workplaces and socialize outside work with their (mainly male) colleagues. In this sense, as home-makers are not formally classifiable based on occupational prestige-based scales (such as the Position Generator), but are nonetheless indispensable network members to people, we hypothesized that health status among women would be more closely linked to responses to the Resource Generator than men. That is, when applied to the identification of specific domains of social capital that promote health, we hypothesized that the Resource Generator would be more closely tied to the health of women than men (compared to the Position Generator). Indeed, as social resources are dependent on the cultural context, it is important to examine gender difference in the effects of social capital as measured with Resource Generator and test whether this concept can be applied to non-Western populations.

Section snippets

Participants

The study was based on the Okayama Social Capital Survey conducted by the City Government of Okayama in February 2009 (Iwase et al., 2012; Ueshima et al., 2010). Okayama is a mid-sized city with a total adult (aged 20–80 years) population of 518,758 (City Government of Okayama, 2009). The postal survey of 4000 residents of that city used a two-stage random sampling design to select 20 elementary school districts out of 87. Within each school district, 200 residents aged 20–80 years were

Results

Characteristics of the study subjects (740 men and 1035 women) are shown in Table 1. The City Government of Okayama reported that the mean age of the residents aged 20–80 years was 48.5 years and 51.6% were women, therefore our sample respondents were somewhat skewed towards more women (58.3%) and older ages (men; 51.6 years, women; 49.6 years) compared to the Okayama Population (City Government of Okayama, 2009). Among them 121 men (16.4%) and 167 women (16.1%) reported poor health. The

Discussion

The present study suggests that among both Japanese men and women, individual social capital measured by the Japanese version of the Resource Generator is associated with self-rated health even after taking into account potential confounders, and that individuals who have access to more personal resources benefit from their network-based resources. In sub-scales, we found differential patterns for men and women. That is, women tended to exhibit a more clear-cut dose–response relationship

Conclusions

Our findings provide further evidence of the potential utility of the Resource Generator for evaluating the relationship between individual-level access to social capital and self-rated health, although we also note that longitudinal studies remain sparse and we cannot exclude reverse causation due to the cross-sectional design of our study. Future research should test the robustness of our findings through longitudinal research as well as the use of objective health outcomes.

References (40)

  • M. Webber et al.

    Social capital and the course of depression: six-month prospective cohort study

    Journal of Affective Disorders

    (2011)
  • D.W. Brown et al.

    Associations between short- and long-term unemployment and frequent mental distress among a national sample of men and women

    Journal of Occupational and Environmental Medicine

    (2003)
  • Centers for Disease Control and Prevention

    Self-reported frequent mental distress among adults – United States, 1993–2001

    MMWR Morbidity and Mortality Weekly Report

    (2004)
  • Centers for Disease Control and Prevention

    Behavioral risk factor surveillance system questionnaire

    (2008)
  • City Government of Okayama

    The age distribution by sex from the Basic Resident Registration, the statistical report of the City Government of Okayama in 2009

    (2009)
  • J.M. Cortina

    What is coefficient alpha? An examination of theory and applications

    Journal of Applied Psychology

    (1993)
  • C.A. Cott et al.

    Determinants of self rated health for Canadians with chronic disease and disability

    Journal of Epidemiology & Community Health

    (1999)
  • H.D. Flap

    Creation and returns of social capital: a new research program

    La Revue Tocqueville

    (1999)
  • H.D. Flap

    Creation and returns of social capital

  • T. Harpham

    The measurement of community social capital through surveys

  • Cited by (47)

    • Social capital and the well-being of migrants to Australia: Exploring the role of generalised trust and social network resources

      2020, International Journal of Intercultural Relations
      Citation Excerpt :

      The Resource Generator specifically covers a range of common network resources such as access to education, prestige, political influence, financial resources, personal skills development and general support. Studies using the Resource Generator have shown access to resources to be associated with greater self-reported health (Kobayashi, Kawachi, Iwase, Suzuki, & Takao, 2013) and found an inverse relationship with common mental disorders (Webber & Huxley, 2007). To measure migrants’ access to network resources, we used the Resource Generator which is based on the original version used in the Social Survey of the Networks of the Dutch (SSND) (Flap, Snijders, & van der Gaag, 2000; Van Der Gaag & Snijders, 2005).

    • Neighbourhood relationships moderate the positive association between family caregiver burden and psychological distress in Japanese adults: a cross-sectional study

      2020, Public Health
      Citation Excerpt :

      It is well known that rich neighbourhood connections and supportive communities have a positive impact on residents' health.20,21 Rich neighbourhood relationships can buffer psychological distress by providing social support and helping people to access resources.22 For family caregivers, social support and easy access to care and medical care via neighbourhood relationships may help to maintain psychological health by reducing loneliness and helping to solve caregiving problems.

    • Dynamic analysis of the sustainable development capability of coal cities

      2020, Resources Policy
      Citation Excerpt :

      This leads to the shrinking of manufacturing industry in coal cities and the formation of an industrial pattern with a single primary processing industry as the main industry (Świąder, 2018). In addition, the lack of human capital is an important factor hindering the optimization of the industrial structure and sustainable development (Kobayashi et al., 2013). Most coal cities pay insufficient attention to investment in education, insisting that resources are the most important asset and neglecting the accumulation of human capital.

    View all citing articles on Scopus
    View full text