Investments in social capital—implications of social interactions for the production of health

https://doi.org/10.1016/S0277-9536(02)00242-3Get rights and content

Abstract

This paper develops a theoretical model of the family as producer of health- and social capital. There are both direct and indirect returns on the production and accumulation of health- and social capital. Direct returns (the consumption motives) result since health and social capital both enhance individual welfare per se. Indirect returns (the investment motives) result since health capital increases the amount of productive time, and social capital improves the efficiency of the production technology used for producing health capital. The main prediction of the theoretical model is that the amount of social capital is positively related to the level of health; individuals with high levels of social capital are healthier than individuals with lower levels of social capital, ceteris paribus. An empirical model is estimated, using a set of individual panel data from three different time periods in Sweden. We find that social capital is positively related to the level of health capital, which supports the theoretical model. Further, we find that the level of social capital (1) declines with age, (2) is lower for those married or cohabiting, and (3) is lower for men than for women.

Introduction

The social-capital literature marks a renewed interest in collective characteristics that shape individual and group outcomes. Social capital is a feature of the social context, not of the individual actors within the social context. In the economic literature, the relevance of social capital for economic development and growth has been recognised (Olson, 1982; Dasgupta, 1988; Scully, 1988; Grier & Tullock, 1989; North, 1990; Bardhan, 1995; Campos & Root, 1996; Stiglitz, 1996). In the recent public-health literature, high levels of social capital have been shown to be associated with lower total mortality (Kawachi et al., 1996; Kawachi, Kennedy, Lochner, & Prothrow-Stith, 1997; Wilkinson, 1996; Kaplan, Pamuk, Lynch, & Balfour, 1996), lower cardiovascular mortality and lower mortality from malignant neoplasms (Kawachi et al., 1997), lower mortality from accidents and suicide (Kawachi et al., 1996), more benign health-related behaviours (Lindström, Hanson, & Östergren, 2001) and lower crime rates (Sampson, Raudenbush, & Earls, 1997; Kennedy, Kawachi, Prothrow-Stith, Lochner, & Gupta, 1998; Kawachi, Kennedy, & Wilkinson, 1999).

A community with high social capital has been characterised and defined by Putnam as a community with high levels of civic engagement/social participation/social interaction, high levels of trust, political equality, and social structures that serve to enhance cooperation between its citizens (Putnam, 1993). Social capital thus consists of all the networks, norms, structures and institutions which facilitate social interaction (Grootaert, 1998; Lavis & Stoddart, 2002; Paldam, 2000). Social capital has mostly been operationalised using measures of social participation/social interaction or trust.

Most studies have used theoretical frameworks and corresponding statistical models derived from a “macro” perspective on the study of social capital, i.e. the focus has been the level of social capital in countries, states (in USA) or larger geographical areas (see most references above; see also Fukuyama, 1999; Putnam, 2000), and not the building, functioning or decline of social capital at the “micro” level of interactions and trust among individuals and groups. Furthermore, there has also been an emphasis on the decline of social capital in recent decades (Fukuyama, 1999; Putnam, 2000). In fact, there seems to be a lack of symmetry in the state of our knowledge of social capital. We have extensive knowledge concerning the forces that tend to destroy social capital but, regrettably, rather few notions of the factors and mechanisms that help to build it (Kawachi & Berkman, 2000).

Social capital can also be studied from a “micro” perspective, i.e. the levels of social interactions and trust among individuals and groups (Coleman, 1990; Manski, 2000). In fact, such a “micro” approach may plausibly contribute to our knowledge of the factors and causal mechanisms that help to build social capital. In the micro-economic model developed in this paper to study the building of social capital at the “micro” level of analysis, we consider the social interaction/social participation type of social capital and its interaction with the health and health investments of individuals and within families. Social interaction—in various degrees—is common to all economic agents. Social interaction is beneficial for at least two reasons: first, it is intrinsically valuable, i.e., social interaction contributes directly to the utility of its performers. Second, social interaction may improve the allocation of resources by improving information sharing, coordination of activities, and collective decision-making.

Several empirical studies have found a connection between social relationships and health (e.g. Berkman & Syme, 1979; House, Landis, & Umberson, 1988; Wolf & Bruhn, 1992; Shye, Mullooly, Freboorn, & Pope, 1995; Litwin, 1998; Berkman, 2000). A social network may, especially in times of stress (e.g. bad health), provide the individual with emotional, instrumental and informational support. Thereby it functions as an expansion of the individual's own resources. A social network may further exercise regulation and control over health-related behaviours such as drinking and smoking. Thus, health-related behaviour and social interaction are interrelated. However, no dynamic theoretical model of the relationship between health and social interaction has been proposed in the literature, in spite of the empirical evidence in favour of its existence being quite persuasive (Lavis & Stoddard, 2002).

The density of a social network is not given once and for all but corrodes through time if it is not properly maintained. Links to people with whom you never interact will probably become weaker and weaker. Therefore, in order to develop and maintain his or her social capital, an individual has to devote scarce resources (time and money) to social interaction.

The dominating micro-economic theoretical model for health-related behaviour is the Grossman model (1972a, b). In Grossman's original model, a singular individual's demand for health and health investments were analysed. Later, the model was extended to include the health behaviour of different agents in a family. Thus, Jacobson (2000) analysed the case where health-related behaviour was derived from common family preferences, Bolin, Jacobson, and Lindgren (2001a) the case where family members are Nash-bargainers, and Bolin, Jacobson, and Lindgren (2002c) the case where family members act strategically. Bolin, Jacobson, and Lindgren (2002a) recognised that agents outside the family sphere (e.g. an employer) may also have incentives for investing in the health of a family member.

In Becker (1974) people interacted with others in order to increase their own productivity of a consumed commodity. By combining the basic features of Grossmans (1972a), Grossmans (1972b) and Becker's (1974) ideas, and relying on our own previous work (Jacobson, 2000; Bolin et al., 2001a), we construct a model in which the family (household) simultaneously demands and invests in both health- and social capital. By definition, a family invests in social capital by forming and maintaining links to other people. The returns on this investment are both direct, as socialising yields utility, and indirect, as the social capital extends the family's resources. In our case the indirect effect will be captured, as the social capital facilitates the production of health within the family. Conceivable pathways for this mechanism run via provision of support and health regulation and control.

This paper develops a theoretical model of the family as producer of health and social capital, and utilises the empirical framework developed by Bolin, Jacobson, and Lindgren (2002b) and a set of individual panel data from three different time periods in Sweden to estimate the theoretical model. The paper proceeds as follows. First, we present our theoretical framework. The empirical model is developed in the following section. Third, the data is presented. Fourth, we describe the statistical methods employed in the paper. Fifth, the empirical model is specified in terms of the statistical method, and, sixth, the model is estimated. The paper concludes with a discussion of the results.

Section snippets

Theoretical framework

The dominant theoretical model for analysing individual health behaviour within health economics is the “demand-for-health” model introduced by Grossmans (1972a), Grossmans (1972b). According to this model, the individual is the ultimate producer of his or her own health. Naturally, the notion of “producer of health” does not mean that the individual determines his or her state of health—heredity, environment, and chance are three factors which may interfere—but rather that the individual can

Predictions

The inclusion of social capital in the demand-for-health model makes it possible to analyse the relationship between the family's demand for, and investments in health and social capital. The first-order condition, (7), implies that the net marginal cost of health investments decreases with the level of social capital. Thus, the stock of health capital is larger in families where the stock of social capital is high, ceteris paribus.

Predictions concerning the effect of age, wage, and education

The empirical model

Our empirical model follows Bolin et al. (2002b), based on Grossman (2000).8, 9

The data

A set of individual panel data based on the Swedish survey of living conditions, Undersökningar av levnadsförhållanden (ULF), was used; for details, see Bolin et al. (2002b). The data set includes health-related and background information for a panel consisting of approximately 3800 individuals, aged 16–84 yr old, for the years 1980/81, 1988/89, and 1996/97; the response rate is normally 80–85 per cent. Three different periods of observation are neccessary for a non-biased estimation of the

The statistical method and model specification

As the indicator of the unobserved variable health, we chose self-assessed health. As the social-capital indicator we chose whether or not the respondent has a close friend. The health-capital indicator was reported as an ordinal ranking, and, hence, we estimated an ordered probit model, whereas the indicator for social capital was reported as a binary response and was estimated by a probit model.17, 18

Empirical results and discussion

We begin this section by presenting our results and move on to a discussion in which we relate our results to results obtained in previous studies of the demand for health and health investments. The estimated marginal effects in the health equation are reported in Table 3 and the estimated marginal effects for the social-capital equation in Table 4.20

Discussion

In this paper we utilised a previously developed theoretical model of simultaneous production of both health and social capital (Bolin, Nystedt, & Lindgren, 2001) and the empirical framework developed by Bolin et al. (2002b) to examine theoretically and empirically the formation of health and social capital. We found that (a) social capital had a positive effect on the level of health and (b) that the level of social capital declined with age, that those married or cohabiting were more likely

Acknowledgements

Financial support by Handelsbankens forskningsstiftelser is gratefully acknowledged. The development and use of the date base were facilitated by research grants to Björn Lindgren from the Swedish Social Research Council, the Vårdal Foundation, the Swedish National Institute of Public Health, and the Medical Faculty of Lund University.

References (66)

  • M. Lindström et al.

    Socioeconomic differences in leisure-time physical activityThe role of social participation and social capital in shaping health-related behaviour

    Social Science and Medicine

    (2001)
  • M. Lindström et al.

    Individual and neighbourhood determinants of social participation and social capitalA multilevel analysis of the city of Malmö, Sweden

    Social Science and Medicine

    (2002)
  • H. Litwin

    Social network type and health status in a national sample of elderly Israelis

    Social Science and Medicine

    (1998)
  • D. Shye et al.

    Gender differences in the relationship between social network support and mortality

    Social Science and Medicine

    (1995)
  • T.C. Antonucci et al.

    An examination of sex differences in social support among older men and women

    Sex Roles

    (1987)
  • P. Bardhan

    Research on poverty and development 20 yr after redistribution with growth

  • G.S. Becker

    A theory of social interaction

    Journal of Political Economy

    (1974)
  • L. Berkman

    Social support, social networks, social cohesion and health

    Social Work in Health Care

    (2000)
  • L. Berkman et al.

    Emotional support and survival after myocardical infarctionA prospective, population-based study of the elderly

    Annals of Internal Medicine

    (1992)
  • L. Berkman et al.

    Social networks, host resistance, and mortalityA 9-yr follow-up study of Alameda County residents

    American Journal of Epidemiology

    (1979)
  • E. Bock et al.

    Suicide among the elderly. Isolating widowhood and mitigating alternatives

    Journal of Marriage and the Family

    (1972)
  • Bolin, K., Gerdtham, U., & Lindgren, B. (2002). A note on dynamic mis-specification in the empirical estimation of the...
  • Bolin, K., Jacobson, L., & Lindgren, B. (2001b). How stable are the results of the Grossman model? Testing different...
  • Bolin, K., Jacobson, L., & Lindgren, B. (2002b). The demand for health and health investments in Sweden 1980/81,...
  • Bolin, K., Nystedt, P., & Lindgren, B. (2001). Investments in social capital: Implications of social interactions for...
  • J. Campos et al.

    The key to the Asian miracleMaking shared growth credible

    (1996)
  • J. Coleman

    Foundations of social theory

    (1990)
  • P. Dasgupta

    Trust as a commodity

  • Deaton, A. (2001a). Health, inequality, and economic development. Working paper 8318, National Bureau of Economic...
  • Deaton, A. (2001b). Relative deprivation, inequality, and mortality. Working paper 8099, National Bureau of Economic...
  • F. Fukuyama

    The great disruption

    (1999)
  • W.H. Greene

    Econometric analysis

    (1997)
  • Grootaert, C. (1998). Social capital: The missing link? Social capital initiative. Working paper No. 3, The World...
  • Cited by (142)

    • The effect of cash transfers on mental health: Opening the black box – A study from South Africa

      2020, Social Science and Medicine
      Citation Excerpt :

      Health investments could act as mediators of the cash-transfer-mental health relationship. Extensions of the Grossman model include further channels such as social capital and socio-economic status (Bolin et al., 2003; van Kippersluis and Galama, 2013). The original Grossman model and its extensions point to four possible mediation channels of the unconditional cash transfer effect on mental health: social capital, socio-economic status, lifestyle choices, and physical health.

    • Firearm access and adolescent health: Safety in numbers?

      2020, SSM - Population Health
      Citation Excerpt :

      Furthermore, the presence of social capital has been shown to be associated with positive health outcomes. The community relationships shown through social capital are associated with fewer risk behaviors, such as smoking and binge drinking (Bolin, Lindgren, Lindström, & Nystedt, 2003). Social capital from family and school relationships are also negatively associated with delinquent behavior (Dufur, Hoffmann, Braudt, Parcel, & Spence, 2015).

    View all citing articles on Scopus
    View full text