JECH Online First, published on January 13, 2017 as 10.

1136/jech-2016-208313
Glossary

Twenty years of social capital and health research:

a glossary
S Moore,1 I Kawachi2
1
Department of Health ABSTRACT network analysis (SNA) methods to measure social
Promotion, Education and Research on social capital in public health is resources and networks, while highlighting inequal-
Behavior, Arnold School of
Public Health, University of
approaching its 20th anniversary. Over this period, there ities in access to social resources. In both
South Carolina, Columbia, have been rich and productive debates on the definition, approaches, social capital is seen as an ecological-
South Carolina, USA measurement and importance of social capital for public level property (eg, interpersonal, organisational,
2
Department of Social & health research and practice. As a result, the concepts neighbourhood and policy) with individual-level
Behavioral Sciences, Harvard T. and measures characterising social capital and health health consequences.10 While these approaches
H. Chan School of Public
Health, Boston, Massachusetts, research have also evolved, often drawing from research differ, they are not mutually exclusive with some
USA in the social, political and behavioural sciences. The researchers seeking to bridge the two approaches.10
multidisciplinary adaptation of social capital-related The following glossary does not recommend one
Correspondence to concepts to study health has made it challenging for approach over another, but does aim to provide
Dr S Moore, Department of
Health Promotion, Education researchers to reach consensus on a common theoretical researchers with a set of key terms and concepts
and Behavior, Arnold School of approach. This glossary thus aims to provide a general prominent in recent research on social capital and
Public Health, University of overview without recommending any particular approach. health.
South Carolina, 915 Greene Based on our knowledge and research on social capital
St. (Discovery 1), Rm. 529,
and health, we have selected key concepts and terms BONDING, BRIDGING AND LINKING SOCIAL
USC, Columbia, SC 29208,
USA; that have gained prominence over the last decade and CAPITAL
mooreds4@mailbox.sc.edu complement an earlier glossary on social capital and The three related concepts of bonding, bridging
health. and linking social capital refer to the group con-
Received 1 September 2016
Revised 14 December 2016 texts in which social capital flows.4 Bonding social
Accepted 29 December 2016 capital refers to resources that are accessed within
INTRODUCTION networks or groups having generally similar
Social capital refers to the resources to which indi- characteristics (eg, class, race/ethnicity, age), thus
viduals and groups have access through their social possibly reinforcing exclusive social identities.5 6 11
networks.1 Social capital has been a key concept in Bridging social capital, on the other hand, refers to
public health and social epidemiology for 20 years. those social resources that may be accessed across
Figure 1 illustrates the number of articles with groups of different socioeconomic or sociodemo-
‘social capital’ in their titles published since 1997 graphic characteristics.5 6 11 Linking social capital
and found in the PubMed database. Research on is the norms of respect and networks of trust con-
social capital and health has given rise to important necting individuals and groups across formal or
debates on the conceptualisation, measurement and institutionalised structures of authority and
ultimately the value of social capital for health pro- power.5 6 11 All three forms may lead to improved
motion. There have been a number of reviews and health through different means, but they may also
commentaries on the meaning and utility of social lead to negative health consequences, particularly
capital in health research.2–7 Yet, few studies have bonding social capital. For example, having
offered a conceptual glossary of the key bonding social capital within a setting of few or
approaches, concepts and measures characterising limited resources may provide certain benefits (eg,
the field. Baum and Ziersch’s8 glossary on social sense of belonging), but it can also act as a liability,
capital is an important exception. While many con- particularly if social obligations become excessive
cepts and debates identified by Baum and Ziersch and down levelling norms are prevalent.12
remain relevant (eg, levels of analysis), there have
been important developments in social capital and COGNITIVE SOCIAL CAPITAL
health research in the intervening years. Cognitive social capital refers to those measures
Social capital and health is multidisciplinary with that assess people’s perceptions of trust, reciprocity
studies often drawing from theories and concepts and support.4 Among these cognitive measures,
in the social, political and behavioural sciences. trust has been central to research on social capital
This has made for a rich body of empirical evi- and health.13 Trust, which might be defined as the
dence, but it has also made it challenging to foster belief or confidence that someone or something is
a common approach. Two main approaches have reliable, truthful and honest, is often distinguished
characterised public health research: cohesion and between its general and particular form.
To cite: Moore S, network approaches.6–10 Cohesion approaches tend Generalised trust, or what Putnam9 calls ‘thin
Kawachi I. J Epidemiol
Community Health Published
to emphasise the cognitive or structural side of trust’, is often measured by asking respondents the
Online First: [ please include social capital through questions about trust in question: ‘Generally speaking, would you say that
Day Month Year] others, perceptions of social belonging and integra- most people can be trusted, or that you can’t be
doi:10.1136/jech-2016- tion, and levels of civic or social participation. too careful in dealing with people?’13 14 Such ques-
208313 Network approaches tend to rely on formal social tions are less about individual trustfulness as much
Moore S, Kawachi I. J Epidemiol Community Health 2017;0:1–5. doi:10.1136/jech-2016-208313 1
Copyright Article author (or their employer) 2017. Produced by BMJ Publishing Group Ltd under licence.
 Glossary

also suggested that the negative consequences are more likely to

arise within closed networks, which limit their members access
to broader, more diverse sets of resources and may enforce
negative norms and behaviours more strongly on their
members.10

NETWORK SOCIAL CAPITAL

‘Network social capital’ refers to measures of social capital that
come from the use of formal social network measurement
instruments and methods. From an ego network approach,
three main instruments have been used to measure network
social capital: name, position and resource generators. Name
generators ask respondents (ie, egos) to name others (ie, alters)
Figure 1 Number of articles published in PubMed with ‘social with whom they are connected along a certain relational dimen-
capital’ in the title from 1997 to 2015. sion (eg, discuss important matters). Once the alters’ names or
pseudonyms have been provided, respondents may be asked a
series of questions about each alter (eg, gender, age,
as a person’s perceptions of the trustworthiness of the social health-related behaviours) so as to gather information about the
environment.12 13 Particularised trust, or ‘thick trust’, is the composition of respondents’ networks. Finally, respondents may
trust that comes from specific interpersonal relations.9 be asked whether their alters know or interact with each other.
Particularised trust questions often appear in the form of ques- This final set of questions provides information on ego network
tions about a person’s trust in their neighbours or people in density (ie, number of ties present in ego’s network as a propor-
their neighbourhood.13 Extensive research has documented tion of number of possible ties present).22–24 Using name gener-
positive associations between generalised and particularised trust ator data, network social capital might be measured in terms of
and health.15 16 Researchers have argued that trust represents an the compositional heterogeneity or the average or maximum
important ‘moral resource’ that inheres within networks and level of social resources accessible24 25 or the structure of the
makes possible collective action and social solidarity.12 Others ego network (eg, structural holes).26 Name generators, particu-
have suggested however that generalised trust may be an inad- larly those asking about discussant networks, tend to capture a
equate proxy for ‘network-accessed resources’ since cognitive person’s stronger ties.27 Position generators ask respondents to
and network measures have been shown to be uncorrelated and identify whether they have a social tie to another person who
have independent associations with health.14 17 18 Further holds some type of position or occupation in society.28 29 A pos-
research is needed to discern the importance of trust and other ition generator often consists a list of occupations sampled from
cognitive measures, including whether they reflect access to psy- a larger set of occupations in society, with these occupations
chosocial, moral or social resources. having different levels of prestige or social value assigned to
them. Three measures may be calculated using the position gen-
erator: (1) diversity or extensity of social ties (ie, the number of
COHESION
different positions accessed), (2) upper reachability (ie, the
Kawachi and Berkman12 define social cohesion as a broader
highest ranked position accessed) and (3) range (ie, the differ-
concept than social capital, encompassing two inter-related
ence between the highest and lowest rank positions accessed).28
features of society: (1) the absence of latent social conflict—
Position generators were originally used in the study of job
whether in the form of economic inequality, racial/ethnic
attainment and social mobility, but have been increasingly used
tensions or other forms of polarisation; and (2) the presence of
in population health research.14 20 The position generator may
strong social bonds—measured by levels of trust, norms of reci-
better capture a person’s weaker social ties.25 28 Resource gen-
procity and the abundance of civic ties that bridge divisions in
erators provide respondents with a fixed list of specific resources
society. Social capital is a critical ingredient of social cohesion,
and ask them to identify whether they have access to a specific
but a community can possess strong social capital through its
resource, and the tie strength (as indicated by social role—kin,
individual members without those resources contributing to
friend or acquaintance) by which they access the resource.25
overall social cohesion (or in the case of some groups, such as
The list of resources may be grouped into different domains (eg,
the Ku Klux Klan, they can actually be destructive of social
prestige and education-related resources), and examined separ-
cohesion at large). The types of social resources emphasised by
ately to assess their importance for health outcomes. With
the social cohesion approach include: (1) the ability of a com-
important exceptions,30 31 resource generators have yet to be
munity to engage in collective action (collective efficacy); (2) the
fully integrated into health research.
ability of the group to enforce social norms (informal social
control) and (3) the ability of the community to express
RESOURCES
solidarity.12
Lin28 defines resources as material or symbolic goods that con-
tribute to basic human sustenance or have been ascribed with
NEGATIVE SOCIAL CAPITAL sociocultural meaning. Personal resources refer to those
While most empirical research has documented the benefits of resources that individual actors have in their own possession
social capital, researchers have also recognised that social capital and control. Personal resources may come into an actor’s pos-
can have downsides.15 19 20 Negative social capital refers to the session through ascription (eg, inheritance), acquisition (eg,
damaging effects that social capital may have for health. earning an educational degree) or exchange.28 Social capital
Negative consequences might arise within particular contexts research focuses on those resources embedded and accessed
due to ‘restricted opportunities’, ‘excessive demands’, ‘limited within social networks. Social resources may be classified into
freedoms’ and ‘down-levelling pressures’.21 Researchers have those resources that are accessible through (1) an actors’ direct
2 Moore S, Kawachi I. J Epidemiol Community Health 2017;0:1–5. doi:10.1136/jech-2016-208313
 Glossary

and indirect social ties (ie, social capital) or (2) the social iden- nuclear family and the importance of family social capital in
tity and recognition that may come from belonging to the same rearing children, viewed family social capital as consisting of
social class or group (ie, cultural capital).28 Social resources may three structural elements: (1) the strength of the relationship
be accessible to individuals or the collective as a result of strong between the adult(s) (ie, parent or guardian) and child, (2) the
social connections. For example, the stock of trust within a col- strength of the relationship between the adults who act as
lective (eg, a community) is a type of ‘moral’ resource which parents, and (3) the continuity of these relationships over
lubricates collective action, mutual assistance, reciprocity time.38 39 Neighbourhood social capital has been defined as
exchanges and the ability to enforce norms. Absent trust, it is social resources inherent within community networks, consisting
difficult for actors to engage in reciprocity exchanges or collect- of four main forms: (1) social support, (2) social leverage, (3)
ive action. Individual A does a favour for B (eg, lend money) informal social control, and (4) neighbourhood organisation
because A trusts that B will return the favour in the future. and participation.40 Workplace social capital refers to the social
Similarly, A will volunteer to perform a service for the collective resources accessible within work or occupational environments,
because he/she trusts that others in the group will also do their a setting in which social capital may be particularly salient
part, and that free riders will be sanctioned. for the health of working-age adults.41 42 Of particular
Psychosocial resources represent a range of individual emo- interest within the workplace setting has been the difference
tional or cognitive states and capacities emerging from the inter- between horizontal (employee-to-employee) and vertical
action of social, behavioural and psychological factors. (employee-to-employer/supervisor) ties, suggesting that the form
Psychosocial resources may be seen as (1) being shaped by social that one’s social capital takes in the workplace can have differ-
structures and contexts and (2) mediating the effects of social ential health impacts.43
structures on health.32 These include optimism, perceived
control, coping style and social support.33 SOCIAL NETWORK ANALYSIS
Social networks refer to the patterns of social ties existing
SOCIAL CAPITAL INEQUALITIES among a set of actors. Actors may consist of individuals, organi-
Social capital inequalities in health refer to systematic variations sations, groups or other salient social units. Social ties are the
in health arising from the differential accessibility that persons links connecting different actors; they may be informal or
or groups have to social resources.34 Lin28 suggests that inequal- formal, and can involve a range of types, including the exchange
ities in social capital arise from capital and return deficits. of expressive or instrumental resources, behavioural interactions,
Capital deficit refers to the relative inaccessibility of social or biological ties.44 SNA examines the pattern or structure that
resources for one group compared with another. For example, emerges from these social ties, and the influence that this struc-
persons with low income tend to have a lower quantity and ture may have on the outcomes of individual actors as well as
quality of resources accessible to them through their networks the network as whole. Three main approaches are used in SNA:
than those with high income. Assuming similar levels of capital, ego, chain (eg, snowball sampling) and sociometric approaches.
return deficit focuses on the processes by which certain groups Ego network approaches examine the social connections of indi-
gain greater benefit from their social capital compared with vidual actors, often focusing on the compositional features (eg,
others. Such processes may involve differences in the capacity of percentage of kin) of an actor’s network. Sociometric
certain groups to mobilise capital or the responses that environ- approaches examine the network structure characterising the
ments or institutions may have to the capital so mobilised.28 interactions of a delimited, bounded group of actors, asking
each actor within that group to report on their ties to every
SOCIAL CAPITAL INTERVENTIONS other group member. Chain approaches (eg, snowball or
Social capital interventions in health represent singular or mul- respondent-driven sampling) begin with an initial set of actors
tiple activities undertaken within or across different levels or who are asked to nominate or report on others with whom they
sectors of a system, with the aim of improving health through have ties of a specific kind (ie, their first-order zone); these
changes in a person’s or group’s access or capacity to mobilise actors are in turn asked to nominate others (ie, second-order
social capital.35 Social capital might serve as the target, channel zone), continuing on until a certain number of waves have been
or mediator, or the segmenting variable of an intervention, sampled or a point of saturation has been reached.44 45 Chain
depending on the intervention aims.15 35 36 For example, as the methods have often been applied to the study of hard-to-reach
segmenting variable, the intervention might be delivered differ- populations. Further information about SNA methods can be
ently to those groups having low versus high social capital. This found in a related glossary.22
may be seen in the differential capacity of communities with
low and high social capital to respond to or recover from a dis- STRUCTURAL SOCIAL CAPITAL
aster and the need to have in place different disaster manage- The word ‘structural’ in structural social capital refers to the
ment programmes and policies for such groups.12 In terms of presence or absence of formal opportunity structures or activ-
policy interventions, public policies may be evaluated according ities in which individual actors might develop social ties and
to whether they (1) are designed to build or support social build social networks. Measures of structural social capital at
capital directly, (2) incorporate social capital considerations into the macrolevel aim to capture the number or density of civic or
other policy assessments, and (3) redesign and improve policies neighbourhood associations, clubs, or other associational activ-
by leveraging current social capital.37 ities available to individual actors, whereas measures at the
microlevel aim to capture whether and, if so, the degree to
SOCIAL CAPITAL SETTINGS which individuals participate in social networks,46 associations
Three main settings have caught the attention of researchers on and civic events. Social participation, a commonly used measure
social capital: family or household, neighbourhood, and work- in this regard, has been shown associated with a range of health
place settings. Coleman38 39 defined familial social capital pri- outcomes and conditions, although these associations are some-
marily as the relations between children and parents (and other times inconsistent across studies. In addition, although social
family members if present). Coleman, whose focus was on the participation may be considered a more distal, upstream factor
Moore S, Kawachi I. J Epidemiol Community Health 2017;0:1–5. doi:10.1136/jech-2016-208313 3
 Glossary

influencing health through network or cognitive capital, most 7 Moore S, Haines V, Hawe P, et al. Lost in translation: a genealogy of the ‘social
studies have treated participation as a dimension of social capital’ concept in public health. J Epidemiol Community Health 2006;60:729–34.
8 Baum FE, Ziersch AM. Social capital. J Epidemiol Community Health
capital rather than an upstream cause of social capital. Further 2003;57:320–3.
research could seek to broaden measures of social participation 9 Putnam RD. Tuning in, tuning out: the strange disappearance of social capital in
to include more informal forms and assess the relationship America. PS Polit Sci Polit 1995;28:664–83.
between structural social capital and the other forms of social 10 Rostila M. The facets of social capital. J Theory Soc Behav 2010;41:308–26.
11 Poortinga W. Community resilience and health: the role of bonding, bridging, and
capital.
linking aspects of social capital. Health Place 2012;18:286–95.
12 Kawachi I, Berkman L. Social capital, social cohesion, and health. In: Berkman L,
TIE STRENGTH Kawachi I, Glymour MM, eds. Social epidemiology. New York: Oxford University
Tie strength can be defined in terms of the emotional intensity, Press, 2014:290–319.
13 Abbott S, Freeth D. Social capital and health: starting to make sense of the role of
intimacy, reciprocity and temporal duration characterising the
generalized trust and reciprocity. J Health Psychol 2008;13:874–83.
relationship between actors.47 Tie strength is an important 14 Carpiano RM, Fitterer LM. Questions of trust in health research on social capital:
concept in social capital research since it may reflect (1) the what aspects of personal network social capital do they measure? Soc Sci Med
types of resources that may be accessed as well as (2) the likeli- 2014;116:225–34.
hood that the resources accessed may, in fact, be mobilised. 15 Kawachi I, Subramanian SV, Kim D, eds. Social capital and health. New York, NY:
Springer, 2008:1–26.
Strong ties may be more easily mobilised than weak ties, but 16 Lindström M. Does social capital include trust? Commentary on Carpiano and
they often reflect homogeneous, insular social relationships that Fitterer. Soc Sci Med 2014;116:235–6.
can limit access to social and economic prospects, information, 17 Moore S, Bockenholt U, Daniel M, et al. Social capital and core network ties: a
and resources that lie beyond one’s close social circle.48 Weak validation study of individual-level social capital measures and their association with
extra- and intra-neighborhood ties, and self-rated health. Health Place
ties, in contrast, are often indispensable for bridging diverse
2011;17:536–44.
social groups and transmitting information, and facilitating 18 Carpiano RM, Hystad PW. ‘Sense of community belonging’ in health surveys: what
transactions across greater social distances, ultimately providing social capital is it measuring? Health Place 2011;17:606–17.
the basis for social integration.47 Direct measures of tie strength, 19 Portes A. Downsides of social capital. Proc Natl Acad Sci USA 2014;111:18407–8.
such as ‘closeness’, are preferable but not always available. In 20 Moore S, Daniel M, Gauvin L, et al. Not all social capital is good capital. Health
Place 2009;15:1071–7.
such cases, the social role (eg, kin, friend or acquaintance) char- 21 Portes A. Social capital: its origins and applications in modern sociology. Annu Rev
acterising the tie might act as a general proxy for tie strength. Sociol 1998;24:1–24.
Research has suggested that the assumption that non-kin ties are 22 Hawe P, Webster C, Shiell A. A glossary of terms for navigating the field of social
weaker than kin ties is accurate, although their predictive ability network analysis. J Epidemiol Community Health 2004;58:971–5.
23 Marsden P. Network data and measurement. Annu Rev Sociol 1990;16:435–63.
may be limited due to variations in tie strength within kin and
24 Campbell K, Lee B. Name generators in surveys of personal networks. Soc Networks
non-kin groups.49 1991;13:203–21.
25 Van Der Gaag M, Snijders TAB. The resource generator: social capital quantification
CONCLUSION with concrete items. Soc Networks 2005;27:1–29.
26 Burt RS. Brokerage and closure: an introduction to social capital. New York: Oxford
Due in part to the rich disciplinary diversity and conceptual University Press, 2005.
debates characterising research on social capital and health, the 27 Marin A. Are respondents more likely to list alters with certain characteristics? Soc
field remains vibrant, attracting new researchers interested in the Networks 2004;26:289–307.
link between social capital and population health and health 28 Lin N. Social capital. Cambridge: Cambridge University Press, 2001.
inequalities. The rise in intervention research and longitudinal 29 Lin N, Dumin M. Access to occupations through social ties. Soc Networks
1986;8:365–85.
studies promises to provide a more complete understanding of 30 Webber MP, Huxley PJ. Measuring access to social capital: the validity and reliability
the social, psychological and behavioural mechanisms linking of the Resource Generator-UK and its association with common mental disorder.
social capital to health. It has also made possible the investiga- Soc Sci Med 2007;65:481–92.
tion of the importance and effect of social capital on health 31 Kobayashi T, Kawachi I, Iwase T, et al. Individual-level social capital and self-rated
health in Japan: an application of the Resource Generator. Soc Sci Med
across the life course and the transmission of social capital bene- 2013;85:32–7.
fits between generations. Future research on social capital and 32 Martikainen P, Bartley M, Lahelma E. Psychosocial determinants of health in social
health might also seek to integrate qualitative and mixed epidemiology. Int J Epidemiol 2002;31:1091–3.
methods to understand better the various strategies and contexts 33 Taylor SE, Seeman TE. Psychosocial resources and the SES-Health relationship. Ann
in which individuals and groups access and mobilise social N Y Acad Sci 1999;896:210–25.
34 Moore S, Stewart S, Teixeira A. Decomposing social capital inequalities in health.
capital for health benefits. J Epidemiol Community Health 2014;68:233–8.
35 Moore S, Leroux JS, Salsberg J. Advancing social capital interventions from a
Contributors SM and IK both contributed to the conceptualisation, writing and network and population health perspective. In: Kawachi I, Takao S, Subramanian
editing of the manuscript. SV, eds. Global perspectives on social capital and health. New York, NY: Springer,
Competing interests None declared. 2013:189–204.
36 Valente TW. Network interventions. Science 2012;337:49–53.
Provenance and peer review Not commissioned; externally peer reviewed.
37 Productivity Commission. Social capital: reviewing the concept and its policy
implications. Melbourne, Australia: Commission Research Paper, 2003.
REFERENCES 38 Coleman J. Foundations of social theory. Cambridge: Belknap Press of Harvard
1 Bourdieu P. Forms of Capital. In: Richardson JG, ed. Handbook of theory for the University Press, 1990.
sociology of education. Westport, CT: Greenwood Publisher, 1986:241–58. 39 Coleman JS. Social capital in the creation of human capital. Am J Sociol 1988;94:
2 Macinko J, Starfield B. The utility of social capital in research on health S95–120.
determinants. Milbank Q 2001;79:387–427. 40 Carpiano RM. Toward a neighborhood resource-based theory of social
3 Hawe P, Shiell A. Social capital and health promotion: a review. Soc Sci Med capital for health: can Bourdieu and sociology help? Soc Sci Med
2000;51:871–85. 2006;62:165–75.
4 Harpham T, Grant E, Thomas E. Measuring social capital within health surveys: key 41 Kouvonen A, Oksanen T, Vahtera J, et al. Low workplace social capital as a
issues. Health Policy Plan 2002;17:106–11. predictor of depression: the Finnish Public Sector Study. Am J Epidemiol
5 Szreter S, Woolcock M. Health by association? Social capital, social theory, and the 2008;167:1143–51.
political economy of public health. Int J Epidemiol 2004;33:650–67. 42 Oksanen T, Kawachi I, Kouvonen A, et al. Workplace determinants of social capital:
6 Kawachi I. Commentary: social capital and health: making the connections one step cross-sectional and longitudinal evidence from a Finnish cohort study. PLoS ONE
at a time. Int J Epidemiol 2006;35:989–93. 2013;8:e65846.

4 Moore S, Kawachi I. J Epidemiol Community Health 2017;0:1–5. doi:10.1136/jech-2016-208313

 Glossary
43 Oksanen T, Kouvonen A, Vahtera J, et al. Prospective study of workplace social 46 Eriksson M, Ng N. Changes in access to structural social capital and its influence on
capital and depression: are vertical and horizontal components equally important? self-rated health over time for middle-aged men and women: a longitudinal study
J Epidemiol Community Health 2010;64:684–9. from northern Sweden. Soc Sci Med 2015;130:250–8.
44 Wasserman S, Faust K. Social network analysis: methods and applications. 47 Granovetter M. The strength of weak ties. Am J Sociol 1973;78:1360–80.
New York: Cambridge University Press, 1994. 48 Smith K, Christakis N. Social networks and health. Annu Rev Sociol
45 Heckathorn DD. Respondent-driven sampling: a new approach to the study of 2008;34:405–29.
hidden populations. Soc Probl 1997;44:174–99. 49 Marsden PV, Campbell KE. Measuring tie strength. Soc Forces 1984;63:482–501.

Moore S, Kawachi I. J Epidemiol Community Health 2017;0:1–5. doi:10.1136/jech-2016-208313 5