ARTICLE IN PRESS Social Science & Medicine 61 (2005) 943–964 www.elsevier.com/locate/socscimed Social capital and mental health: An interdisciplinary review of primary evidence Astier M. Almedom! Department of Biology, Tufts University, 165 Packard Avenue, Medford, MA 02155, USA Available online 27 March 2005 Abstract An interdisciplinary interrogation of primary evidence linking social capital and mental health sought to establish: (1) ‘quality of evidence’ (assessed in terms of study design, methods used to address stated questions, rigor of data analysis, and logic and clarity of interpretation of results), and (2) applicability of the evidence to public health policy and practice with respect to mental health. It is found that social capital, a complex and compound construct, can be both an asset and a liability with respect to mental health of those in receipt of and those providing services and other interventions. The most meaningful assessment of social capital or components thereof may examine individual access to rather than possession of social capital, a property of groups, and therefore an ecological variable. Theoretical advances in research on social capital serve to identify mainly two types of social capital: bonding (between individuals in a group) and bridging (between groups). Each type of social capital has cognitive and/or structural component(s) and may operate at micro and/or macro level(s). Effective mental health policy and service provision may build or strengthen bridging social capital and benefit from both bonding and bridging social capital where either or both exist. Established indicators of social capital are amenable to quantitative and qualitative assessment, preferably in tandem. However studies that employ combined research design are rare or non-existent. Interdisciplinary multi-method investigations and analyses are called for in order to unravel mechanisms whereby social capital and mental health might be meaningfully associated. r 2005 Elsevier Ltd. All rights reserved. Keywords: Social capital; Mental health; Interdisciplinary research; Mental health policy; Health services Introduction Leading academic public health journal editorials, commentaries, and special issues continue to draw their readers’ attention to ‘‘social capital’’ and ‘‘mental health’’ or psychosocial variables (Muntaner, 2004; Kelleher, 2003; Greenberg & Rosenheck, 2003; Sartorius, 2003; Dannenberg, Jackson, & Frumkin, 2003; Saegert & Evans, 2003; Jackson, 2003; McKenzie, !Tel.: +1 617 627 3249; fax: +1 617 627 3805. E-mail address: astier-m.almedom@tufts.edu. Whitley, & Weich, 2002; Henderson & Whiteford, 2003). A compound and complex construct, social capital continues to appeal to the intellect and imagination of public health scholars, policy makers and practitioners alike, all of whom are immersed in the wider debate on poverty, health inequalities and social exclusion (Carlson & Chamberlain, 2003; Pearce & Davey Smith, 2003; Krishna, 2002; Moss, 2002; Pilkington, 2002; Muntaner, Lynch, & Davey Smith, 2001; Hawe & Shiell, 2000; Lynch, Due, & Muntaner, 2000; Baum, 1999, 2000; Kawachi & Kennedy, 1999; Leeder & Dominello, 1999; Lomas, 1998; Edwards & 0277-9536/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.socscimed.2004.12.025 ARTICLE IN PRESS 944 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 Foley, 1998; Edwards, Foley, & Diani, 2001; Wilkinson, 1996). The afore-mentioned authors constitute a representative sample of protagonist and antagonist theoretical and empirical stances on ‘‘social capital’’, often used as an umbrella term embracing social cohesion, social support, social integration and/or participation, among several other social determinants of health in general and mental health in particular. The substance of debate has been marked by polarized political overtones and mixed philosophical undertones. Scarcity of primary data purposely gathered to investigate associations between social capital and health and/or mental/emotional wellbeing has been a major constraint. However, this situation is rapidly improving, witness the steady rise in the number of papers on social capital indexed in MEDLINE and elsewhere between 1992 and 2002; and there are hints of reconciliation between hitherto polarized camps with opposing theories and interpretations of research evidence—See Putnam, 2004; Kawachi, Kim, Coutts, & Subramanian, 2004. Social capital is now integral to global discussions of sustainability and collective management of (common) resources, as exemplified by the recent special issue of Science which revisited (and reprinted) Hardin’s ‘‘Tragedy of the Commons’’—See Pretty, 2003; Hardin, 1968, 2003. Limitations in understanding of the multifaceted concepts of ‘health’, ‘community’ and ‘participation’ remain. These are all central to the social capital discourse, and by definition difficult to assess solely by means of quantitative methods of investigation and analysis (Cowley, 1997, 1995). They are all dynamic and process-oriented and not static or linear outcomeoriented phenomena, hence the need for researchers to carefully consider meanings already assigned to these terms and define their own use of them in any given context. Previous studies have highlighted common pitfalls in the use/misuse of ‘‘community’’ and ‘‘community participation’’ in health and social research (McDowell, Spasoff, & Kristjansson, 2004; Drevdahl, 2002; Bryson & Mowbray, 1981; Jewkes & Murcott, 1996; Hawe, 1996; Hawe, 1994; Fowler, 1991; Tumwine, 1989). Participatory and qualitative research evidence may shed light on otherwise intractable associations (see Cattell, 2001; Morrow, 1999, 2001; Wilson, 1997), but these have not featured prominently in the social capital and mental health debate. The turn of century has seen official designation of mental health as a ‘‘global burden of disease’’, with widespread depression accounting for most of the ‘‘burden’’ (WHO, 2001). A renewed shift of focus from curative to preventive measures has come about through multi-disciplinary enquiry into the mechanisms whereby social support and effective and efficient health care delivery may promote health in general and mental health in particular (Sartorius, 2002; Harpham, Grant, & Thomas, 2002; McKenzie et al., 2002; Kawachi & Berkman, 2001; Berkman, Glass, Brissette, & Seeman, 2000). This paper sets out to investigate what is known so far about the associations between social capital and mental health, what remains unknown (if not unknowable), and what possible policy and practice implications might be gleaned from available primary evidence. The limits of definition and measurement Social capital and mental health are both compound and complex terms which require multidimensional definitions and corresponding multi-method means of investigation and analysis. Concerning social capital, two distinct schools of thought are currently prominent in the published literature: Robert Putnam’s communitarian line of political thinking (Putnam, 1995) and Pierre Bordieu’s social theory of forms of capital (Bourdieu, 1986) amplified by James Coleman’s exposition of ‘‘family social capital’’(Coleman, 1988). These have both been subjects of critical discussion in the behavioral and social sciences (see Edwards & Foley, 1998; Edwards et al., 2001). Within social epidemiology/psychiatry, definitions of social capital are intertwined with the measurement scales used to quantify it, and there is little consistency in the literature. However, it is clear that social capital is not synonymous with (although it may well embrace and indeed be manifested by) social participation, social integration, social cohesion, and/or social support individuals can access or be barred from on account of their membership in groups and/or formal and informal institutions. Both theoretical and empirical analyses of social capital in a broad array of disciplines and sub-disciplines (including sociology, social epidemiology, and political sciences) identify two types of social capital: bonding and bridging. Each of these has two components (some authors call them ‘‘forms’’ of social capital): structural and cognitive operating at micro (individual—person or family/household) and/or macro (ecological—i.e. neighborhood, community, formal or informal group) levels—see Table 1 for a sketchy outline; and Macinko & Starfield, 2001; Hawe & Shiell, 2000 for substantive reviews. As these authors have pointed out, the language of social capital spans the realms of economic metaphor and political rhetoric. Definitions are usually extracted from the key protagonists (Putnam, Bourdieu, and/or Coleman) in précis form. Brief definitions capture little of the meaning with which constructs such as social capital are imbued. Researchers have taken the definition of their choice and run with it in different directions to extend the original meaning of this compound and complex construct; generating more questions than answers in the process. For the purposes of this review, a brief consideration of the roots of Putnam’s, Bourdieu’s and Coleman’s expositions of social capital is outlined. ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 945 Table 1 Social Capital: types, components and levels Type (direction) Component Level Bonding (horizontal) Structural (social networks); Cognitive (social control/efficacy; shared values; mutual trust and norms of reciprocity). Micro (individual, family/household) Bridging (horizontal: between different community and/or voluntary groups; and/ or vertical: between such groups and statutory as well as non-statutory organizations with power to make decisions on the distribution and/or allocation of public goods and services. Structural (access to public goods and services, amenities); Cognitive (participation; sense of belonging; decision-making capacity). Macro (statutory and/or voluntary organizations–local, national, international). Note: Szreter and Woolcock (2004) have introduced Linking social capital as a sort of diagonal (rather than vertical) bridge across power differentials. This proposition has generated varying responses ranging from support, intrigue, and outright opposition (see for instance Kawachi et al., 2004; Ellaway, 2004; Navarro, 2004). The social capital thesis of Robert Putnam and colleagues was developed around the workings of democracy at the level of local and regional government in Italy during its decentralization ‘‘experiment’’ which began in 1970. Devolution of power from the central government to the regions had opened up for citizens unparalleled opportunities for participation in political decision-making processes. ‘Social capital’ constituted one chapter out of six in Putnam et al.’s treatise Making Democracy Work (Putnam et al., 1993), but it was the key chapter, the anchor without which the preceding five chapters may have floated astray. Putnam et al.’s longitudinal data chart development and change in Italy’s diverse local government and civic regional institutions over a 20-year period of study. Local small-scale informal institutions such as rotating credit associations built on the basis of trust and reciprocity primarily between close friends and family relations feature significantly in this classic work. Rotating credit and saving associations (ROSCA) are well known, primarily women’s institutions which provide effective informal social security for their members. ROSCA exist practically in all parts of the world and they provide comprehensive social support (emotional, cognitive and material) directly to women (and thereby to their families) in ways that formal institutions normally cannot (see Ardener & Burman, 1995). However, neither Putnam nor his disciples consider carefully the workings of ROSCA, and whether or not ‘‘levels of social capital’’ can be adequately measured without incorporating sitespecific ethnographic narrative and analysis of genderspecific informal institutions such as these in Italy or in other countries. Putnam’s popularized version of social capital or lack of it in the USA (Putnam, 2000) obscures the salient points of his earlier work, and has justifiably generated rebuttals (Edwards et al., 2001; Edwards & Foley, 1998) including in mental health circles (Pevalin, 2003). Individuals and groups with material assets would be expected to both generate and benefit from the structural and cognitive components of social capital differently from those without. The ‘haves’ would be better placed to reciprocate goods and services amongst themselves compared to the ‘have-nots’. In countries like the USA where the haves and have-nots are structurally and socially segregated, there would be increased likelihood that the have-nots will be more miserable (and thus suffer worse emotional/mental ill health) than they might in Italy where the effects of such inequality may be mitigated by strong cultural influences of communitarian (as opposed to individualistic) society. More complex and intractable issues present themselves when gender, class and/or race come into the equation. Prominent scholars and practitioners have both tackled virtually insoluble questions that revolve around health inequality and social exclusion in public health generally (Wilkinson, 1996) and mental health specifically (Sayce, 2000). This paper argues that contemporary social capital and mental health discourse would do well to be inclusive rather than exclusive of historical and sociocultural narratives in order to avoid going round and round in circles of arguments between epidemiologists, however sociologically and/or anthropologically aware. If the premise that the fundamental goal of research is to portray as accurate a representation of reality as possible holds, then it would behove all researchers to seek information from the best possible source(s) using the most reliable method(s). As different academic disciplines and sub-disciplines carry with them their own strengths and limitations/biases in research, it is ARTICLE IN PRESS 946 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 necessary for policy makers, practitioners and other users of research findings (including the study participants themselves where power structures and other factors do not relegate them to mere passive respondents to survey questionnaires) to employ inter-disciplinary analyses of available evidence where it exists and to actively seek and/or commission it where it is lacking. Research on social capital and mental health may be particularly prone to misrepresenting study populations and distorting their realities, with possible detrimental public health and social policy consequences. Consider a juxtaposition of the history of civic participation/democracy and social capital in Putnam et al.’s Italy with Franco Basaglia’s movement of democratization of psychiatry and the process of deinstitutionalization of mental health patients during the second half of the last century—Table 2. The point here is that measures of ‘social capital’ in isolation from political and economic historical context of any given society are bound to produce only partial accounts of reality; and by implication, evidence that may be inapplicable (if not outright wrong and harmful) to policy and practice. Turning to the origins of Bourdeiu’s and Coleman’s notions of social capital, theory precedes empirical evidence. Pierre Bourdieu’s work on the ‘‘forms of capital’’ (human, cultural and social) was informed by a brief spell of quasi-ethnographic fieldwork in Algeria followed by extensive analysis and theorizing back in France (Bourdieu, 1986; see also Calhoun et al., 1993; Robbins, 1991). It is beyond the scope of this paper to discuss the impact of Bourdieu’s social theory upon sociologists and other researchers on both sides of the Atlantic, suffice it to say that it has had significant influence on Coleman’s exposition of ‘‘family social capital’’ (Coleman, 1988). Although Bourdieu’s ideas originated in observations of communitarian social capital (not dissimilar to Putnam’s) in rural Algeria, its interpretation and uptake in the USA has been distinctly individual. There remains much confusion in the literature as to whether social capital, a common good, can meaningfully be discussed at the individual level with respect to disease causation and/or health promotion (see for instance Davey Smith & Lynch’s most recent commentary, Davey Smith & Lynch, 2004). This is not unrelated to the ongoing crisis of identity and soul-searching among epidemiologists witnessed over the past decade; bringing to the fore cause-effect ambiguities around social determinants of health/illhealth; and the limitations of epidemiological training, theory, and methods (Krieger, 1994; Susser & Susser, 1996a, b; Shy, 1997; Walker, 1997; Morabia, 1998; Mcpherson, 1998; Pearce, 1999; Wall, 1999; Davey Smith & Ebrahim, 2001). This paper acknowledges these ongoing conversations in social epidemiology, and seeks to contribute new perspectives grounded in bio-medical anthropology. It is proposed that while it is reasonable to assess (qualitatively and quantitatively) individual access to available stocks of social capital, it may be problematic both theoretically and empirically to ‘measure’ individual levels of social capital; or derive from individual measures aggregate quantities of social capital said to be at the disposal of groups. This paper maintains that researchers from outside looking in are bound to present very different portrayals of reality from investigators/facilitators who engage their study participants in participatory co-investigations of social capital and mental health. Human beings thrive more in groups than in isolation (Berkman et al., 2000; Berkman, 1995). However, social ties may constrain individual freedom and sense of wellbeing; and there may be trade-offs between the safety of cohesive ties and the flexibility of weak ties (Gargiulo and Benassi, 2000; Kawachi & Berkman, 2001; Portes & Landolt, 1996; Granovetter, 1973). In Putnam’s popularized language (after de Souza Briggs), bonding social capital helps people to ‘‘get by’’, while bridging social capital enables them to ‘‘get ahead’’. These two functions of social capital may operate at micro levels of individuals, their families and social networks; or macro levels of formal and informal institutions to which individuals and/or groups subscribe (Table 1). While structural components of social capital are relatively easy to quantify, elements of cognitive social capital may only be fully examined by means of qualitative and participatory methods of investigation and analysis. Several researchers have grappled with questions of definition and quantitative measurement of social capital, and there are now established and adaptable scales for measuring collective efficacy in terms of informal social control, social cohesion and trust (Sampson, Raudenbush, & Earls, 1997; Saguaro Seminar, 2000; Harpham et al., 2002), as well as issue and/or site-specific scales developed for the purposes of assessing neighborhood quality (Yang, Yang, & Shih, 2002; Ross & Mirowsky, 1999). However, these do not necessarily incorporate meanings assigned to notions of ‘‘trust’’ (thick and thin), and ‘‘reciprocity’’ by their study participants/respondents themselves. Ideally, ethnographic/qualitative investigations should precede and inform the design of quantitative measurement scales; and/or coincide with and/or follow in order to help interpret statistical data. Qualitative and participatory assessments of social capital are few and far between (Morrow, 2001; Cattell, 2001; Wilson, 1997), and this presents a serious limitation on the extent to which health and social capital relationships can be properly understood. Defining and measuring mental health is equally challenging. Mental health encompasses a wide range of categories of illness and social behavioral disorders ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 947 Table 2 A juxtaposition of the ‘discovery’ of social capital and the democratization of mental health services in Italy Robert Putnam (with Robert Leonardi & Raffaella Nanetti), 1993 Franco Basaglia American political scientists: academic researchers/observers of the ‘‘Italian Experiment’’ of political decentralization and devolution during 1970–1989. Politically active Italian social psychiatrist, founder/pioneer of Italy’s national movement Psychiatria Democratica (democratic psychiatry) for the abolition of the institution of manicomio, lunatic asylum. Noted that bonding and bridging ties within and between informal and formal institutions in the 6 regions studied–Lombardia, Veneto, Emiglia-Romagna, Lazio, Puglia, and Basilicata—constituted a ‘‘moving picture’’. EmigliaRomagna, was rated the most ‘‘efficient’’ region in implementing policy at the grass-roots. Indicators of efficiency included bureaucratic responsiveness, leaders’ support for political equality, and provision of publicly supported [child] day care centers. Successfully instigated deinstitutionalization of the mentally ill, and reform of mental health services in two regions of the North during 1974–1980, Basaglia encountered unforeseen problems which indirectly implicate communitarian social capital alla Putnam et al. (1993): Some inmates did not want to leave the asylum because they had lost their sense of independence and capacity to reconnect with their families and communities, however cohesive. Basaglia ‘‘soon discovered that the ‘open door’ merely reminded the patients of the fact of their exclusion and rejection by the world outside. Instead of taking the cue to freedom and autonomy offered by the open door, the newly ‘liberated’ inmates at Gorizia remained passive and imprisoned by an internalized image of the asylum that was part of their new sense of selfy the open door produced fewer escapes, less ‘acting out’, and the great quagmire of patient gratitude to the benevolent doctor/father.’’ (Scheper-Hughes & Lovell, 1986, p.164). Observed that the ‘‘most civic’’ regions (as measured by voter turnout, an indicator of participation in political decisionmaking) of the North (Lombardia and Emiglia-Romagna) had the longest tradition of social trust and reciprocity garnered by rotating credit associations. These two particular northern regions also had stronger and more dynamic economies than southern regions, and were best equipped to use the decentralization and devolution experiment to their citizens’ best advantage. Southern local government institutions did not have the structures in place to use the Cassa per il Mezzogiorno (Fund for the South) public investments in the same way. Worked through gradual processes of creating ‘open community’ where former inmates became integrated in the local community, some working as staff of the hospital as nurses and administrators. When Basaglia moved from Trieste [where he successfully implemented the practice reforms he first directed in Gorizia] in the North to Rome in the South [presumably from high to low social capital levels], his legal triumph in demedicalizing and decriminalizing the mentally ill [Law 180, passed by the Italian Parliament in 1978 and described in Basaglia, 1980] run into some serious problems. In ScheperHughes & Lovell’s words (1986), ‘‘Elsewhere in the South, deliberate misapplication of the law resulted in a form of ‘dumping’ that Italians call ‘wild discharges’–i.e. patients literally bused off hospital grounds without any discharge plans or material or social resources. As a result, homelessness seems to have increased in several cities.’’ (1986: 172). Basaglia’s open door had turned into a ‘revolving door’, a common occurrence in other countries, including UK, USA and Canada (see Sayce, 2000; Szasz, 2003; Beaulieu et al., 2002). Noted that two regions, Emiglia-Romagna (in the North), ‘‘blessed [and still is, as always]y with virtually the most civic culture in all of Italy’’ and Calabria (South), ‘‘cursed [and still is, as always] by perhaps the least civic of Italian regional cultures–feudal, fragmented, alienated, and isolated’’ epitomized the divide between haves and have-nots in Italy. ‘‘By the 1980s, Emiglia-Romagnay was on its way to becoming the wealthiest region in Italy and among the most advanced in Europe, while Calabria was the poorest region in Italy and among the most backward in Europe.’’ (p. 154) including psychosis, anxiety, depression and substance misuse/addiction. One aspect of social capital, namely, informal social control in terms of adherence to shared norms of behavior has been shown to be positively associated with health promoting behaviors such as smoking cessation and lack of indulgence in excessive alcohol consumption (Lindström, Hanson, & Ostergren, 2000, 2001; Weitzman & Kawachi, 2000). However, social participation and access to common goods can also lead individuals to their early grave; if for instance, their closest emotional ties happen to be immersed in health demoting activities such as excessive smoking and alcohol consumption. This paper adopts the holistic definition of mental health espoused by Sartorius (2001, p. 101), ‘‘the state of balance that individuals establish within themselves and ARTICLE IN PRESS 948 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 between themselves and their social and physical environment’’. Two fundamental assumptions are made at the outset: one, health is conceptually ambiguous and defiant of objective definition, quantification and bureaucratic appropriation (Cowley, 1997) and therefore any measurable associations between social capital and mental health can only be approximate. Qualitative assessments of social capital in terms of ‘‘resources for health’’ (Cowley & Billings, 1999) may complement quantitative approximations; and two, public health policy (and practice) has historically been influenced (if not driven) by contemporary socio-political discourse, the latest of which is imbued with social capital (see Szreter, 2002a, b). Mental health policy and practice, critical components of this larger picture, remain central to the social capital discourse. There are fundamental limits to building social capital through top-down social engineering, but public policy and practice may support and strengthen bottom-up processes of social action to promote health, collective efficacy and economic prosperity. This review attempts to forward an inter-disciplinary framework of analysis of primary evidence that links social capital and mental health with a view to discerning the implications for inter-sectoral policy and practice with special reference to mental health. Method Main electronic bibliographic databases were searched for ‘social capital and mental health’; ‘social capital and psychosocial’ and ‘social capital and depression’ appearing in the summary/abstract, text and/or list of key words in peer-reviewed journal articles published and indexed by December 2003 (earliest indexing is 1966). Items resulting from the electronic search were hand-sifted in order to follow-up cited references and contact selected authors—see Box 1 for a step-by-step description of the review conduct. A final short list of a dozen studies reporting primary data on primary indicators of social capital and mental health were grouped into four categories for the purposes of thematic analysis: (A) social capital and the mental health and/or social behavior of children and adolescents; (B) social capital and adult emotional/mental health; (C) social capital and senior citizens’ mental health and/or emotional wellbeing; and (D) social capital and mental health service provision—see Table 3A–D. Results and discussion The twelve studies reviewed here reflect current state of affairs in social capital and mental health research, policy and practice. As expected, due to the compound and complex nature of both ‘social capital’ and ‘mental health’, multiple definitions and measurement scales/ assessment tools have been employed. Definitions of ‘metal health’ range from externalizing and/or internalizing behavior problems in children and young people (Beyers, Bates, & Pettit, 2003; Caughy, O’Campo, & Muntaner, 2003; Drukker, Kaplan, & Feron, 2003; Moffitt et al., 2002; van der Linden, Drukker, & Gunther, 2003); to social withdrawal, anxiety and depression (non-clinical, non-referred) in adolescents and young adults (Harpham, Grant, & Rodriguez, 2004; Stevenson, 1998); and ‘maternal depression and symptoms of antisocial personality disorder’ (Moffitt et al., 2002); ‘emotional wellbeing’ (Rose, 2000a; Cotterill & Taylor, 2001); and ‘psychological distress’(Mitchell & LaGory, 2002) in adults and senior citizens. Measurement scales and tools of assessment employed include, Child Behavior Check List (CBCL); Child Health Questionnaire (CHQ-CF87); interviews with children, adolescents, and/or their teachers, and/or parents/ primary carers; Revised Rutter Scale; Diagnostic & Statistical Manual of Mental Disorders (DSM-IV); Short form Multiscore Depression Index (SMDI); Teacher Report Form (TRF); Self-report Questionnaire (SRQ-20); Diagnostic Interview Schedule (DIS-IV); Short Michigan Alcoholism Screening Test (SMAST); General Health Questionnaire (GHQ-12); Mirowsky & Ross’ psychological scale; and also semi-structured interviews and observations. This means that each study needs to be examined in relation to itself vis a vis contemporary social capital and mental health debates and dilemmas, and in relation to other studies under review only with reference to policy and/or practice implications of the findings, if any. Measurement scales for social capital used correspond to the particular definition(s) adopted by each study. These include a Dutch translation and adaptation of Informal Social Control (ISC) and Social Cohesion and Trust (SC&T) scales; Neighborhood Social Capital scale (NSC), Kinship social support (KSS) and Fear of calamity scale (FOC); Adapted Social Cohesion and Trust scale (A-SCAT); interviews with youth, teachers and parents; Psychological Sense of Community (PSOC); and Putnam’s community social capital benchmark survey. A number of the studies reviewed measure two or more types and components of social capital, namely, the structural and/or cognitive components of bonding and bridging social capital are measured in geographically delineated urban areas. However, notions of ‘‘the shared social environment’’ are inconsistent across these studies. For example, ‘‘neighborhood’’ can mean ‘‘census block’’ (Caughy et al., 2003) or ‘‘census tract’’ (Beyers et al., 2003; Mitchell & LaGory, 2002) or ‘‘postcode’’ (Steptoe & Feldman, 2001). Only one study ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 949 Box 1 Literature search strategy and conduct of review Step 1 Five electronic bibliographic databases: CINAHL (1983–; cumulative index to Nursing and Allied Health Literature); HealthSTAR (1975–; Health administration, technology and research); MEDLINE (1966–; bibliographic database indexing approximately 4,5000 biomedical journals; worldwide coverage); PsychInfo (1974–; index of professional and academic literature in psychology and related disciplines); and Web of Science (1974– ; Institute of Scientific Information citation databases: Science Citation Index Expanded; and Social Sciences Citation Index) were initially searched for the phrases ‘social capital’ and ‘mental health’ appearing together in the title, abstract, key words and/or body of the text. These searches were re-run by replacing ‘mental health’ with ‘psychosocial’ and with ‘depression’. A decision was made to retain ‘social capital’ as a valid search term in its own right, and not to replace it with synonyms such as social integration and/or neighborhood quality. The results were hand sifted to identify items of specific relevance to mental illness (anxiety and depression were selected as primary indicators; and smoking and/or alcohol or substance misuse and/or welfare dependence as secondary). Non-English publications were included. Step 2 Additional review items were picked up from the reference lists of items found in Step 1; from Biosis Preview (1980–; a comprehensive life sciences database incorporating biological abstracts) PubMed citations; and by searching selected journals such as Social Science and Medicine for relevant articles ‘in press’. The electronic journal, Evidence-based Mental Health was screened for studies that may particularly have influenced current mental health policy and/or practice. Book reviews, dissertation abstracts, conference abstracts, and edited book chapters were set aside. Step 3 The items resulting from Steps 1 and 2 were divided into two broad categories: Review articles, journal editorials, commentaries and/or correspondences Research papers presenting primary or secondary data A final short-list of 12 papers presenting primary data on primary indicators of social capital and mental health was examined in detail. The following two questions guided the review: What is quality of evidence presented and/or marshaled to argue the case for positively or negatively linking social capital and mental health? What, if any, are the practical implications of the evidence for mental health policy and/or practice? uses the term neighborhood in an ‘‘ecologically meaningful’’ way, and recognizes that ‘‘perceived neighborhood’’ (according to the study participants) differs in meaning from the researchers’ use of the term (Drukker et al., 2003). The significance of access to and use of different types, components and levels of social capital varies across the life course. Geographical area-based social cohesion and informal social control translates into a sense of freedom and safety that is conducive to healthy cognitive and emotional development and socialization of children and adolescents (Ross, Reynolds, & Geis, 2000; Davis, 1998; Sampson et al., 1997). This is important for the physical safety, emotional security and wellbeing of senior citizens as well (Lindström, Merlo, & Östergren, 2003; Klinenberg, 2002). Residential social capital may be more critical to families (specifically women) with young children and to the elderly than to relatively young adults without dependants. Therefore empirical links between social capital and mental health are considered below with reference to specific stages of the lifecourse. The sub-grouping of studies in Table 3(A–C) is however fluid, as some studies belong in more than one sub-group. For example, Harpham et al.’s study includes adolescents and young adults (15–25 year olds), Moffitt et al. report on young mothers and their twin children, and Steptoe and Feldman’s sample has a very wide age range: 18–94 years, with a mean of 52 and SD of 18 years). Rather than listing these studies twice in Table 3, their ‘dual’ focus is discussed in the text only. Social capital and mental health and/or social behavior of children and youth Family and neighborhood social capital are evidently important determinants of children’s and adolescents’ 950 Table 3 Studies with primary data linking social capital and mental health of adolescents and young children Indicator of social capital & scale(s) used Indicator of mental health & scale(s) used Key findings Policy/practice implications & remarks Stevenson (1998) Anonymous city, USA Cross-sectional N ¼ 160 African American youth in an unnamed city in the northeast; correlations and multiple regression, mixed models Neighborhood social capital (NSC scale), Kinship social support (KSS scale) and Fear of calamity (FOC scale) Emotional adjustment/mental health (guilt, cognitive difficulty, sad mood, irritability, low self esteem, instrumental helplessness, social introversion, low energy, pessimism, learned helplessness); MDI (short form) m fear of potential violent calamity and k symptoms of global depression in girls compared to boys. Gender differences in access to neighborhood social capital and use of emotionally adaptive strategies, including social introversion. Supportive and watchful neighborhoods (m social capital) can make up for lower levels of family social support. Author argues in favor of building and strengthening structural components of social capital. His recommendation that families and social networks to which a child belongs need to be connected to larger networks of ‘fictive kin’ resonates with progressive education, social welfare and health policies . Beyers et al. (2003) Nashville (TN), Knoxville (TN), (IN) USA Longitudinal, two cohort study of children aged 5 followed into adolescence, age 13 (N ¼ 440) evenly distributed among 3 southern cities), multi-level Structural disadvantage, residential instability, concentrated affluence (census-based measures); parental monitoring and involvement (interview with youth and with parents, parental monitoring and activity scores) Externalizing behavior (e.g. ‘gets into fights’, ‘disobedient at school’) as reported by teachers (grades 6–8); TRF (34 items including aggression and delinquency scales) Neighborhood structure contributed to socialization of adolescents by moderating the effects of parental monitoring or lack thereof. Authors point out that their findings may not be generalized to African American families or disadvantaged youth as the majority of the sample consisted of white and middleclass families. Caughy et al. (2003) Baltimore city(MD), USA Cross-sectional N ¼ 200 African American mothers/ care givers of 3–4.5 year olds in 39 neighborhoods; singlelevel regression models Parental sense of community (interview with mothers/care givers; PSOC-G (general) and PSOC-K (‘knows neighbors’) scales Child behavior problems; CBCL scores for internalizing (anxiety, depression, withdrawal), externalizing (aggression) and total problem behavior score. Contradictory evidence: m wealth in residential area ¼ k social capital/ level of attachment with/sense of community in mothers/ carers ¼ k behavioral/mental health problems among 3–4 year olds; and yet k levels of neighborhood impoverishment and k maternal social capital also ¼ m child behavior/ mental health problems. Potentially harmful policy and practice implications as children as young as 3 and 4 may be labeled ‘aggressive’ or ‘depressed’. No information on how the African American mothers/care givers’ interviewed define their own communities; and whether or not they problematise their young children’s behavior. Drukker et al. (2003) Maastricht, The Netherlands Longitudinal cohort study of 11–12 year old children (N ¼ 3401) living in 36 ‘‘ecologically meaningful’’ neighborhoods, to be followed into adulthood; multilevel regression models. Neighborhood informal social control, social cohesion and trust; translated and adapted ISC and SC&T scales (with 5 site/culture-specific questions added). General mental health and behavior (aggression, delinquency, hyperactivity, impulsivity and social withdrawal), self esteem; CHQ-CF87 Children’s mental health and behavioral problems specifically associated with neighborhood levels of informal social control. Promising prospects for a sound evidence-base for mental health policy and practice as authors are aware of and responsive to the limitations of epidemiological survey research. ARTICLE IN PRESS Study design, sample size & unit(s) of analysis A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 Author (year), location of study site(s) Studies with primary data linking social capital and mental health of young, adults (including mothers with young children) Indicator of social capital & scale(s) used Indicator of mental health & scale(s) used Key findings Policy/practice implications & remarks Rose (2000) Russian Federation Cross-sectional, multistage randomly stratified sample, N ¼ 1904 adults age 18 and up; Multiple regressions. Multiple indicators of social integration and individuals’ cumulative use of networks including church attendance, trust in people, sense of control over one’s life. New Russia Barometer (NRB) surveys. Emotional wellbeing (12 months recall) 5 point lickert scale: ‘‘in the past year, would you say your emotional health has been very good, good, average, poor or very poor?’’ NRB surveys. Social capital increases physical and emotional health more than human capital; Human and social capital together can easily raise the individual’s selfreported health from just below average on the scale to approaching good health. Author argues that public policy intervention to increase household incomes coupled with autonomy of social capital networks from government would secure emotional and physical health benefits for Russians across all age groups. Steptoe and Feldman (2001) London, UK Cross-sectional N ¼ 658 postal questionnaire survey respondents in the London area; multilevel regression models. Collective efficacy: social cohesion (SC), informal social control (ISC); neighborhood problem scale Feeling unhappy and depressed (GHQ-12, 4point scale) Neighbourhood problems (including litter in the streets, air pollution, noise, vandalism and disturbance by neighbors or youngsters) correlated with poor self-rated health, psychological distress and impaired physical function independent of age, sex, neighbourhood SES, individual deprivation, and social capital. Authors point out that the cross-sectional design and low response rate make it difficult to go beyond recommendations for further research. Moffitt et al. (2002) England & Wales, UK Longitudinal, N ¼ 1116 women who became mothers of twins in 1994–95; 562 of whom were o 20 yr old at the time of their twins’ birth. Neighborhood social cohesion and Trust, informal social control. Mother’s mental health history, symptoms of antisocial personality disorder, maternal depression; SMAST; DISIV; Children’s cognitive ability and prosocial behavior; TRF; DSM-IV. ‘Personality traits’ suggestion that younger mothers were less ‘conscientious’ and with more ‘problematic’ mother-child relationships; Mother and teacher reports show m’inattentionhyperactivity’ in children of younger mothers; equal participation in prosocial activity. Authors make explicit policy recommendations for prevention of teen childbearing, and support for teenage mothers to gain access to child care, education, housing, employment and mental health services. They call for comprehensive, ‘multimodal’ interventions. ARTICLE IN PRESS Study design, sample size & unit of analysis A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 Author (year), location of study site(s) 951 952 Table 3 (continued ) Indicator of social capital & scale(s) used Indicator of mental health & scale(s) used Key findings Policy/practice implications & remarks Mitchell and LaGory (2002) Birmingham (AL) USA Cross-sectional N ¼ 222 households (30 Censusblocks) Individual’s extent of participation in the community (bonding) and strength of trust and bridging ties (Social capital community Benchmark survey) Mental distress (Mirowsky & Ross’ psychological scale) Strong bonding ties within group; weak bridging ties with other groups; mparticipation ¼ m mental distress due to increased demands on time & resources; m‘mastery’ more important than social capital in mitigating mental distress. Findings concerning ‘mastery’ are potentially useful for designing support interventions to promote individual mastery (and by implication collective resilience) in cohesive inner-city African American communities. Harpham et al. (2004) Cali, Colombia Cross-sectional N ¼ 1168 young people; factor analysis and logistic regression models Social cohesion and Trust (thick and thin); A-SCAT scale Global depression and anxiety (non-clinical); SRQ-20 manxiety and depression in females; low levels of education and employment are more significant risk factors for mental ill health than social capital. In the presence of violence (as a variable), social capital has no statistical effect on mental health variables. Guarded policy and practice recommendations. More research in progress. Studies with primary data linking social capital and mental health with reference to senior citizens Author (year), location of study site(s) Study design, sample size & method(s) of analysis Instrument used to assess social capital & mental health Key findings Policy/practice implications & remarks Cotterill and Taylor (2001) Plymouth, UK Cross-sectional, N ¼ 95 participants of Plymouth HAZ-funded Age Well Project in six locations; N ¼ 10 non-project participants and N ¼ 10 staff from voluntary organizations involved in AW Project. Study assessed reported ‘‘social health’’: in terms of social participation, social networks and interpersonal interaction. Housebound elderly people benefited from opportunities for social interaction, but did not want to spoil the atmosphere of social gatherings by ‘‘talking about what was wrong with them’’; health information generated fear and threatened day- Authors highlight the complex and contradictory consequences of unwelcome health information and the welcome social interactions to combat isolation and loneliness in order to promote older ARTICLE IN PRESS Study design, sample size & unit of analysis A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 Author (year), location of study site(s) Author (year), location of study site(s) Study design, sample size & unit of analysis Indicator of social capital & scale(s) used Indicator of mental health & scale(s) used Key findings Policy/practice implications & remarks Rosenheck et al. (2001) USA Cross-sectional, one year follow-up observational study of 18 sites participating in the ACCESS (Access to Community Care and Effective Services and Supports) program for homeless seriously mentally ill patients (entered in the study in two cohorts) in the USA (9 States). N ¼ 2668 (mean age 38.5 yr; 64.4% male; 45.3% African American. Cross-sectional, casecontrol (mental health service users versus nonusers), N ¼ 262 children (56 cases and 206 controls) living in Maastricht neighborhoods; Multilevel logistic regression models. Number of club meetings attended in past 12 months; number of community projects worked on; number of participants in volunteer work; general belief that other people are honest; proportion of adults who voted in the 1994 and 1996 elections. Clinical diagnoses of mental illness: psychiatric problems and substance abuse problems m community social capital associated with greater system integration and greater access to assistance from a public housing agency and to a greater probability of exit from homelessness at 12 months. Collaboration between service providers and service integration may improve outcomes for homeless mentally ill people. More data expected to become available at the end of the ACCESS demonstration period to guide policy and practice more specifically. Neighborhood informal social control, social cohesion and trust; translated and adapted ISC and SC & T scales (with 5 site/culturespecific questions added). Not specified. van der Linden et al. (2003) Maastricht, The Netherlands No associations between environmental factors, or systems integration and psychiatric problems. More children from lower SES neighborhoods seen by mental health care services; Neighborhood social capital (social cohesion & trust) mitigate the effects of lower SES and children’s coming into contact with mental health services. This study makes explicit justification for ‘early intervention’: parenting and family support strategies. Authors argue that prevention/ programs for high-risk children should seek to alleviate neighborhood deprivation by creating safe areas for children to play and for their parents to meet and increase social cohesion. ARTICLE IN PRESS people’s sense of wellbeing and happiness. A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 to-day coping strategies. AW project participants engaged in the active management of their sense of wellbeing by avoiding some topics of information in order to stay happy. Studies with primary data linking social capital and mental health with reference to health care and service provision 953 ARTICLE IN PRESS 954 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 development, health and wellbeing. Both individual and ecological factors are at play, warranting plurality of methods and levels of investigation and analysis. Stevenson (1998) defines social capital as ‘‘the sum total of positive relationships including families and neighbors that serve as buffers to the negative influences within one’s immediate environment.’’ (p. 48) He then presents a careful account of mechanisms whereby race, psychological sense of belonging and neighborhood economic deprivation interact to shape mental and emotional health and wellbeing of adolescents in an anonymous American city located in the North-east. This study addressed three questions: ‘‘(a) do African American youth who live in self-reported unsafe neighborhoods show higher levels of depression? (b) are there gender differences according to perception of calamity, social capital and depression? and (c) do adolescents from supportive families and neighborhoods demonstrate healthier psychological outcomes compared to adolescents who have only one of these supports?’’ (p. 49). These are all critical questions around which subsequent discussions of childhood and adolescent education, mental health, employment, and civic participation in the USA have revolved (See Goodman, Huang, & Wade, 2003; Dika & Singh, 2002; Hill & Herman-Stahl, 2002; Crystal & DeBell, 2002; Saegert, Winkel, & Swartz, 2002; Larson, 2000; Vimpani, 2000; Lerner, 2000; Ross et al., 2000). Stevenson’s insightful analysis highlighted the need for interventions to recognize and bolster existing support systems available to adolescent boys and girls living in racially segregated socioeconomically disadvantaged urban quarters. Stevenson observed gender differences in perceptions of potential calamity and expressions of fear. Adolescent girls were more likely to express fear of calamity and benefit from access to neighborhood social capital than their fearless male counterparts. Girls were less likely to report depressive ideation including lethargy, instrumental helplessness, and cognitive difficulties even when they lived in high-risk locations. Being fearful of violent calamity and articulating this fear is shown to be an emotionally adaptive strategy teen-age girls use to both generate and access social capital. Moreover, social isolation resulting from fear of violent calamity may promote resilience (p. 56). Stevenson couches his crime prevention and mental health promotion policy and practice recommendations in a comprehensive discussion in favor of building neighborhood social capital and healthy communities through adult supervision and care of adolescents (see also Stevenson, 1997). This resonates with and complements other studies reporting differential levels of depression according to racial identity associated with the social stressors of discrimination and social exclusion (see Campbell & McLean, 2002; Chakraborty & McKenzie, 2002; Hill & HermanStahl, 2002; Sharpley, Hutchinson, & McKenzie, 2001). Furthermore, recent developments in mental health policy research and practice add credence to Stevenson’s forward-looking approach to mental health promotion among marginalized groups (see López, 2003; Manson, 2003). Beyers et al.’s longitudinal study (Beyers et al., 2003) conducted in three southern cities of the USA (Nashville, TN, Knoxsville, TN, and Bloomington, IN) independently reinforces Stevenson’s call for concerted efforts to build and strengthen structural and cognitive social capital through prevention/intervention programs. This study addressed two questions: ‘‘(i) do neighborhood structural disadvantage, concentrated affluence, and residential instability relate to initial levels of and/or growth in adolescence externalizing behavior after controlling for individual and family factors? and (ii) do gender and parenting practices differentially affect the development of externalizing behaviors depending on the social structure of neighborhoods in which families reside?’’ (p. 36) Jennifer Beyers and her team use Coleman’s definition of social capital as ‘‘yphysical presence of adults in the family and the quality of relations among family members’’ (p.46), and describe family level collective efficacy as connectedness of social networks among resident adults and youths (after Sampson et al., 1997). They confine their investigation to externalizing behavior problems among youth, and conclude that while neighborhood structure does not directly impact externalizing behavior, it contributes to the socialization of adolescents via the moderating effects of parental monitoring. The authors are careful to point out that their findings are not generalizable to African American youths and/or low SES densely populated urban American neighborhoods, as this category constituted only 17% of their study sample across three southern American cities. However, their findings resonate with ‘‘neighborhood-effect’’ studies of SES in relation to adolescent behavior and mental health, most notably Aneshensel & Sucoff’s evidence (1996) from Los Angeles neighborhoods. Caughy et al. (2003) focus on African American mothers/carers of young children in a racially segregated American city (Baltimore, Maryland) and find that the mother/carer’s ‘‘lack of attachment to community was a risk factor for behavior problems for children living in wealthy communities but, a protective factor for children living in highly impoverished neighborhoods.’’ (p. 231). This study demonstrates a somewhat muddled view of social capital. Social capital (bonding and not bridging type) is investigated in this study in relation to neighborhood ‘context’ with contradicting results. Margaret Caughy and her team use ‘census block’ as a proxy for neighborhood, and do not attempt to examine the meaning of ‘‘community’’ in the context of their study site and sample of respondents. Their suggestion that weak neighborhood ties may be indicative of weak ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 community ties and that African American mothers and/or pre-school children may be better off without their communities is questionable. ‘‘Contextual analysis’’ without enquiry into the meanings and boundaries of the community in question presents a serious limitation given what is already known about the issues of community, particularly in the context of health research (see for example Drevdahl, 2002; Jewkes & Murcott, 1996; Hawe, 1994). Furthermore, ‘‘knowing’’ and/or ‘‘trusting’’ neighbors enough to delegate childcare responsibilities to them may be a red herring if the respondents’ sense of community is non-geographically defined. Other studies have found neighborhood economic deprivation to be associated with increased levels of disorder that may result in psychological distress among residents who do not lack social ties (Ross, 2000). Moreover, analysis of nationally representative data on socializing with neighbors and with friends outside the neighborhood (United States, 1974–1996) had found a secular trend of decline in the importance of neighborhood-based social ties, and hints of a rise in the importance of non-neighborhood social ties (Guest & Wierzbicki, 1999), demonstrating the subtlety of meanings of social interaction that certain measures of social capital may simply miss. Non-spatial social ties include ‘befriending in cyberspace’ provided by voluntary organizations such as the Samaritans in the UK whose services have extended beyond telephone and face-toface contact through email (see for instance Bale, 2001). It would be reasonable to suggest that social cohesion in the context of poverty and structural disadvantage poses mental health risks to women either because they tend to be giving more than receiving, or because they may be constrained by the norms and expectations of their social ties (Kawachi & Berkman, 2001), but Caughy et al.’s study does not consider such possibilities. Mindy Fullilove (1998), a social psychiatrist, analyzing the insights of insiders has demonstrated that building social cohesion and collective efficacy in four different American inner-city locations was beneficial for women, because ‘‘women have major responsibility for raising childreny The importance of social connections is not simply a matter of social intercourse, but more profoundly a matter of getting women’s work done. Loss of social cohesion in the larger community will make women’s work more onerous. Conversely, improvements in social organization create networks that allow women to share responsibilities and aid each other.’’ (p. 76) Caughy et al.’s suggestion that ‘‘being alone might be better’’ thus runs counter to Fullilove’s, Stevenson’s, and Beyers et al.’s assertions. The latter highlight positive aspects of social capital with respect to the behavioral development and social adjustment of children and youth; while the former expressly set out to find non-salutary effects of communitarian social capital on individual wellbeing. Caughy et al.’s study is likely to fuel the ongoing 955 politically charged debate in epidemiology regarding social capital and public health in general and mental health in particular. A simplistic interpretation of Caughy et al.’s evidence may also serve to perpetuate stigma through undue labeling of black children and young people. Lessons from the struggle to reconcile past erosion of cultural and social capital of minority groups with present day notions and realities remain critical (see Hunter, 1998). It thus behoves researchers to ensure that meanings assigned to key terms such as community and neighborhood are properly examined before conclusions are drawn from results of statistical analysis, however advanced. In sharp contrast, Drukker et al. (2003) define neighborhood in an ‘‘ecologically meaningful’’ way, and demonstrate care in fine-tuning their chosen measurement scale for specific components of social capital to suit their study participants. These authors adopted Sampson et al.’s ISC and SC&T scale (Sampson et al., 1997) and translated it into Dutch, adding five new questions in order to make it specifically relevant to Maastricht (small city) neighborhoods. This study benefits from and reinforces a related case-control study of children’s mental health services in Maastricht (van der Linden et al., 2003) which is discussed in Social capital and mental health service and care provision below. Drukker et al.’s longitudinal study was designed to investigate associations between SES and social capital; and how these influence behavior and quality of life of children on the brink of adolescence. The study design is robustly eco-epidemiological, and the baseline evidence indicates that children living in ‘‘better’’ economic and social capital (low instability) neighborhoods enjoy better quality of life, better general and mental health and exhibit more pro-social behavior as they embark on adolescence. The evidence pin-points children’s mental health and social behavior association with one particular aspect of social capital: informal social control. Social capital and adult mental health and emotional wellbeing Papers summarized in Table 3B include two crosscultural studies of social capital and emotional/mental health (Rose, 2000a; Harpham et al., 2004). Richard Rose’s New Russia Barometer (NRB) study (Rose, 2000a) sets out to find out whether it is human capital (education, subjective social status, and household income), or social capital (social integration, formal and informal links with others, someone to rely on if ill, etc.), or both human and social capital combined which primarily determine individual health (p. 1423). Rose’s NRB questionnaire was designed to measure ‘‘different forms of networking, some familiar in Russia and unfamiliar in the West, and some common to both ARTICLE IN PRESS 956 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 types of societies’’ and ‘‘administered to a full-scale multi-stage randomly stratified sample covering the whole of the Russian Federation, urban and ruraly’’ (p. 1425). This study presents purposely collected data on social capital in the Russian Federation; an improvement on previous studies such as for example, Kennedy, Kawachi, and Brainerd (1998) which involved secondary analysis of survey data, ‘‘retro-fitting’’ the concept of social capital on data collected for other purposes. However, it is worth noting that Rose’s data on emotional health are subject to recall error. In anthropological and related areas of health research, 12 month recall is considered too long to produce reliable information. Nevertheless, Rose’s multiple regression models showed that human capital could explain 12.3% of the variance in emotional health; while social capital explained 15.7%, and a composite model with human capital and social capital variables together explained 19.3% of variance in emotional health. Social capital significantly influenced involvement in or exclusion from formal and informal networks; friends to rely on when ill; control over one’s own life; and ‘‘trust’’. Younger Russians (o40 years of age) had greater sense of control of their lives compared to their middle-aged and older compatriots. Rose argued that social capital, a multifaceted construct, cannot be reduced to a single measure, and cautioned against using aggregate membership statistics as a proxy for social capital in aggregate analysis because, ‘‘The fullest understanding of the influence of social factors on health is best achieved by recognizing the independent influence of selective forms of both individual and social capital’’ (p. 1431). Rose concluded that public policy can only intervene in economic terms—to ensure sustained growth in household incomes and to promote resilience. It is worth noting here that the prominence of ‘‘anti modern’’ society and culture in contemporary Russia contributes to the complexity of the picture partially presented in this study—see also Rose (2000b, 2001). Steptoe and Feldman (2001) investigated neighborhood-level effects of deprivation and deficit of social capital on self-rated health and psychological distress (measured using the GHQ-12). Neighborhood problems, including litter in the streets, air pollution, noise, vandalism and disturbance by neighbors or youngsters correlated with poor self-rated health, psychological distress and impaired physical function independent of age, gender, neighborhood SES, individual deprivation, and social capital. The study participants represented a ‘‘stable residential population’’ with a very wide age range (18–94 years; M ¼ 52; SD, 18), and the authors posit and confirm that higher SES neighborhoods had higher levels of social capital. This could however be an artifact of postal questionnaire response—a response rate of 24% is low. Descriptive epidemiological studies such as this one tend to be limited, as that they confine themselves to quantitative methods of analysis, and do not adequately investigate underlying context and meaning. Evidence presented by Terrie Moffitt et al. (2002) serves to demonstrate how quantitative data from descriptive epidemiological studies may benefit from existing qualitative data to enhance the quality and applicability of evidence for policy and practice. This study is discussed within the sub-group of reviewed papers on social capital and mental health of adults and young people because the authors expressly focus on and prioritize mother-centered interventions. Moffitt and her team compared younger mothers of twins in England and Wales with older ones in order to examine a wide range of social and behavioral risk factors associated with poor child mental health outcomes. Environmental factors (including younger mothers’ mental health history, biological father’s mental health history, social support for parenting, neighborhood social cohesion, and twins’ cognitive development and behavior at age 5) had negative prognoses for younger mothers and their twins compared to older mothers and their twins. This study’s findings and recommendations merit discussion in the wider context of UK health policy and practice reform. Reducing social exclusion and building social capital have been New Labour’s explicit goals of health service modernization; and reducing (unwanted) teenage pregnancy and mental health promotion focusing on children and young people had been prioritized (see Social Exclusion Unit, 1999a, b). The term ‘‘teenage’’ is not unambiguous, however. It needs careful defining. A qualitative study designed to assess health needs, attitudes and aspirations of young people in South London where teenage pregnancy rated highest in Europe, had revealed that ‘teenage pregnancy’ was a heterogeneous category that embraced cases of underage (unwanted) pregnancy occurring before girls reached the age of consent (sometimes as young as 12) as well as deliberate (wanted) pregnancy among 16–19 year olds who often considered themselves ‘adults’ (Health First, 1999). This latter group disapproved of ‘infantalizing’ approaches to their needs on the part of practitioners in health and social services who summarily problematized teenage childbearing. Parenthood in (late) teens was often a function of life aspirations, economic and social needs—a deliberate choice on the part of girls and young women, mainly in working class families following their own mothers’/role models’ example of early parenting. Considering Moffitt et al.’s findings alongside the qualitative evidence summarized above would strengthen their policy recommendations. Practitioners involved in the allocation of resources to facilitate child care access for ‘‘teenage’’ mothers to enable them to build their human capital through education and employment would gain better understanding of their clients by ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 integrating qualitative research evidence. Lack of communication and coordination between quantitative and qualitative researchers, and between researchers and practitioners has continued to hinder social inclusion and achievement of health improvement policy goals in the UK and other countries such as the USA. The problems are magnified when questions of race and/or immigration status limit the extent to which teenagers (or any other ‘target groups’) may access and benefit from bridging social capital (see for example Geronimus, 2003; Almedom and Gosling, 2003). Mitchell and LaGory (2002) employ Putnam’s Social Capital and Community Benchmark Survey and Mirowsky and Ross’ psychological distress scale to examine how individual level social capital and individual sense of mastery may avert mental distress in an impoverished ‘‘ghetto’’ setting in Birmingham, Alabama. The authors report strong bonding ties within community and weak bridging ties to other groups: 71% of the study participants, pre-dominantly African American, trusted their neighbors, while 32% reported trust in people in general. Women and the unemployed experienced greater numbers of economic and environmental stressors. According to Mitchell and LaGory, bonding social capital significantly increased mental distress, and individual sense of mastery played a more important role than social capital: those with lower levels of mastery experienced more mental distress. It is likely that social cohesion would enhance mastery in individuals and thereby promote collective resilience in the face of socio-economic adversity and absence of bridging social capital. However, the authors appear to ‘blame the victim’ by implying that their study participants’ cooperation with them could have been transferred to social action on the part of the study participants in order to solve social problems. It is possible that the researchers were viewed (by the respondents) as possible links between the community in distress and external structures of power. Other studies have shown that Birmingham, Alabama is among the cities where impoverished as well as betteroff Black neighborhoods demonstrate high levels of political participation (see for example Portney & Berry, 1997). Trudy Harpham and her team (Harpham et al., 2004) developed, tested and validated an adapted form of Sampson et al.’s social capital measurement scale (Sampson et al., 1997) prior to its application in a South American city. Their study builds on earlier work on critical questions of meaning and measurement of social capital and mental health (Harpham et al., 2002; Harpham & Blue, 1995), and is the first of a series of international investigations with promise to contribute towards building evidence-base for mental health and social capital programming in resource poor countries. Harpham et al. (2004) conclude that in the presence of 957 violence, social capital, namely, trust, is not as closely associated with mental health as is socio-economic status, specifically, poverty and unemployment. The distinction between thin and thick trust helps to dissociate personal from structural stressors; however, it is not surprising that in a setting where crime and political violence are widespread, bonding social capital may accrue negative effect on mental health, and may even serve to perpetuate conflict in the absence of, or due to breakdown in bridging social capital. Nevertheless, Harpham et al. found that only 24% of their study participants were ‘‘probable cases of mental ill health’’ and only ‘‘13% of the youth admitted considering suicide in the last month’’ (p. 2272). This may not be as ‘‘disturbing’’ as Harpham et al. suspect, given that a large majority (84%) did not report suicidal ideation, the exact meaning and significance of which is unknown for this sample. This is not to understate the nature of the problem, but to emphasize the need for integration of socio-culturally specific context data to illuminate the statistical analysis, as Harpham’s team appreciate (Rabelo, Alves, & Souza, 1995; Scheper-Hughes, 1992). Similarly, Harpham et al. report of absence of (statistically significant) effect of social capital on mental health may not necessarily be ‘evidence of absence’ (see Alderson, 2004). Harpham et al. (2004) may be constrained by their aim to show measurable effect, in the absence of qualitative ‘‘thick descriptions’’ the relevance of which they recognize (Harpham et al., 2002). Interestingly, Patricia Wilson’s study (Wilson, 1997) which employed participatory action research methods had concluded that ‘‘social capital [in Cali, Colombia] will not be built through social engineering by technical experts. By its nature, it is being built (and rebuilt) humbly in small increments by individuals stepping out of isolation, enjoying connectedness and taking responsibility for their public lives. Nevertheless, there is a role, an opportunity, for professionals who work with people whether in the public, private or nonprofit sector to become catalysts of productive social capital.’’ (p. 758) This is clearly relevant for policy and practice implications of Harpham et al.’s study (Harpham et al., 2004). Taken together, the evidence from Russia (Rose, 2000) and London, England (Steptoe & Feldman, 2001), England and Wales (Moffitt et al., 2002), Alabama, USA (Mitchell & LaGory, 2002), Cali, Colombia (Harpham et al., 2004), confirm earlier research reports showing more reports of depression in women compared to men; implicating social support (giving and receiving differentials) and gender specific economic and social inequalities (see Brown & Harris, 1978; Aneshensel, Estrada, Hansell, & Clark, 1987; Aneshensel, Frerichs, & Clark, 1981; Antonucci & Akiyama 1987; Dohrenwend, Levav, & Shrout, 1992; Pevalin & Goldberg, 2003). Randomized controlled trials have also confirmed ARTICLE IN PRESS 958 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 that social intervention aimed at treatment of depression may be more effective than medical intervention (Harris, Brown, & Robinson, 1999a, b). Building and/or strengthening bonding as well as bridging social capital is therefore salutary for mental health. Social capital and senior citizens’ mental and emotional wellbeing Cotterill and Taylor’s evaluation of Plymouth Health Action Zone’s ‘‘Ageing Well (AW)’’ project (Cotterill & Taylor, 2001) comprises a qualitative study of a portion of a complex inter-sectoral, multi-agency government supported initiative to build social capital. Health Action Zones (HAZ) are area-based British government-initiated interventions to tackle health inequalities and social exclusion, with explicit mandate to build social capital. Policy analysts and practitioners have expressed both support for and concern over the prospects of evaluating such complex initiatives with compound structural and functional opportunities and challenges (Higgins, 1998; Jacobson & Yen, 1998; Powell & Moon, 2001). In response, a national HAZ evaluation commissioned to examine successes and failures of all 26 HAZ all in England with the exception of one in northern Ireland) had proposed combined ‘‘Theories of Change’’ and ‘‘Realistic Evaluation’’ models of evaluation (see Judge, 2000). These did not incorporate specific measures of social capital. Moreover, one of the challenges to local evaluation design has been the absence of baseline data on pre-HAZ levels of social capital against which the success of targeted interventions can be measured. However, the health service modernization programme is said to be progressing steadily, and HAZs are currently in the process of relocating from local Health Authority to Primary Care Trust (PCT) settings in order to accomplish institutional ‘‘Whole Systems’’ change. It is worth noting here that HAZ funding timeframe of seven years may be too short to effect real change. As Putnam, Leonardi, and Nanetti (1993) observed from the Italian experiment, the development of effective democracy and meaningful civic engagement involves lengthy processes of public discussion, reasoning and decision-making for which government-led, time limited and funding-bound initiatives hardly allow. Cotterill and Taylor’s qualitative assessment of effectiveness of a social capital building intervention (2002) exposes the contradictory effects of dissemination of health information intended to empower senior citizens (which threatens their emotional wellbeing by introducing fear about their health) and building bonding social capital to reduce isolation and thereby promote mental health. Enabling senior citizens to generate bonding and bridging social capital in order to ‘‘manage health information’’ thrown at them by health professionals with whom they have unequal power relationships may indicate positive overall outcome. This study brings to the fore inherent problems in social engineering, namely, the contradictions of ‘empowerment’ and target-driven health promotion activities aimed at the production of statistically significant measurable results in time for local and/or general election campaigns. It is well known that social capital in terms of reciprocity, availability of social networks and access to social support involves delicate negotiations, time-intensive processes of social interaction and individually crafted balances between dependence and autonomy (see Liang, Krause, & Bennett, 2001; Krause, 1997; Antonucci, Fuhrer, & Jackson, 1990). External interventions may thwart more rather than enhance these salutary processes. The UK social and health modernization policy has set in train processes of decentralization and devolution of public health (Evans, 2003) which may serve to empower health workers at the expense of excluded groups for whom prospects of social inclusion and civic participation may be a long way away (see for instance Almedom & Gosling, 2003). While advances in operational research (OR) herald promise of real integration of participatory and crosscultural multimethod (Taket & White 1994, 1998, 2000; White & Taket, 1994), translation of research into action may be pie in the sky. Real improvements in health and social development are likely to progress at a slow and arduous pace as and when the poor and marginalized gain control over their own health and social welfare. As Alinsky (1965) had related clearly and incisively, ‘‘poverty of economy’’ and ‘‘poverty of power’’ makes for a virtually insurmountable double disadvantage unless the poor themselves have a direct say in political decision making; as the very institutions charged with fighting poverty may continue to sustain self serving institutions. In spite of progressive policy and practice reforms, tackling health inequalities and social exclusion in the UK faces the same obstacles that Alinsky observed in the USA. If the research process itself is considered to be public service in that researchers have a duty to seek to provide unbiased data and interpretations thereof, more needs to be done to ensure data quality. The studies reviewed above have used data gathering instruments with varying degrees of ‘measurement error’. There are great advantages as well as inherent problems with large scale survey design involving unsupervised data collection procedures such as for instance in the widely used method of postal questionnaire-based surveys. Relatively small-scale qualitative investigations can serve as checks and balances (see for instance, Cattell & Herring, 2002). It is of grave concern that well-established proponents of social capital research to whom key indicators and robust measures are attributed do not include qualitative and participatory methods in their ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 research plans. For example, suggestions by Sampson, Morenoff, and Gannon-Rowley (2002) that video recording ‘‘systematic social observation’’ data conducted by ‘‘observers’’ driving or riding in their sports utility vehicle (SUV) at 3–5 miles per hour down the streets of sampled Chicago neighborhoods, as a reliable (but expensive) method is alarming to anthropologists and community development field workers. Fundamental philosophical differences among researchers within and across disciplines and sub-disciplines in the medical and social sciences have important influences on data quality, methods of analysis and interpretation and ultimate assignment of meaning to research evidence. There is a need for these concerns to be part of current conversations in social epidemiology (see Little, 1998). Social capital and mental health service and care provision The former WHO Mental Health Division head Norman Sartorius’ valedictory appeal for social capital highlights a two-way process whereby efficient and effective mental health services help to build and/or strengthen social capital in the communities they serve, and are in turn built and strengthened by the social capital of service users (2002, 2003). Rosenheck, 959 Morrissey, and Lam (2001) and van der Linden et al. (2003) independently reinforce Sartorius’ views. Rosenheck et al. (2001) demonstrate effectively that structural bonding and bridging social capital in mental health and housing service integration ‘‘reflect the state of civic culture in the community at large.’’ (p. 701). This supports Sartorius’ argument (Sartorius, 2002) and is borne out by the findings of other studies (see Ahern & Hendryx, 2003; Hendryx, Ahern, Loverich, & McCurdy, 2002; Hendryx & Ahern, 2001). Similarly, van der Linden et al.’s report of children’s use of mental health services substantiates the view that deficit in social capital in the shared social environment contributes to increased exposure of children to mental health services. It could be surmised that measures of mental health service use by children and young people indicate ‘‘social deficit’’ rather than ‘‘social capital’’, but deficit models are considered stigmatizing and unhelpful for policy/ practice improvements (Kwame McKenzie, personal communication, 2004). It remains the duty of researchers to think through the research process to its end point of informing policy and practice responsibly. In summary, this evaluative review serves to derive from the findings a set of guidelines for interdisciplinary research aimed at unraveling the complex associations between social capital and mental health. What is Box 2 Locating anthropology among the disciplines and sub-disciplines currently engaged in interdisciplinary enquiry into Social capital and mental health. What is already known about this topic a Social capital is a multi-dimensional construct which embraces both social support and social cohesion among other social determinants of health in general and mental health in particular. b The search for coherence and consistency in meaning and measurement of social capital has been riddled with conceptual and methodological shortcomings. What this study adds a The results of this evaluative review point to the need to promote interdisciplinary investigation and analysis. b Insight of the practical needs and requirements of decision makers and practitioners in public health and social policy mandates integration of qualitative data with quantitative ones in the interest of building improved/meaningful ‘evidence-base’ for health policy and practice. ARTICLE IN PRESS 960 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 known so far about the associations between social capital and mental health is outlined herewith. Neighborhood safety is a function of informal social control, social cohesion and trust whereby prevention of vandalism and violent crime, parental active involvement in children’s and adolescents’ activity generates collective efficacy. Residents’ sense of physical and mental or emotional wellbeing cannot be disaggregated into separate categories or promoted by means of social in the absence of economic interventions. Attempts to evaluate effectiveness of interventions which aim to build social capital and promote mental health require research design that combines both qualitative and quantitative investigation and analysis. Such a study design would have to include not only spatial definitions of key terms such as neighborhood and community, but also investigation of the meanings assigned to them across time and space. Participatory methods of inquiry would serve to identify the most pertinent issues for investigation—i.e. defining research questions and framing them meaningfully through active involvement of study participants in the research process itself—and to enable interpretation of quantitative data to the nearest approximations of reality as experienced by a wide range of stakeholders—social epidemiology is the locus of interdisciplinary research and policy development in social capital and mental health, to which Anthropology is expected to make significant contributions—see Box 2. Democracy as we know it may not support the conditions which make public discussion and reasoning that are inclusive of the views of marginalized, excluded, or non-dominant groups as the economically powerful and politically best connected will continue to win votes. Health policy may also dictate evidence making rather than the other way round (see for example Davey Smith, Ebrahim, & Frankel, 2001), but researchers must still strive for public discussion and public reasoning in order to facilitate learning and understanding the issues at hand, lest stagnation sets in (Sen, 2004). The ultimate challenge is for researchers to devise robust methods and tools of integrating quantitative and qualitative evidence and to increase the legitimacy and uptake of the latter by health and social policy makers and practitioners. Above all, unless eco-epidemiology integrates participatory, qualitative approaches, both protagonists and antagonists of social capital may be doomed to seemingly interminable crosspurpose talk. Acknowledgements This study was supported by the Henry R. Luce Professorship in Science and Humanitarianism and the Critical Thinking Program at Tufts University. The author wishes to thank her students Melissa Rosen and Jennifer Mendel for their participation in the literature survey. Ichiro Kawachi and Kwame McKenzie provided generous feedback on earlier stages of this review. Tufts Reference Librarian Regina Fisher Raboin gave invaluable logistical support. Helpful comments and suggestions were gratefully received from two anonymous reviewers. References Ahern, M., & Hendryx, M. S. (2003). Social capital and trust in providers. Social Science & Medicine, 57, 1195–1203. Alderson, P. (2004). Absence of evidence is not evidence of absence. British Medical Journal, 328, 476–477. Alinsky, S. D. (1965). The war on poverty—political pornography. Journal of Social Issues, 21, 41–47. Almedom, A. M., & Gosling, R. (2003). The health of young asylum seekers and refugees in the United Kingdom: reflection from research. In P. Allotey (Ed.), The health of refugees: public health perspectives from crisis to settlement (pp. 169–184). Melbourne: Oxford University Press. Aneshensel, C. S., Estrada, A. L., Hansell, M. J., & Clark, V. A. (1987). Social psychological aspects of reporting behavior: lifetime depressive episode reports. Journal of Health and Social Behavior, 28, 232–246. Aneshensel, C. S., Frerichs, R. R., & Clark, V. A. (1981). Family roles and sex differences in depression. Journal of Health and Social Behavior, 22, 379–393. Aneshensel, C. S., & Sucoff, C. A. (1996). The neighborhood context of adolescent mental health. Journal of Health and Social Behavior, 37, 293–310. Antonucci, T. C., & Akiyama, H. (1987). An examination of sex differences in social support among older men and women. Sex Roles, 17, 737–749. Antonucci, T. C., Fuhrer, R., & Jackson, J. S. (1990). Social support and reciprocity: A cross-ethnic and cross-national perspective. Journal of Social and Personal Relationships, 7, 519–530. Ardener, S., & Burman, S. (Eds.). (1995). Money-go-rounds: the importance of rotating savings and credit associations for women. Oxford: Berg. Bale, C. (2001). Befriending in cyberspace—challenges and opportunities. Crisis, 22(1), 10–11. Basaglia, F. (1980). Problems in law and psychiatry: the Italian experience. International Journal of Law and Psychiatry, 3, 17–37. Baum, F. (1999). Social capital, Is it good for your health? Journal of Epidemiology and Community Health, 53, 195–196. Baum, F. (2000). Social capital, economic capital and power, further issues for a public health agenda. Journal of Epidemiology and Community Health, 57, 409–410. Beaulieu, A., Morin, P., Provencher, H., et al. (2002). Le travail comme déterminant social de la santé pour les personnes utilisatrices des services de santé mentale (notes de recherché). Sante Mentale au Quebec, 27, 177–193. Berkman, L. F. (1995). The role of social relations in health promotion. Psychosomatic Medicine, 57, 245–254. ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 Berkman, L. F., Glass, T., Brissette, I., & Seeman, T. E. (2000). From Social integration to health: Durkheim in the new millennium. Social Science & Medicine, 51, 843–857. Beyers, J. M., Bates, J. E., Pettit, G. S., et al. (2003). Neighborhood structure, parenting processes, and the development of youths’ externalizing behaviors: a multilevel analysis. American Journal of Community Psychology, 31, 35–53. Bourdieu, P. (1986). The forms of capital. In J. Richardson (Ed.), Handbook of theory and research for the sociology of education (pp. 241–258). New York: Greenwood. Brown, G. W., & Harris, T. (1978). Social origins of depression: a study of psychiatric disorder in women. New York: Free Press. Bryson, L., & Mowbray, M. (1981). ‘Community’: The Sprayon Solution. Australian Journal of Social Issues, 16, 255–269. Campbell, C., & McLean, C. (2002). Ethnic identities, social capital and health inequalities: factors shaping AficanCaribbean participation in local community networks in the UK. Social Science & Medicine, 55, 643–657. Carlson, E. D., & Chamberlain, R. M. (2003). Social capital, health, and health disparities. Journal of Nursing Scholarship, 4, 321–325. Cattell, V. (2001). Poor people, poor places, and poor health: the mediating role of social networks and social capital. Social Science & Medicine, 52, 1501–1516. Cattell, V., & Herring, R. (2002). Social capital and well-being: Generations in an East London neighbourhood. Journal of Mental Health Promotion, 1(3), 8–19. Caughy, M. O., O’Campo, P. J., & Muntaner, C. (2003). When being alone might be better: neighborhood poverty, social capital, and child mental health. Social Science & Medicine, 57, 227–237. Chakraborty, A. P. U., & McKenzie, K. (2002). Does racial discrimination cause mental illness? British Journal of Psychiatry, 180, 475–477. Coleman, J. S. (1988). Social capital and the creation of human capital. American Journal of Sociology, 94, S95–S120. Cotterill, L., & Taylor, D. (2001). Promoting mental health and wellbeing amongst housebound older people. Quality in Ageing, 2, 32–46. Cowley, S. (1995). Health-as-process: a health visiting perspective. Journal of Advanced Nursing, 22, 433–441. Cowley, S. (1997). Public health values in practice, the case of health visiting. Critical Public Health, 7, 86. Cowley, S., & Billings, J. R. (1999). Resources revisited: salutogenesis from a lay perspective. Journal of Advanced Nursing, 29, 994–1004. Crystal, D. S., & DeBell, M. (2002). Sources of civic orientation among American youth: trust, religious valuation, and attributions of responsibility. Political Psychology, 23, 113–132. Dannenberg, A. L., Jackson, R. J., Frumkin, H., et al. (2003). The impact of community design and land-use choices on public health: a scientific research agenda. American Journal of Public Health, 93, 1500–1508. Davey Smith, G., & Ebrahim, S. (2001). Epidemiology—is it time to call it a day? International Journal of Epidemiology, 30, 1–11. Davey Smith, G., Ebrahim, S., & Frankel, S. (2001). How policy informs the evidence: ‘‘evidence based’’ thinking can 961 lead to debased policy making. British Medical Journal, 322, 184–185. Davey Smith, G., & Lynch, J. (2004). Commentary: social capital, social epidemiology and disease etiology. International Journal of Epidemiology, 33, 691–700. Davis, M. (1998). Ecology of fear. New York: Vintage. Dika, S. L., & Singh, K. (2002). Applications of social capital in educational literature: a critical synthesis. Review of Educational Research, 72, 31–60. Dohrenwend, B. P., Levav, I., Shrout, P. E., et al. (1992). Socioeconomic status and psychiatric disorders: the causation-selection issue. Science, 255, 946–952. Drevdahl, D. J. (2002). Home and border: the contradictions of community. Advanced Nursing Sciences, 24, 8–20. Drukker, M., Kaplan, C., Feron, F., et al. (2003). Children’s health-related quality of life, neighbourhood socio-economic deprivation and social capital: a contextual analysis. Social Science & Medicine, 57, 825–841. Edwards, B., & Foley, M. W. (1998). Civil society and social capital beyond Putnam. American Behavioral Scientist, 42, 124–139. Edwards, B., Foley, M. W., & Diani, M. (Eds.). (2001). Beyond tocqueville: civil society and the social capital debate in comparative perspective. Hanover and London: Tufts University, University Press of New England. Ellaway, A. (2004). Commentary: Can subtle refinements of popular concepts be put into practice? International Journal of Epidemiology, 33, 681–682. Evans, D. (2003). Taking public health out of the ghetto: the policy and practice of multi-disciplinary public health in the United Kingdom. Social Science & Medicine, 57, 959–967. Fowler, R. B. (1991). The Dance with Community. University Press of Kansas. Fullilove, M. T. (1998). Promoting Social Cohesion to Improve Health. Journal of American Medical Women’s Association, 52, 72–76. Gargiulo, M., & Benassi, M. (2000). Trapped in your own net? Network cohesion, structural holes, and the adaptation of social capital. Organization Science, 11, 183–196. Geronimus, A. T. (2003). Damned if you do: culture, identity, privilege, and teenage childbearing in the United States. Social Science & Medicine, 57, 881–893. Goodman, E., Huang, B., Wade, T. J., et al. (2003). A multilevel analysis of the relation of socioeconomic status to adolescent depressive symptoms: does school context matter? Journal of Pediatrics, 143, 451–456. Granovetter, M. S. (1973). The Strength of Weak Ties. American Journal of Sociology, 78, 1360–1380. Greenberg, G. A., & Rosenheck, R. A. (2003). Managerial and environmental factors in the continuity of mental health care across institutions. Psychiatric Services, 54, 529–534. Guest, A. M., & Wierzbicki, S. L. (1999). Social ties at the neighborhood level: two decades of GSS evidence. Urban Affairs Review, 35, 92–111. Hardin, G. (1968). The Tragedy of the Commons. Science, 162, 1243–1248 Reprinted 2003 in Science 302, 1243–1248. Harpham, T., & Blue, I. (Eds.). (1995). Urbanization and mental health in developing countries. Aldershot: Avebury. Harpham, T., Grant, E., & Rodriguez, C. (2004). Mental health and social capital in Cali, Colombia. Social Science & Medicine, 58, 2267–2277 (accessed October 2003). ARTICLE IN PRESS 962 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 Harpham, T., Grant, E., & Thomas, E. (2002). Measuring social capital within health surveys: key issues. Health Policy and Planning, 17, 106–111. Hawe, P. (1994). Capturing the meaning of ‘community’ in community intervention evaluation: some contributions from community psychology. Health Promotion International, 9, 199–210. Hawe, P. (1996). Needs assessment must become more changefocused. Australian and New Zealand Journal of Public Health, 20, 473–476. Hawe, P., & Shiell, A. (2000). Social capital and health promotion: a review. Social Science & Medicine, 51, 871–885. Harris, T., Brown, G. W., & Robinson, R. (1999a). Befriending as an intervention for chronic depression among women in an inner city. 1: randomized controlled trial. British Journal of Psychiatry, 174, 219–224. Harris, T., Brown, G. W., & Robinson, R. (1999b). Befriending as an intervention for chronic depression among women in an inner city. 2: role of fresh-start experiences and baseline psychosocial factors in remission from depression. British Journal of Psychiatry, 174, 225–232. Health First (Community Health South London) (1999). Health Needs and Perceptions of Young People in Lambeth Report prepared by Carrick James market Research, June 1999, London. Henderson, S., & Whiteford, H. (2003). Social capital and mental health. Lancet, 362, 505–506. Hendryx, M. S., & Ahern, M. M. (2001). Access to mental health services and health sector social capital. Administration and Policy in Mental Health, 28, 205–218. Hendryx, M. S., Ahern, M. M., Loverich, N. P., & McCurdy, A. R. (2002). Access to health care and community social capital. Health Services Research, 31, 85–101. Higgins, J. (1998). HAZs warning. Health Service Journal, 24–25. Hill, N. E., & Herman-Stahl, M. A. (2002). Neighborhood safety and social involvement: associations with parenting behaviors and depressive symptoms among African American and Euro-American mothers. Journal of Family Psychology, 16, 209–219. Hunter, E. (1998). Ways of seeing: changing directions in the health of indigenous populations. Clinical Child Psychology & Psychiatry, 3, 519–530. Jackson, L. E. (2003). The relationship of urban design to human health and condition. Landscape and Urban Planning, 64, 191–200. Jacobson, B., & Yen, L. (1998). Health action zones offer the possibility of radical ideas which need rigorous evaluation. British Medical Journal, 316, 164. Jewkes, R., & Murcott, A. (1996). Meanings of Community. Social Science & Medicine, 43, 555–563. Judge, K. (2000). Testing evaluation to the limits: the case of English Health Action Zones. Journal of Health Service Research Policy, 5, 3–5. Kawachi, I., & Berkman, L. F. (2001). Social ties and mental health. Journal of Urban Health-Bulletin of the New York Academy of Medicine, 78, 458–467. Kawachi, I., & Kennedy, B. P. (1999). Income inequality and health: pathways and mechanisms. Health Services Research, 34, 215–227. Kawachi, I., Kim, D., Coutts, A., & Subramanian, S. V. (2004). Commentary: reconciling the three accounts of social capital. International Journal of Epidemiology (accessed). Kelleher, C. C. (2003). Mental health and ‘‘the Troubles’’ in Northern Ireland: implications of civil unrest for health and wellbeing. Journal of Epidemiology and Community Health, 57, 474–475. Kennedy, B., Kawachi, I., & Brainerd, E. (1998). The role of social capital in the Russian Mortality Crisis. World Development, 26, 2029–2043. Klinenberg, E. (2002). Heat wave: a social autopsy of disaster in Chicago. Chicago: University of Chicago Press. Krause, N. (1997). Anticipated support, received support, and economic stress among older adults. Journal of Gerontology: Psychological Sciences, 52B, 284–293. Krieger, N. (1994). Epidemiology and the web of causation: Has anyone seen the spider? Social Science & Medicine, 39, 887–903. Krishna, A. (2002). Active social capital: tracing the roots of development and democracy. New York: Columbia University Press. Larson, R. W. (2000). Toward a psychology of positive youth development. American Psychologist, 55, 170–183. Leeder, S., & Dominello, A. (1999). Social capital and its relevance to health and family policy. Australian and New Zealand Journal of Public Health, 23, 424–429. Lerner, R. M. (2000). Developing civil society through the promotion of positive youth development. Developmental and Behavioral Pediatrics, 21, 48–49. Liang, J., Krause, N. M., & Bennett, J. M. (2001). Social Exchange and Well-Being: Is Giving Better Than Receiving? Psychology & Aging, 16, 511–523. Lindström, M., Hanson, B. S., Ostergren, P., et al. (2000). Socioeconomic differences in smoking cessation: the role of social participation. Scandinavian Journal of Public Health, 28, 200–208. Lindström, M., Hanson, B. S., & Oestergren, P. (2001). Socioeconomic differences in leisure-time physical activity: the role of social participation and social capital in shaping health related behaviour. Social Science & Medicine, 52, 441–451. Lindström, M., Merlo, J., & Östergren, P.-O. (2003). Social capital and sense of insecurity in the neighborhood: a population-based multilevel analysis in Malmö, Sweden. Social Science & Medicine, 56, 1111–1120. Little, M. (1998). Assignments of meaning in epidemiology. Social Science & Medicine, 47, 1135–1145. Lomas, J. (1998). Social capital and health: implications for public health and epidemiology. Social Science & Medicine, 47, 1181–1188. López, S. R. (2003). Reflections on the Surgeon General’s Report on Mental Health, Culture, Race, and Ethnicity. Culture, Medicine and Psychiatry, 27, 419–434. Lynch, J., Due, P., & Muntaner, C. (2000). Social capital—Is it a good investment strategy for public health? Journal of Epidemiology and Community Health, 54, 404–408. Macinko, J., & Starfield, B. (2001). The utility of social capital in research on health determinants. Milbank Quarterly, 79, 387–427. Manson, S. M. (2003). Extending the boundaries, bridging the gaps: crafting mental health, culture, race and ethnicity, a ARTICLE IN PRESS A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 supplement to the surgeon general’s report on mental health, culture, race, and ethnicity. Culture, Medicine and Psychiatry, 27, 395–408. McDowell, I., Spasoff, R. A., & Kristjansson, B. (2004). On classification of population health measurements. American Journal of Public Health, 94, 388–393. McKenzie, K., Whitley, R., & Weich, S. (2002). Social capital and mental health. British Journal of Psychiatry, 181, 280–283. McPherson, K. (1998). Wider ‘causal thinking in the health sciences’. Journal of Epidemiology and Community Health, 52, 612–614. Mitchell, C. U., & LaGory, M. (2002). Social capital and mental distress in an impoverished community. City & Community, 1, 195–215. Moffitt, T. E. & the E-Risk Study Team. (2002). Teen-aged mothers in contemporary Britain. Journal of Child Psychology and Psychiatry, 43, 727–742. Morabia, A. (1998). Epidemiology and bacteriology in the 1900: who is the handmaid of whom? Journal of Epidemiology and Community Health, 52, 616–617. Morrow, V. (1999). Conceptualizing social capital in relation to the well-being of children and young people: a critical review. Sociological Review, 47, 744–765. Morrow, V. (2001). Using qualitative methods to elicit young people’s perspectives on their environments: some ideas for community health initiatives. Health Education Research, 16, 255–268. Moss, N. E. (2002). Gender equity and socioeconomic inequality: a framework for the patterning of women’s health. Social Science & Medicine, 54, 649–661. Muntaner, C. (2004). Commentary: social capital, social class, and the slow progress of psychosocial epidemiology. International Journal of Epidemiology, 33, 674–680. Muntaner, C., Lynch, J., & Davey Smith, G. (2001). Social capital, disorganized communities, and the third way: understanding the retreat from structural inequalities in epidemiology and public health. International Journal of Health Services, 31, 213–237. Navarro, V. (2004). Commentary: Is capital the solution or the problem? International Journal of Epidemiology, 33, 672–674. Pearce, N. (1999). Epidemiology as a population science. International Journal of Epidemiology, 28, S1015–S1018. Pearce, N., & Davey Smith, G. (2003). Is social capital the key to inequalities in health? American Journal of Public Health, 93, 122–129. Pevalin, D. (2003). More to social capital than Putnam. British Journal of Psychiatry, 182, 172–173. Pevalin, D. J., & Goldberg, D. P. (2003). Social precursors to onset and recovery from episodes of common mental illness. Psychological Medicine, 33, 299–306. Pilkington, P. (2002). Social capital and health: measuring and understanding social capital at a local level could help to tackle health inequalities more effectively. Journal of Public Health Medicine, 24, 156–159. Pretty, J. (2003). Social capital and the collective management of resources. Science, 302, 1912–1914. Portes, A., & Landolt, P. (1996). The downside of social capital. American Prospect, 26, 18–21 94. 963 Portney, K. E., & Berry, J. M. (1997). Mobilizing minority communities: social capital and participation in urban neighborhoods. American Behavioral Scientist, 40, 632–644. Powell, M., & Moon, G. (2001). Health action zones: the ‘third way’ of a new area-based policy? Health and Social care in the Community, 9, 43–50. Putnam, R. (1995). Bowling alone: America’s declining social capital. Journal of Democracy, 6, 65–78. Putnam, R. D. (2000). Bowling alone: the collapse and revival of American community. New York: Simon & Schuster. Putnam, R. D., Leonardi, R., & Nanetti, R. Y. (1993). Making democracy work, civic traditions in modern Italy. New Jersey: Princeton University Press. Rabelo, M. C. M., Alves, P. C. B., & Souza, I. M. A. (1995). The many meanings of mental illness among the urban poor in Brazil. In T. Harpham, & I. Blue (Eds.), Urbanization and mental health in developing countries (pp. 167–192). Aldershot: Avebury. Rose, R. (2000a). How much does social capital add to individual health? A survey study of Russians. Social Science & Medicine, 51, 1421–1435. Rose, R. (2000b). Uses of social capital in Russia: modern, pre-modern, and Anti-modern. Post-Soviet Affairs, 16, 33–57. Rose, R. (2001). When government fails: social capital in an antimodern Russia. In B. Edwards, M. W. Foley, & M. Diani (Eds.), Beyond tocqueville (pp. 56–69). Hanover and London: Tufts University, University Press of New England. Rosenheck, R., Morrissey, J., Lam, J., et al. (2001). Service delivery and community: social capital, service systems integration, and outcomes among homeless persons with severe mental illness. Health Services Research, 36, 691–710. Ross, C. E. (2000). Neighborhood disadvantage and adult depression. Journal of Health and Social Behavior, 41, 177–187. Ross, C. E., & Mirowsky, J. (1999). Disorder and decay: the concept and measurement of perceived neighborhood disorder. Urban Affairs Review, 34, 412–432. Ross, C. E., Reynolds, J. R., & Geis, K. J. (2000). The contingent meaning of neighborhood stability for residents’ psychological well-being. American Journal of Social Review, 65, 581–597. Saegert, S., & Evans, G. W. (2003). Poverty, housing niches, and health in the United States. Journal of Social Issues, 59, 569–589. Saegert, S., Winkel, G., & Swartz, C. (2002). Social capital and crime in New York city’s low income housing. Housing Policy Debate, 13, 189–226. Saguaro Seminar. (2000). The social capital benchmark survey. Cambridge, MA: Harvard University, Kennedy School of Government. Sampson, R. J., Morenoff, J. D., & Gannon-Rowley, T. (2002). Assessing ‘‘neighborhood effects’’: social processes and new directions in research. Annual Review of Sociology, 28, 443–478. Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: a multilevel study of collective efficacy. Science, 277, 918–924. Sartorius, N. (2002). Fighting for mental health. Cambridge: Cambridge University Press. ARTICLE IN PRESS 964 A.M. Almedom / Social Science & Medicine 61 (2005) 943–964 Sartorius, N. (2003). Social capital and mental health. Current Opinion in Psychiatry, 16, S101–S105. Sayce, L. (2000). From psychiatric patient to citizen: overcoming discrimination and social exclusion. London and St. Martin’s Press, New York: Macmillan. Scheper-Hughes, N. (1992). Death without weeping: the violence of everyday life in Brazil. Berkley: University of California Press. Scheper-Hughes, N., & Lovell, A. M. (1986). Breaking the circuit of social control: Lessons in public psychiatry from Italy and Franco Basaglia. Social Science & Medicine, 23, 159–178. Sen, A. (2004). Feminism, human rights and development. Lecture delivered at Tufts University, April 28, Medford. Sharpley, M., Hutchinson, G., McKenzie, K., et al. (2001). Understanding the excess of psychosis among the AfricanCaribbean population in England, Review of current hypotheses. British Journal of Psychiatry, 178, s60–s68. Shy, C. (1997). The failure of academic epidemiology: witness for the prosecution. American Journal of Epidemiology, 145, 479–484. Social Exclusion Unit (1999a). Teenage Pregnancy Report presented to Parliament by the Prime Minister by Command of Her Majesty (June 1999), London. Social Exclusion Unit (1999b). Bridging the Gap: New opportunities for 16–18 year olds not in education, employment or training London. Steptoe, A., & Feldman, P. J. (2001). Neighborhood problems as sources of chronic stress: development of a measure of neighborhood problems, and associations with socioeconomic status and health. Annals of Behavioral Medicine, 23, 177–185. Stevenson, H. C. (1998). Raising safe villages: culturalecological factors that influence the emotional adjustment of adolescents. Journal of Black Psychology, 24, 44–59. Stevenson, H. C., Jr. (1997). ‘‘Missed, dissed, and pissed’’: making meaning of neighborhood risk, fear and anger management in urban black youth. Cultural Diversity and Mental Health, 3, 37–52. Susser, M., & Susser, E. (1996). Choosing a future for epidemiology: I. Eras and paradigms. American Journal of Public Health, 86, 668–673. Susser, M., & Susser, E. (1996). Choosing a future for epidemiology: II. From black box to Chinese boxes and eco-epidemiology. American Journal of Public Health, 86, 674–678. Szasz, T. (2003). The psychiatric protection order for the ‘‘battered mental health patient’’. British Medical Journal, 327, 1449–1451. Szreter, S. (2002a). Health, class, place and politics: social capital and collective provision in Britain. Contemporary British History, 16, 27–57. Szreter, S. (2002b). The state of social capital: bringing power, politics, and history back in. Theory and Society, 31, 573–621. Szreter, S., & Woolcock, M. (2004). Health by association? Social capital, social theory and the political economy of public health. International Journal of Epidemiology, 33, 650–667. Taket, A., & White, L. (1994). Doing community operational research with multicultural groups. Omega, International Journal of Management Science, 22, 579–588. Taket, A., & White, L. (1998). Experience in the practice of one tradition of multimethodology. Systemic Practice and Action Research, 11, 153–168. Taket, A., & White, L. (2000). Partnership & Participation. Chichester: Wiley. Tumwine, J. K. (1989). Community participation as myth or reality: a personal experience from Zimbabwe. Health Policy and Planning, 4, 157–161. van der Linden, J., Drukker, M., Gunther, N., et al. (2003). Children’s mental health service use, neighbourhood socioeconomic deprivation, and social capital. Social Psychiatry & Psychiatric Epidemiology, 38, 507–514. Vimpani, G. (2000). Child development and civil society—does social capital matter? Developmental and Behavioral Pediatrics, 21, 44–47. Walker, A. M. (1997). ‘‘Kangaroo Court’’: invited commentary on Shy’s ‘‘The failure of academic epidemiology: witness for the prosecution’’. American Journal of Epidemiology, 145, 485–486. Wall, S. (1999). Epidemiology in transition. International Journal of Epidemiology, 28, S1000–S1004. Weitzman, E. R., & Kawachi, I. (2000). Giving means receiving: the protective effect of social capital on binge drinking on college campuses. American Journal of Public Health, 90, 1936–1939. White, L., & Taket, A. (1994). The death of the expert. Journal of Operational Research Society, 45, 733–748. Wilkinson, R. G. (1996). Unhealthy societies: the afflictions of inequality. London: Routledge. Wilson, P. (1997). Building social capital: a learning agenda for the twenty-first century. Urban Studies, 34, 745–760. World Health Organization. (2001). World Health Report, Mental Health, New Understanding, New Hope. Geneva: WHO. Yang, M. J., Yang, M. S., Shih, C. H., et al. (2002). Development and validation of an instrument to measure perceived neighbourhood quality in Taiwan. Journal of Epidemiology & Community Health, 56, 492–496.