A noted scientist on the radio during Science Week suggested that we are moving from an age of physics to an age of biology - from quantum mechanics to genetic engineering. In the same vein, it would seem that social science is witnessing a move from an age of economics to an age of sociology - from rational markets to civil societies. While simple characterisations such as these rarely stand up to rigorous inquiry, there is clearly growing political and policy concern about the sustainability of community networks and social civility in many advanced western democracies today. The fragility of civil society in Australia has been pinpointed by falling levels of trust, of individuals and of institutions (Hughes, Bellamy and Black 1998), and by a falling participation rate in voluntary work (Lyons and Fabiansson 1998).

There is now mounting international evidence that social relations of a particular quality and nature are central to creating sustainable communities. Social relationships, which are characterised by high degrees of mutual trust and reciprocity, are argued to sustain better outcomes in the economy, democracy and civil society. These sorts of social relationships are said to be laden with social capital - the norms and networks that enable people to act collectively.

Few would disagree that an ebbing civic spirit is somehow tied up with the fundamental changes occurring in family life. Yet few have actually examined how family life is changing to affect this civic spirit. Within the social capital literature it is generally assumed that the mere practice of family life will generate norms of trust and reciprocity but the actual circumstances under which this takes place are under researched. Community life, rather than family life, has typically been the focus of social capital researchers. This work has pointed to the 'decline of community' (a long-standing theme in classical sociology) but has tended to assume that family life has been unchanging. It is as if the colloquial saying 'If you can't trust your family, who can you trust', has been glibly accepted as social scientific fact.

Empirical change in family life has been ignored in part because examination of family life and social capital has not been set within a theorised framework of social change. This paper aims to review the conceptual definition of social capital and to examine understanding of the relation between family life and social capital - first, as it occurs within family networks, and second as it bridges from family networks beyond into the community. Arguing that these linkages are under-researched and under-theorised, this paper concludes by drawing upon Giddens's and Beck's notions of post-traditional society and risk society respectively, to outline a framework for understanding the changing nature of family life and how it may be linked to the ebb and flow of social capital.

The lineage of the concept of social capital has been traced back as far as 1916, to an article by L.F. Hanifan, a West Virginia school reformer (Woolcock 1998). Although the idea of social capital is central to sociology's inquiry into the nature of social order (Wall et al. 1998: 303), the concept does not reappear, per se, until Jane Jacobs' work on urban planning in the 1960s. North American social science in the 1970s also witnessed passing reference to social capital in the work of economist Glenn Loury (1977), as he critiqued neoclassical theories of racial income inequality (Portes 1998: 2). But it is James Coleman's (1988) work on school participation in Chicago, that is credited as being the spark for much of the contemporary debate, which was then fuelled significantly in the 1990s by Putnam's (1993) study of democratic institutions in Italy. Whilst Coleman's work (1988, 1990) has been the foundation for much of the recent conceptual debate, particularly in North America, there has also been something of a rediscovery of Bourdieu's (1986) work on social capital within the European sociological theatre.

Bourdieu on social capital

For Bourdieu, 'Capital can present itself in three fundamental guises: as economic capital, which is immediately convertible into money and may be institutionalised in the form of property rights; as cultural capital, which is convertible, on certain conditions, into economic capital and may be institutionalised in the form of educational qualifications; and as social capital, made up of social obligations ('connections'), which is convertible, in certain conditions, into economic capital and may be institutionalised in the form of a title of nobility.' (1986: 243).

That social capital may be institutionalised in the form of a good family name, a title, does not mean that it is the preserve of the 'well to do'. Social capital, as a network of connections, is not a natural given or a social given, but something that must be worked for on an ongoing basis. It 'is the product of investment strategies, individual or collective, consciously or unconsciously aimed at establishing or reproducing social relationships that are directly useable in the short or long term' (Bourdieu 1986: 251). These relationships can be in the neighbourhood, the workplace, or amongst kin.

Bourdieu defines social capital as 'the aggregate of the actual or potential resources which are linked to . . . membership in a group - which provides each of its members with the backing of the collectivity owned capital' (1986: 249). The emphasis is upon social networks that provide access to a group's resources. The outcome of this social capital is ultimately economic reward reaped through ongoing participation in the network as mutual benefits accrue. Social capital is therefore a means, through social connections, to resources which are keenly sought in capitalist societies.

Bourdieu's particular application of the concept social capital (note, not his definition of the concept) relates to understanding how individuals draw upon social capital to improve their economic standing in capitalist societies. In such societies, Bourdieu argues, economic capital is the fundamental resource and his concern is with how social capital, and cultural capital, may be instrumental in increasing an individual's economic capital. It is important to note that social capital is conceptualised as a resource to action with increased economic capital the outcome of actually drawing upon that resource. This conceptual distinction between social capital as a resource to action and the outcomes of that resource is an important one that, as we shall see below, has become blurred in the 'measurement rush'.

Coleman on social capital

In comparison with Bourdieu, Coleman uses different terms to define social capital, uses the concept for a different purpose, and works within a different theoretical tradition, but fundamentally constructs the same theoretical concept.

Coleman defines social capital, not by what it is, but by what it does, or by its function. 'The function identified by the concept of 'social capital' is the value of these aspects of social structure to actors as resources that they can use to achieve their interests' (Coleman 1988: S101). This definition is effectively the same as Bourdieu's connections or group memberships (Coleman's 'aspects of social structure') that provide access to a group's resources for each member of the group (Coleman's 'resources that they use to achieve their interests'). Again social capital is defined as a resource - it is a stock, a store, a basis from which other actions, collective actions, might be undertaken to meet mutual interests.

The 'aspects of social structure' Coleman (1988) refers to comprise obligations and expectations, information channels, norms and effective sanctions that constrain and or encourage certain kinds of behaviour and these 'exist in the relations among persons' (1988: S100-101). 'If A does something for B and trusts B to reciprocate in the future, this establishes an expectation in A and obligation on the part of B.' (Coleman 1988: S102). Information is said to be important as a basis for action and norms, and effective sanctions are said to facilitate certain forms of action and constrain others. A feature of social structures, rather than social relations, said to be important to the generation of social capital is closure, said by Coleman to mean sufficient ties among members of a group to ensure the observance of the group's norms.

Coleman's application of the concept social capital, again not his definition, is concerned to understand the role of norms and sanctions, within family and community networks, that facilitate the attainment of human capital. He is particularly concerned to understand variation in the educational achievements of high school children and the role that family support plays in this. In this application of the concept, Coleman is very close to Bourdieu (Portes 1998: 5; Harriss and De Renzio 1997: 921). Both see social capital as a means to increasing an individual's resources. If Bourdieu is interested in social capital as a resource to economic capital for individuals, in a range of social settings, Coleman is interested in how social capital in family and community networks, is a resource to human capital for individuals.

Putnam on social capital

A third key author in recent social capital debates is Putnam (1993, 1993a, 1995, 1996, 1998). Putnam is interested in how social capital works at the regional level to shore up democratic institutions and economic development. Putnam operationalises the concept of social capital at a different social scale to both Bourdieu and Coleman, although his definition of the concept social capital is drawn directly from Coleman.

Putnam defines social capital as 'trust, norms and networks' that facilitate cooperation for mutual benefit (1993: 167). Although the terminology again shifts somewhat, the definition effectively follows Coleman's (and therefore Bourdieu's) notion of social capital being those networks and norms that facilitate collective action. Norms of trust and reciprocity within networks are the capital resources, which are inherently social, the outcomes of which are various forms of collective action. Social capital is not collective action per se, but rather the norms and sanctions of trust and reciprocity residing within social networks, that enable dilemmas of collective action to be solved.

Whilst Coleman states that this facilitation by social capital could result in outcomes for individuals or in 'system-level behaviours' (1988: S101), his own empirical work was focused at the individual level on human capital. Putnam's focus, though, is upon the system level behaviours as he is concerned to explain economic and political development at regional and national levels. This focus upon outcomes for regions and nations distinguishes Putnam's work empirically from that of Coleman and Bourdieu, but not in terms of the fundamental definition of the concept social capital. It is simply that the concept is applied at a broader social scale.

The aspect of social capital said by Putnam to result in differing regional or national political and economic outcomes, is the norm of generalised reciprocity that fuels social trust. This is the trust possessed by community members that their short-term altruistic actions that contribute to others' welfare, will be rewarded at some point in the future. Networks of civic engagement (neighbourhood associations, choral societies, cooperatives, sports clubs, mass-based parties, etc.) are an essential element of social capital as they foster robust norms of reciprocity. These norms in turn sanction those who do not reciprocate. Such sanctions become widely effective as the trustworthiness, or not, of a particular individual is communicated across overlapping social networks

Putnam's empirical measurement of social capital has been criticised for being reductionist, as it places undue weight upon participation in voluntary associations and assumes that reciprocity and trust will flow from this participation (Harriss and de Renzio 1997; Putzel 1997). A danger of this approach to measuring social capital is that the concept loses its specificity as a resource to action and becomes redefined, in the 'measurement rush', as the action itself (see Newton 1997). Harriss and de Renzio (1997) and Putzel (1997) also criticise Putnam for prescribing a linear causal link between civic engagement in voluntary associations and improved democratic institutions, rather than allowing for a dynamic interrelationship between the two. As a consequence they argue that Putnam wrongly excludes a role for democratic institutions, such as government, in the creation of social capital. 

Putnam's definition of the concept social capital may well be the same as Bourdieu's and Coleman's, but its application at a broader social scale does add complexity to the original formulation. Applying the concept of social capital to a regional or national scale raises the question of whether social capital has, necessarily, to result in outcomes that are mutually beneficial to all in the region or the nation - that is must social capital result in common good outcomes?

To this end Putzel (1997) argues the precise definition of social capital must be sensitive to analytically distinct elements. 'There is a need to distinguish carefully between what might be seen as the mechanics of trust and the political content and ideas transmitted through such networks and embodied in such norms. We can see how the mere existence of networks and norms underpinning trust between individuals or groups can facilitate exchange by reducing risks and making behaviour more predictable. But whether or not these networks will contribute to democracy has much more to do with the political ideas and programs transmitted through them.' (Putzel 1997: 941-942).

This distinction between the mechanics of social capital and the content of the ideas transmitted points to the notion that social capital may have negative effects for some parties - what is referred to as the 'dark side of social capital' (Putzel 1997). The notion of dark side outcomes is integral to the original formulation of social capital by Bourdieu and Coleman. They focussed upon how social capital enabled individuals to gain a competitive advantage, either in economic or human capital terms. Arguably a gain for some through the use of social capital is a loss, or potential loss, for others. Putnam's (1993) work on Italian regions has given us the oft-cited example of the mafia as a 'family' that draws heavily on particular forms of social capital to produce positive outcomes for those in the 'family' but potentially 'dark side' outcomes for those outside. Indeed, a recent foreword by Putnam to a journal's theme issue on social capital, identifies both 'public-regarding' and 'private-regarding' purposes in the use of social capital.

In fact, Putnam identifies six dimensions to social capital:

1. formality - there are both formal and informal types of civic engagement;
2. purpose - some institutions are public-regarding, some are private regarding;
3. bridging - bonds of trust and reciprocity can bridge cleavages in society or, conversely, bring like-minded or like-ethnic individuals together;
4. immediacy - trust may stem from immediate, face-to-face connections or generalised anonymous bonds;
5. strandedness/intensity - at one end of this spectrum are durable, intense, and multistranded networks (i.e., people know each other through multiple, overlapping networks), at the other are the weaker, more fleeting bonds that might be created from a day of volunteering together; and
6. social location - neighbourhood ties represent the place-based end of the social capital spectrum, while Internet groups represent the function-based end. (Putnam 1998: vi).

Each of these conceptual dimensions is important in considering the role of family life in the construction of social capital. Important because they acknowledge the different social groups and scales within which social capital might rest. Thus, there is not only formal civic engagement through voluntary organisations but informal civic engagement amongst kin, friends and neighbours. Stocks of family-based social capital are perhaps more likely to be 'private regarding' than 'public regarding', focusing upon the welfare of family members. Family-based social capital is likely to be an important aspect of family bonding and stability, but under what conditions does this social capital become a resource to 'bridge' beyond the family networks? The dimension of 'immediacy' recognises the difference between trust of familiars, of family members, and trust of anonymous strangers. Family networks are likely to be durable and intense within built-in obligations for regular contact. Finally, 'social location' needs to extend from the household to the neighbourhood and as far as the virtual world of the internet.

Summary of key concepts of social capital

Theoretical specification of just what social capital is has progressed, then, to a stage of recognising multiple dimensions, which now awaits some much needed empirical confirmation. Table 1 provides a simple summary of some of the key conceptual contours of social capital. The work of Bourdieu, Coleman and Putnam is compared in relation to how they define social capital, the purpose for which they employ the concept, and the social scale at which they apply it. The definitions are clearly overlapping and, as argued above, effectively refer to the same social processes.

Social capital, then, is a resource to collective action. That resource comprises the norms and sanctions of trust and reciprocity that operate within social networks. The structural components of networks such as 'size', 'density' and the extent of 'closure' and relational aspects such as 'inequality' shape the social capital capacity of a network. The outcomes of the social capital within a network comprise a variety of forms and scales of collective action. Here lies the key difference in the work of Bourdieu, Coleman and Putnam - they employ the concept of social capital to understand different types of collective action and pitch their analyses at different social scales. Such differences in 'purpose' and 'analysis' should not be mistaken, however, for contradictory definitions of the concept social capital.

Source: Australian Institute of Family Studies, 2000.

Typically it is simply asserted that family life is a bed-rock of social capital. Putnam (1995: 73) states that 'the most fundamental form of social capital is the family'. Bourdieu (1993: 33) sees the family as the main site of accumulation and transmission of social capital. Fukuyama (1999: 17) asserts that 'families are obviously important sources of social capital everywhere'. Newton (1997: 579) states that 'the family may also be the most fundamental source of social capital'.

However, the nature of family life that supports social capital is at no point detailed by these authors. There is a sense in which the role of the family has been idealised 'as the most productive site of social capital and therefore a pillar of civic virtue and democracy' (Putzel 1997: 945). The nature of this idealisation is that it is assumed that families will provide models of good relationships and civic virtues (Cox 1995: 28-29). The feminist critique of 'the family' as a power structure imbued with fundamental gender inequalities is ignored. That family practices and the norms that govern them may well be changing and that these changes might underlie an ebbing civic spirit has largely escaped attention.

An extension of the view of the family as the bedrock of social capital is that of the family as antithetical to social capital in the community at large. In this view family life is 'familist', meaning 'the elevation of family and kinship ties above other sorts of social obligations' (Fukuyama 1999: 37). In these circumstances family ties become too strong, and crowd out the weaker ties of community. Allegiance is to family and kin and trust does not extend beyond the bounds of the family. 'In some cases, there appears to be something of an inverse relationship between the bonds of trust and reciprocity inside and outside the family; when one is very strong, the other tends to be weak.' (Fukuyama 1999: 17-18)

That the relatively 'weak' ties of community may be more supportive of a vibrant civil society than the strong ties of the family, is a point first theorised in Mark Granovetter's (1973) The strength of weak ties. This position is also employed by Putnam (1993) to support his argument that horizontal relationships of association are more effective at generating social capital than strong ties of friendship or family - although this does seem to contradict his statement, cited above, that 'the most fundamental form of social capital is the family' (1995: 73) (see Harriss and de Renzio 1997: 924). In Australian work on social capital Cox (1995: 31) and Latham (1997: 13) each make the point that family life may limit world-views and sociability. Secondary associations such as sports clubs and mutual aid societies, they suggest, are likely to have a greater social utility from a social capital perspective.

Somewhat ironically, the conservative writer Fukuyama, who labels rising divorce and illegitimate birth rates 'The Great Disruption', finds some solace in these trends, as a solution to the problem of familism. 'Although the breakdown of a family in itself constitutes a loss of social capital, this breakdown may actually lead some family members to greater levels of association with people and groups outside the family . . . it is possible that the weakening of bonds within contemporary western families leads to an increase in social ties outside the family.' (Fukuyama 1999: 117-118)

Overall the social capital literature tends to under emphasise the role of families in constructing social capital, be this only within family networks or beyond to community networks. Instead it tends to over emphasise the comparatively small amount of time individuals spend working for voluntary associations. This voluntarism, however, is unlikely to be the most important network for the development of norms of trust and reciprocity (Newton 1997: Levi 1996). As Norton (1997: 5-6) argues: 'If we want to create a more comprehensive theory of the relationship between civil society and social capital we cannot overlook institutions with far greater participation rates than voluntary associations: families, schools and firms. In all likelihood it is these institutions, rather than voluntary associations, that are the principal creators of social networks.'

To the extent that social capital and family life have been examined, it has been in two ways. The first examines the construction of social capital within family networks and the importance of this for children's developmental outcomes. Using Putnam's multi-dimensional definition of social capital, the focus has been upon bonding social capital within the informal, intense and durable, face-to-face connections of the household, resulting in private-regarding outcomes.

The second treatment examines the role of families in constructing social capital beyond family networks, where family-based social capital bridges into networks beyond the household, into the informal and formal arenas of economy, polity and community, where strangers are trusted and public regarding outcomes result. The latter is the dominant theme within the social capital literature, which contains an implicit hypothesis that to have strong communities, one must have strong families.

The following two sections examine each of these literatures and demonstrate the limited understanding that has been developed of family life and social capital.

Social capital within the family

One way in which the relationship between families and social capital has been investigated is by examining the importance of social capital, within the family, to a range of individual outcomes. Here the focus is upon bonding social capital within the informal, intense and durable, face-to-face connections of the household, resulting, typically, in private-regarding outcomes. Coleman (1988, 1988a) examines the importance of social capital within the family to children's education outcomes. Amato (1998) and Furstenberg (1995) range more broadly to examine the importance of social capital for children's cognitive and social development.

Coleman's (1988, 1988a) work shows how social capital within the family is important to education outcomes for children. The presence of social capital is said to be vital to the transfer of human capital from parents to children: 'If the human capital possessed by parents is not complemented by the social capital embodied in family relations, the human capital of the parents becomes irrelevant to the child's educational growth' (Coleman 1988a: 384). As a part of this analysis Coleman examines the aspects of family life that appear important to social capital. The social capital of the family is measured through the strength of the relations between parents and children (Coleman 1988: S110). This is said to depend upon the 'physical presence of adults in the family and on the attention given by adults to the child' (Coleman 1988: S111).

Coleman finds that the ratio of parents to children in the household, the extent of the family network, has an important effect upon the likelihood of a child dropping out of high school. A single parent with five children was more likely to witness one of those children dropping out of high school than a two-parent family with two children (Coleman 1988: S112). A second finding is that 'the frequency of talking to parents about personal experiences shows essentially no relation to dropping out' (Coleman 1988: S113). The third key finding is that the mother's expectation that the child would go to college, is associated with the likelihood of the child dropping out of high school.

A number of points need to be made with regard to the measures of social capital used by Coleman, and the subsequent findings. Two key components of social capital are networks, an objective component, and norms, a subjective component. Coleman measures the network within the family by examining the parent to child ratio. Whilst the analysis of parent-child ratios demonstrates a strong association between this variable and dropping out of high school, Coleman uses the extremes of this variable to make his point. He does so by comparing high school 'drop-out' rates of children from 'two-parent-two-child' families with those from 'one-parent-five-child' families. By using this measure rather than, say, a comparison of two parent-two-child families with two-parent-four-child families, Coleman confounds measurement of the parent-child ratio with all the associated characteristics of sole parenthood, such as lower education levels and poverty. Coleman elides between a parent-child ratio variable and a family structure variable. His statement that 'the most prominent element of structural deficiency in modern families is the single-parent family' (Coleman 1988: S111), implies that it is family structure and not the parent-child ratio per se that he wished to analyse.

To measure the strength of the norms of cooperation within the family, Coleman measures the quality of parent-child interaction. The indicator used by Coleman is the frequency of discussions with parents about personal matters - not a particularly proximate measure of the norms of trust and reciprocity within the family. The quality of parent-child interaction is effectively reduced to a quantitative indicator, ignoring the nature of the interaction. For example, it is unknown from this indicator whether the parent-child interaction took the form of an argument and represented a failure to communicate. So despite Coleman's avowed intention to measure the quality of parent-child interaction, the indicators he has available (constructed retrospectively), prevent him from doing so.

Coleman states: 'Even if adults are physically present, there is a lack of social capital in the family if there are not strong relations between children and parents' (Coleman 1988: S111). Given this, it is surprising that on finding no relationship between the frequency of discussions with parents about personal matters and high school drop out rates, that Coleman reverts to use of the simple parent-child ratio to demonstrate the importance of social capital within the family. This approach is clearly flawed, even on Coleman's own terms.

Regardless of the limitations of Coleman's analysis, his work is important in pointing to the possible variations in the relationship between family life and social capital. First, and quite simply, levels of social capital within families vary according to a range of factors. Second, Coleman identifies parent-child ratios as important in explaining variation in levels of social capital within families. Third, the values held by a family, in this particular instance the mother's expectation of the child going to college, can be an important indicator of the level of social capital within the family. Coleman also locates his analysis of social capital and family life within a context of change - the growth of sole-parent families - and is concerned to understand the implications of this for social capital, rather than simply assuming that the family is static and that if there is a social capital deficit its cause rests elsewhere.

Furstenberg and Hughes (1995) explore the relationship between measures of social capital and indicators of young adult success, through a longitudinal study of children of teenage mothers. Using 18 retrospectively created measures of social capital their results suggest that social capital helps youth negotiate their way out of disadvantage. In particular completion of high school, enrolment in college and socioeconomic status appear significantly related to at least half of the social capital measures used by Furstenberg and Hughes (1995: 586). Fertility, illegal activity and mental health are, however, unrelated.

Particular aspects of social capital, as Furstenberg and Hughes measured it, were found to be more important to youth outcomes than others. For example, 'family cohesion' and 'mother's relationship with her own mother' were consistently related to the youth outcomes, whilst 'contact with other kin' was not. Mother's educational aspiration for her child was consistently related to positive youth outcomes (Furstenberg and Hughes 1995: 587), yet the presence of a biological father or stepfather, is inconsistently related (Furstenberg and Hughes 1995: 589).

It is debatable to what extent Furstenberg and Hughes' (1995) measures of social capital are indeed that. Many of their retrospectively constructed 'measures' fail to capture the nature of social capital. They do not measure social capital per se - the extent of social networks and whether they are characterised by norms of trust and reciprocity - rather, most of their 'measures' are focused upon the potential outcomes of social capital - various forms of social action. At best these 'measures' are indicators of social capital, the validity of which needs to be established before they can confidently be used as indicators of the presence of social capital. It cannot be assumed that any instance of collective action is necessarily the outcome of social capital. Such action could have been mandated or forced through violence.

Furstenberg and Hughes confuse the potential outcomes of social capital with social capital per se. In fact only three of their 18 'measures' go to the nature of social capital: 'Father in home' as a measure of the size of the network within the household; 'Number of child's friends mother knows' as a measure of the overlapping nature of the family and child's friendship networks; and 'Strong help network' as a measure of the capacity of the network to help solve dilemmas of collective action.

The other 15 'measures' relate to the potential outcomes of social capital. For example, 'family cohesion' measures involvement and exchange with extended family. Social capital could well be the resource being drawn upon to enable such action, but the action itself is not social capital. It may indicate the presence of social capital, but the strength of this relationship needs to be tested before it can be accepted as an indicator of social capital.

Amato (1998) draws upon the concept of social capital to examine men's contributions to their children's lives. Amato uses the quality of the co parental relationship and the parent-child relationship as indicators of social capital within the family. Two key hypotheses drive his inquiry:

'The quality of the co-parental relationship (as reflected in the extent to which fathers model constructive dyadic skills, cooperate with mothers in childrearing, and are emotionally supportive of mothers) is positively associated with children's wellbeing. 

'The quality of the father-child relationship (as reflected in the level of support and effective control exercised by fathers) is positively associated with children's wellbeing.' (Amato 1998: 247). 

Based upon literature review and primary data analysis Amato concludes that each of these hypotheses is supported. 'Studies of two-parent households suggest that fathers benefit children to the extent that they promote a positive marital relationship, and more specifically a positive co parental relationship . . . paternal support and effective control are good for children' (Amato 1998: 258). Via primary data analysis Amato (1998: 266) finds in relation to the co-parental relationship that marital discord (a proxy measure of co-parenting quality) had a direct negative effect on children's life satisfaction and self-esteem. With regard to the parent-child relationship, high levels of support from fathers lowered children's psychological distress and marital discord indirectly increased children's psychological distress.

In a critique of Amato's (1998) paper, Furstenberg (1998) makes the important point that social capital is a sociological construct, not a psychological one. This means that social capital is not reducible to the individual, as it only operates at a shared and collective level (Furstenberg 1998: 296). On this basis, Furstenberg levels the criticism that Amato 'treats the notion of social capital as a psychological property of the dyadic relationship between parent and child, not as a sociological property of the family as a collective social system with a set of social norms and cultural understandings' (Furstenberg 1998: 297). To pitch the application of social capital at the dyadic rather than family unit level is to fail to recognise the potential added understanding of the concept. The focus should be upon the extent to which the norms of trust and reciprocity are shared across the family unit to enable them to undertake particular forms of collective action. By simply re-labelling the quality of dyadic relationships as 'social capital', no new understanding is added to the existing psychological analysis. 

This misunderstanding and misuse of the concept of social capital is common to other analyses. Parcel and Menaghan (1993: 120) argue that 'social capital which inheres in the relations between parents and children should be associated with internalisation of social norms in children' and that 'family social capital promotes norm transmission across generations'. This framing of social capital fails to understand that social capital is not associated with norm internalisation and transmission, as if it were something separate from these processes. Social capital is the internalisation and transmission of particular norms.

Marjoribanks and Kwok (1998) pay similar disregard to the concept of social capital by labelling 40 items associated with parental activities ('My father/mother checks my school diary and schoolwork'; 'My father/mother supports me by attending activities at school') as social capital. These sorts of activities might well be the outcome of social capital within the family. That is, they might occur because of the norms of trust and reciprocity that operate within the family network. But they also might occur due to a complete lack of trust within the family, with the parents believing they have to monitor closely their child's school related activities. These items not only confuse the potential outcomes of social capital, forms of collective action, with social capital per se, but also mistakenly assume that these particular forms of collective action are indeed associated with social capital with no a priori establishment of the fact.

In summary, little work has been done on the circumstances associated with social capital within the family or on the consequences of this for family life. The work that has been done either misconceptualises social capital and/or resorts to using inadequate measures or indicators of social capital. Because of this the currently available findings provide an unclear picture of the construction and role of social capital within family networks. There is, however, some evidence that network size, as measured by parent-child ratios is an important variant of within family social capital. It is possible that such variations could affect social and psychological outcomes in children, though the current inaccuracy of measuring social capital means that this is not properly established.

Social capital beyond the family

A second way in which the relationship between families and social capital has been examined, focuses on what sorts of family circumstances strengthen social capital within civil society. That is, what patterns of family life are associated with robust norms of trust and reciprocity in social networks beyond the family.

In his earlier work Putnam tends to assume that change in the nature and pattern of family life are key to understanding falling levels of social capital: 'The most fundamental form of social capital is the family, and the massive evidence of the loosening bonds within the family (both extended and nuclear) is well known. This trend, of course, is quite consistent with - and may help to explain - our theme of social decapitalisation' (Putnam 1995: 73). Challenging his own assumptions about the role of family life in constructing social capital, Putnam later (1996) systematically reviews the evidence about the likelihood of a link between changing family circumstances and the decline of social capital in US civil society.

Putnam's measurement of social capital has two components. The extent of a norm of social trust is measured using the General Social Survey and World Values Survey question: 'Generally speaking, would you say that most people can be trusted or that you can't be too careful in dealing with people?' A second component, the strength of social networks, is measured by the number of formal group memberships. The family circumstances examined by Putnam, that might be associated with these measures of social capital, include education, economic hardship, residential mobility, time stress, female labour force participation and marriage.

Putnam finds that level of education is the key aspect of family/individual circumstances associated with greater social capital. Even when allowing for the effects of income and social status upon group memberships, education continues to be the primary factor. 'Well educated people are much more likely to be joiners and trusters, partly because they are better off economically, but mostly because of the skills, resources and inclinations that were imparted to them at home and in school' (Putnam 1996: 4-5). And although education level is a stronger predictor of civic engagement than income, economic hardship in the form of lower incomes and feeling financially strapped, holding education constant, does account for lower levels of social trust and civic engagement (Putnam 1996: 7). In an Australian context Hughes, Bellamy and Black (1998) confirm Putnam's finding that higher levels of education are significantly associated with higher levels of trust, but also find that frequency of reading non-fiction books and journals is important.

Residential stability and its correlate home ownership, have also been found to be associated with greater civic engagement (Saunders 1979; Winter 1994). Putnam, however, dismisses a link he at one stage thought important, as a cause of the decline of civic America. Rates of both long-distance and short distance mobility have in fact fallen over the past 50 years. In contrast, Onyx and Bullen's Australian work, finds residential stability to be important to social capital; the longer one has lived in a local community the stronger the neighbourhood ties are likely to be (Onyx and Bullen 1997: 22).

Increasing female labour force participation, the emergence of the dual career household, and associated time stress, represent a third cluster of factors examined by Putnam. His findings are first, women in the labour force are members of more voluntary associations than housewives. Second, women who work part-time are more trusting and more civically engaged than either full-time workers or those who do not work outside the home at all (Putnam 1996: 6). Third, amongst workers, those who work longer hours are more civically engaged (Putnam 1996: 7). These findings, he sums up, make it difficult to conclude that longer working hours, for men or women, are linked to declining social capital.

Limitations of the data in Australia lead to a similar conclusion. Female labour force participation is linked to the decline in volunteering by the Australian Bureau of Statistics (1997) and Warburton (1997). However, Lyons and Fabiansson (1998) demonstrate, to the extent the data allow, that this is unlikely given that the male volunteering rate has fallen more sharply than the female. A significant limitation of the data, however, preventing a conclusive finding either way, is that the data do not disaggregate between women in and out of the labour force. With approximately half of all working age women in the labour force and half not, it is difficult to impugn from volunteering data on all women, just what the impact of increasing labour force participation is. Specifically, we need to know if the fall in the volunteering rate of women is occurring predominantly amongst those in the labour force or not.

In relation to the importance of marriage for social capital, Putnam finds that married men and women are about a third more trusting and belong to about 15 or 25 per cent more groups than comparable single men and women, controlling for differences in education, age and race (Putnam 1996: 8). Interestingly, it is not just that those separated or divorced are less civically engaged and trusting, but those never married as well. Thus falling marriage rates may be as important as rising divorce rates. Putnam (1996: 8) concludes 'successful marriage, especially if the family includes children, is statistically associated with greater social trust and civic engagement'.

Australian work has also found that people with children are more likely to participate in the local community (Onyx and Bullen 1997: 22). Baum et al. (2000) identify the paradoxical role of the presence of children as both a conduit and a barrier to social and civic participation. On the one hand playgrounds, schools and children's sporting and leisure activities brought parents into contact with one another and widened social and civic networks. On the other hand, the responsibilities of child care could leave parents, particularly mothers, housebound and relatively isolated from broader community networks.

With regard to Putnam's particular aims of explaining decline in social capital, significant cohort differences in levels of trust and group membership, pivoting on the time at which television was first introduced, lead Putnam to pin much of the blame on television. Aside from the simple time displacement, the content of television programs is argued to increase levels of pessimism about human nature, thereby lowering trust on 'the commons'. All of the other family circumstances associated with higher or lower social capital pale in comparison with television viewing. 'Controlling for education, income, age, race, place of residence, work status and gender, television viewing is strongly and negatively related to social trust and group membership . . . An increase in television viewing of the magnitude that the United States has experienced in the last four decades might directly account for as much as one quarter to one half of the total drop in social capital' (Putnam 1996: 14).

To sum up Putnam's findings, levels of social capital are associated with the level of education, an absence of economic hardship, successful marriage, and the presence of children, but key causes of the decline in social capital are a generational shift in leisure patterns away from the social to the privatised passive activity of television viewing. 

How robust is Putnam's analysis? Putnam's (1996) measures of social capital are indeed that. He focuses upon the strength of the norm of social trust and measures the size of networks by the number of types of organisations to which people belong. Each of these are key aspects of the definition of social capital as a resource to action. Putnam does not, however, measure the strength of the norm of reciprocity. Furthermore, the measure of trust available to Putnam is limited to only one dimension of trust - trust of strangers. The General Social Survey data on trust, pitch the notion of trust at an abstract level. As Fukuyama (1999: 20-24) points out, this fails to differentiate trust of family members, co-ethnics, co-religious believers and trust of strangers. Whilst, in Putnam's defence, these are the best data available on levels of trust, do they amount to a sufficient base from which to conclude that civic America is disappearing? Given the widespread criticism of this particular claim (see, for example, Ladd 1996; Skocpol 1996), probably not.

The same criticism can be applied to Putnam's measure of networks - group memberships - that it is rather limited. Whilst adequately measuring the strength of a persons networks in formal organisations, although not the number of group memberships per se (Skockpol 1996: 1), it ignores the strength of informal networks through family, kin, neighbours, colleagues and the like. If one really wants to know about the capacity of someone's social networks, to assist as a resource to action, then clearly these sorts of networks are likely to be key.

Skocpol (1996) also makes an important point about the high level of aggregation that Putnam's analyses is pitched at. For example, in assessing the relationship between residential mobility and membership of organisations, Putnam simply relates the two national level trends. A stronger test of this hypothesis would be to disaggregate the data spatially into high and low residential mobility locations, and then to examine the rates of organisational membership in these different sorts of localities.

Very little work has examined the family-related circumstances associated with social capital in the community. Putnam is one of the few to consider the empirical evidence. The limited Australian evidence available has also been noted. Given this limited examination, plus the shortcomings of current data, what we know about the family circumstances and strong community-based social capital nexus is that it is associated with: higher levels of education; less economic hardship; and married life and the presence of children. The importance of residential stability and home ownership, and patterns of labour force participation remains unclear.

The mediating effect of locality

In the same way that aspects of family and household structure may shape the strength of social capital within family networks, so may aspects of locality and neighbourhood structure mediate the capacity of family-based social capital to manifest itself at the community level. That is, neighbourhood characteristics, such as crime rates, can affect the likelihood of family-based social capital bridging to the community. Even where families are well educated, financially well off, and married with children, and thus hypothetically rich in social capital, factors at the locality level may interrupt the expansion of social capital across community networks.

The links between family life and civil society take place within varied geographic settings - urban, suburban, rural, remote, crime-ridden, crime free, ethnically homogeneous or heterogeneous, and so on. These are the sorts of factors associated with particular places rather than particular families that may be important in shaping the vibrancy of civil society.

The significance of urban, suburban and rural living for the quality of social life, has been a continuing theme of sociological inquiry as far as back as the earliest days of the Chicago School, and can even be traced back to the work of the 'founding fathers' of sociology (Saunders 1981). Population size, population density, geographical isolation and residential propinquity have all been considered. Such is the case as well with social capital. Investigators have sought variation in levels of trust and civic engagement across the urban-rural divide. Putnam (1996) includes such variables in his survey of the decline of civic America. He finds that levels of trust in the twelve largest US metropolitan areas were approximately 10 per cent lower than in other towns and cities. With regard to the strength of networks, Putnam (1996) found that organisation memberships were 10 to 20 per cent lower in metropolitan areas in comparison with small rural towns.

The mediation of the relationship between family life and social capital by factors at a locality level has also attracted some attention in Australia. It has been found that people in small rural communities are trusting of local people but not as trusting of 'most Australians' as people in larger rural communities, and considerably less trusting than urban dwellers (Hughes, Bellamy and Black 1998: 7). Although these researchers appear to confuse locality size (what they actually have a measure of) with geographic isolation and density of population (what they refer to but do not have measures of), they conclude that: 'People in small communities may trust each other but not the stranger who comes into their midst. People in larger communities may be more trusting of strangers' (Hughes, Bellamy and Black 1998: 7).

In contrast, although Onyx and Bullen (1998: 20) find distinct differences in levels of social capital and in the mix of social capital factors in the five urban and rural communities they study, they do not attribute these differences to the nature of the localities per se.

Regarding ethnic homogeneity, Sampson, Raudenbush and Earls (1997: 921) examine the effect of violent crime in the United States upon social cohesion among neighbours. They find that levels of social cohesion are higher in neighbourhoods where immigrant concentrations are lower. This could be because ethnic homogeneity within a neighbourhood eases the process of developing shared social norms, due to the absence of language and cultural barriers. It is perhaps the same social logic that engenders familism within family networks.

The extent of income polarisation is also found to impact upon levels of social capital. Testing the hypothesis that spatial variations in income inequality predict the extent of investment in social capital, Kawachi et al. (1997) conclude that: 'A major finding of this study . . . is that the size of the gap between rich and poor is powerfully and negatively related to level of investment in social capital' (Kawachi et al. 1997: 1495). The greater the income inequality, the lower the level of per capita group membership, and the lower the level of social trust on average across the society. Putnam (1993: 224) found the same in his study of Italian regions, with income distribution being more egalitarian in the more social capital rich regions. At a national rather than regional level, a relationship between income inequality and levels of social capital has also been documented. Nations with higher and more equal incomes have been found to have stronger norms of trust and civic engagement than those with lower and less equal incomes (Knack and Keefer 1997: 1251).

Finally, Coleman's (1988) work suggests that the specific nature of local services can affect levels of social capital. The Catholic high schools in the neighbourhoods of Chicago he studied had far lower drop-out rates than the public schools or the other private schools. Coleman attributes this to the religious based community that surrounds such schools which provide 'integenerational closure that is based on a multiplex relation: whatever other relations they have, the adults are members of the same religious body and parents of children in the same school' (Coleman 1988: S114). Any study of the ways in which locality factors may affect the family-social capital dynamic will therefore also need to take into account the nature of the social networks constructed through local service provision.

Drawing these various elements together, the factors associated with particular localities that may mediate the relationship between family life and social capital beyond the family include the urban, suburban or rural nature of the locality, crime rates, ethnic homogeneity, income polarisation, and the nature of local service provision.

So what can we conclude about the family-social capital dynamic? Table 2 lists the factors that appear most likely to be associated with social capital within family networks; the bridging of family-based social capital into community networks; and mediation of the relationship between family based social capital and community networks.

Source: Australian Institute of Family Studies, 2000.

Clearly, current understanding of the family life-social capital nexus is very limited. This is partly because of the early stage at which social capital analysis finds itself, but also because data sets are being retrospectively analysed to construct inaccurate 'social capital' variables that result in poor measurement of the concept. Despite the avowed importance of family life to social capital, it has clearly not been a topic of sustained analysis.

It seems that family life is not itself something that needs to be explained, it has not been regarded as a theoretical subject. There are theoretical accounts that point to changes in the nature of family life, but they regard such change as derived. That is, the changes occurring in family life are being driven by structural changes elsewhere. These theoretical positions accuse 'too much state' (Becker 1981; Murray 1984; Fukuyama 1995, 1999), or alternatively 'too much market' (Wolfe 1989; Sennett 1998) of being the root cause of contemporary problems in family life.

The 'too much state' explanation (typically a neo-liberal position) is that the welfare state has 'crowded out' the roles of families and communities, supplanted tasks that they previously undertook, and left people dependent upon welfare provision. The 'too much market' explanation (typically a social democratic position) is that an over reliance upon market delivery and competition orients individuals to a consumerist 'what's in it for me' mentality, that undermines the cooperation and mutuality of family and community life.

In both these expositions the family is treated as 'something which is affected by other social changes rather than as a source of change itself' (Smart 1997: 306). To redress this situation, this paper concludes by integrating a conceptual focus upon social capital with recent theorised accounts of how and why family life is changing, to provide a theoretical framework that will support a clearer focus upon the family-social capital dynamic.

Risk society

Contemporary social change, theorists argue, is of such a dimension and quality that it amounts to a transformation, with the industrial era of modern society now giving way to 'risk society'. The two key aspects of social change that define risk society are 'manufactured uncertainty' and increased 'individualisation'.

Manufactured uncertainty refers to the risks or hazards produced in the growth of industrialism that now dominate contemporary ways of life (Beck 1996). It is millennium bugs rather than plagues of locusts. Increased individualisation is a compounding factor of manufactured uncertainty in a risk society in which families, communities, class groups and the like, are increasingly less important in setting the social rules or norms by which we live. This exhaustion of collective sources of meaning leads to a situation whereby individuals, rather than general or traditional social rules, are increasingly responsible for setting the boundaries to the ways in which we lead our lives (Beck 1996: 29-30).

Giddens (1992) presents two concepts that help us to understand the nature of family life in a risk society. The pure relationship is 'a social relationship entered into for its own sake, for what can be derived by each person from a sustained association with another; and which is continued only in so far as it is thought by both parties to deliver enough satisfactions for each individual to stay within it' (Giddens 1992: 58 and cited in Smart 1997: 307). In contrast with romantic love, confluent love is 'active, contingent love, and therefore jars with the "for-ever", "one-and-only" qualities of the romantic love complex' (Giddens 1992: 61 and cited in Smart 1997: 307). Romantic love is thus being surpassed by the pure relationship and confluent love. Rather than seeking the perfect partner, the perfect relationship is sought, 'and if one person does not provide it, one moves on until one finds it' (Smart 1997: 307).

Importantly, this conceptualisation pitches the changing nature of family life and intimate relations as a central part of, not just a reflection of, the processes of manufactured uncertainty and increased individualisation in an emergent risk society, or as Giddens prefers to term it, a post-traditional society. The pure relationship and confluent love are not the result of structural change in material circumstances, but rather an effect of changing mentalities (Smart 1996: 307). This focus upon changing mentalities accords with Giddens's project of reintroducing agency into sociological accounts, and allows for family members to be 'active agents rather than resembling the furniture in a household which is moved by external forces' (Smart 1996: 307). 

The emergence of the pure relationship and confluent love has particular implications for the practices of family life. The decision of someone living in a western society to get married has changed dramatically. 'Fifty years ago, someone who decided to marry knew what he or she was doing: marriage was a relatively fixed division of labour involving a specified status for each partner. Now no one quite knows any longer what marriage actually is, save that it is a 'relationship', entered into against the backdrop of profound changes affecting gender relations, the family, sexuality and the emotions' Giddens (1996:153). This uncertainty about marriage can be thought of as a manufactured uncertainty because it is a part of the social restructuring of gender, sexual and family relations.

One outcome associated with the rise of manufactured uncertainty and individualisation in a risk society, is a quantitative change in the diversity of family and household forms. Processes of individualisation lead to the breakdown of more traditional forms of social group which includes the nuclear family. Manufactured uncertainties about marriage (high divorce and falling marriage rates), childbirth (decreasing fertility rates), sexuality and gender roles, lead to a proliferation of other family and household forms, including increasing proportions of one-parent families, and group and living-alone households. As Beck summarises it: 'Marriage can be subtracted from sexuality, and that in turn from parenthood; parenthood can be multiplied by divorce; and the whole thing can be divided by living together or apart, and raised to a higher power by the possibility of multiple residences and the ever-present potentiality of taking-back decisions. This mathematical operation yields a rather large though fluctuating sum on the right side of the equation, and gives some idea of the variety of direct and multiply nested shadow existences that are more and more often concealed today behind the unchanged and so upright words 'marriage' and 'family'.' (Beck 1992: 116).

Nor is it simply that at any one point in time there is an increased diversity of household forms, but that over time any one individual is likely to find their biography littered with an array of living arrangements. Beck (1992: 119) says: 'A broad spectrum of variations on familial and extrafamilial forms of living together will arise and continue to exist side by side. Characteristically, many of these - single life, living together before and during marriage, living in communes, various parenthoods over one or two divorces, etc. - will be integrated as different phases into one overall biography.' 

Not only is there a multiplicity of household forms (the quantitative change), but these diverse forms are increasingly accepted as a part of the complex mosaic of family life in late modernity. There is thus a qualitative change in the social meaning of the diversity of household forms. As the incidence and timing of marriage has altered dramatically over the past 20 years, so have the norms and values associated with it - its social meaning - also been reshaped in the wake of increased individualisation. Consider the example of married life. 'Couples can [now] create for themselves the normative order of their relationship. Thus, if a couple agree to a certain set of boundaries, the important element is sticking to what is agreed rather than following general or traditional rules which are presumed to accompany one's status as husband or wife.' Smart (1997: 307) Individuals are increasingly responsible for setting and negotiating their own life patterns. 

Effectively the norms that govern family life are undergoing change as well. These changes in the norms of family life have been documented empirically by Finch (1989) and Finch and Mason (1993). The nature of this shift is from family norms being primarily obligational to being primarily negotiated (Finch 1989). Responsibilities towards family members were found to be fluid and not fixed - a topic for ongoing negotiation and renegotiation (Finch and Mason 1993: 166-167) that depended upon how one felt about a family member, rather than being predetermined by the role of 'daughter' or 'grandfather'. In this context, as Smart (1997: 308) observes, the taken-for-grantedness of kinship obligations loosens and trust has to be earned. The post-traditional family, or the family in risk society, is signified by 'the subjective meaning of intimate connections rather than formal, objective blood or marriage ties' (Silva and Smart 1999: 7).

Thus, 'family forms have moved away from a fixed or rigid notion of the "proper" family and . . . even inside traditionally structured families, new normative guidelines are emerging' (Silva and Smart 1999: 4). This understanding of changing family life points to changes in the scale of kin based networks and changes in the norms governing family life. With networks and norms being the two fundamental elements of the concept of social capital, and assuming that family life is fundamental to social capital, we have here a basis from which to understand why levels of social capital, in advanced western democracies at least, may be falling. What the precise nature of the change is, and whether it is happening for the hypothesised reasons is a matter for empirical inquiry.

The emergence of risk society, characterised by manufactured uncertainty and increased individualisation, which in the sphere of intimate relations is captured by the concepts of pure relationship and confluent love, identifies quantitative changes in the scale of family networks and qualitative changes in the norms operating within those family networks. With networks and norms being the fundamental elements of the concept of social capital, this risk society theoretical framework, provides a basis for understanding why the role of family life in the construction of social capital may be undergoing change.

Despite social capitalists (Bourdieu, Putnam, Fukuyama) asserting the importance of family life to social capital, our empirical understanding of this process is particularly limited. As this paper has argued, this is in part because family life has not been treated as a theoretical subject. A risk society theoretical framework provides a way forward, most particularly because it explains why family-based networks and norms are experiencing change, and these are the fundamental aspects of the family-social capital dynamic.

• Amato, P. (1998), 'More than money? Men's contributions to their children's lives', Chapter 13 in A. Booth & A. Creuter (eds) Men In Families: When Do They Get Involved? What Difference Does It Make?, Lawrence Erlbaum, New Jersey.
• ABS (1997), Australian Social Trends 1997, Australian Bureau of Statistics, Catalogue No. 4102.0, Canberra.
• Baum, F., Palmer, C., Modra, C., Murray, C. & Bush, R. (2000), 'Families, social capital and health', Chapter 10 in I. Winter (ed.) Social Capital and Public Policy in Australia, Australian Institute of Family Studies, Melbourne.
• Beck, U. (1992), Risk Society: Towards a New Modernity, Sage, London.
• Becker, G. (1981), Treatise on the Family, Harvard University Press, Harvard.
• Bourdieu, P. (1993), Sociology in Question, Sage, London.
• Coleman, J. (1988), 'Social capital in the creation of human capital', American Journal of Sociology, vol. 94, S. 95-120.
• Coleman, J. (1988a), 'The creation and destruction of social capital: implications for the law', Notre Dame Journal of Law, Ethics and Public Policy, vol. 3, pp. 375-404.
• Cox, E. ( 1995), A Truly Civil Society, ABC Books, Sydney.
• Finch, J. (1989), Family Obligations and Social Change, Polity Press, Cambridge.
• Finch, J. & Mason, J. (1993), Negotiating Family Responsibilities, Routledge, London.
• Foley, M. & Edwards, B. (1997), 'Escape from politics? Social theory and the social capital debate', American Behavioural Scientist, vol. 40, no. 5, pp. 550 561.
• Fukuyama, F. (1995), Trust: The Social Virtues and the Creation of Prosperity, Penguin, London.
• Fukuyama, F. (1999), The Great Disruption: Human Nature and the Reconstitution of Social Order, Free Press, New York.
• Furstenberg, F. (1995), 'Social capital and successful development among at risk youth', Journal of Marriage and the Family, vol. 57, pp. 580-592.
• Furstenberg, G. (1998), 'Social capital and the role of fathers in the family', Chapter 15 in A. Booth & A. Creuter (eds) Men In Families: When Do They Get Involved? What Difference Does It Make?, Lawrence Erlbaum, New Jersey.
• Giddens, A. (1982), Profiles and Critiques in Social Theory, Macmillan, London. 
  
  Giddens, A. (1992), The Transformation of Intimacy, Polity Press, Cambridge.
• Giddens, A. (1996), 'Affluence, poverty and the idea of a post-scarcity society', in C. Hewitt de Alcantara (ed.) Social Futures, Global Visions, Blackwell, Oxford.
• Giddens, A. (1998), The Third Way: The Renewal of Social Democracy, Polity Press, London.
• Granovetter, M. (1973), 'The strength of weak ties', American Journal of Sociology, vol. 78, no. 6, pp. 1360-1380.
• Harriss, J. & De Renzio, P. (1997), 'Missing link or analytically missing? The concept of social capital', Journal of International Development, vol. 9, no. 7, pp. 919-937.
• Hughes, P., Bellamy, J. & Black, A. (1998), 'Social capital and religious faith', Zadok Paper, S97, Spring Summer 1998/1999.
• Kawachi, I., Kennedy, B., Lochner, K. & Prothrow-Stith, D. (1997), 'Social capital, income inequality and mortality', American Journal of Public Health, vol. 87, no. 9, pp. 1491-1498.
• Knack, S. & Keefer, P. (1997), 'Does social capital have an economic payoff? A cross-country investigation', The Quarterly Journal of Economics, November, pp. 1251-1288.
• Latham, M. (1997), 'The search for social capital', in A. Norton et al. Social Capital: The individual, Society and the State, Centre for Independent Studies, Sydney.
• Ladd, E. (1996), 'The data just don't show erosion of America's social capital', The Public Perspective, June/July, pp. 1-22.
• Levi, M. (1996), 'Social and unsocial capital: a review essay of Robert Putnam's Making Democracy Work', Politics and Society, vol. 24, no. 1, pp. 45-55.
• Lyons, M. & Fabiansson, C. (1998), 'Is volunteering declining in Australia?', Australian Journal on Volunteering, vol. 3, no. 2, pp. 15-21.
• Morgan, D. (1999), 'Risk and family practices: accounting for change and fluidity in family life', Chapter 2 in E. Silva & C. Smart (eds) The New Family, Sage, London.
• Murray, C. (1984), Losing Ground, Basic Books, New York.
• Newton, K. (1997), 'Social capital and democracy', American Behavioural Scientist, vol. 40, no. 5, pp. 575-586.
• Norton, A. (1997), 'Social capital and civil society: some definitional issues', in A. Norton et al. Social Capital: The Individual, Society and the State, Centre for Independent Studies, Sydney.
• Onyx, J. & Bullen, P. (1997), Measuring Social Capital in Five Communities in NSW: An Analysis, Working Paper No. 41, Centre for Australian Community Organisations and Management, University of Technology, Sydney.
• Onyx, J. & Bullen, P. (1998), Measuring Social Capital in Five Communities in NSW: A Practitioner's Guide, Management Alternatives Pty Ltd, Sydney.
• Parcel, T. & Menaghan, E. (1993), 'Family social capital and children's behavioural problems', Social Psychology Quarterly, vol. 56, no. 2, pp. 120 135.
• Portes, A. (1998), 'Social capital: its origins and applications in modern sociology', Annual Review of Sociology, vol. 24, pp. 1-24.
• Putnam, R. (1995), 'Bowling alone: America's declining social capital', Journal of Democracy, vol. 6, no. 1, pp. 65-78.
• Putnam, R. (1993), Making Democracy Work: Civic Traditions in Modern Italy, Princeton University Press, Princeton
• Putnam, R. (1993a), 'The prosperous community: social capital and public life', The American Prospect, Spring, pp. 35-42.
• Putnam, R. (1996), 'The strange disappearance of civic America', Policy, Autumn, pp. 3-15.
• Putnam, R. (1998), 'Foreword', Housing Policy Debate, vol. 9, no. 1, pp. v viii.
• Putzel, J. (1997), 'Accounting for the "dark side" of social capital: reading Robert Putnam on democracy', Journal of International Development, vol. 9, no. 7, pp. 939-949.
• Sampson, R., Raudenbush, S. & Earls, F. (1997), 'Neighbourhoods and violent crime: a multilevel study of collective efficacy', Science 277, August, pp. 918-924. 
  
  Saunders, P. (1981), Social Theory and the Urban Question, Hutchinson, London.
• Sennett, R. (1998), The Corrosion of Character: The Personal Consequences of Work in the New Capitalism, WW Norton and Co., New York.
• Silva, E. & Smart, C. (1999), 'The "new" practices and politics of family life', Chapter 1 in E. Silva & C. Smart (eds) The New Family, Sage, London.
• Skocpol, T. (1996), 'Unravelling from above', The American Prospect, vol. 25, pp. 20-25.
• Stewart Weeks, M. & Richardson, C. (1998), Social Capital Stories, Centre for Independent Studies, Sydney.
• Valenzuela, A. & Dornbusch, S. (1994), 'Familism and social capital in the academic achievement of Mexican origin and Anglo adolescents', Social Science Quarterly, vol. 75, no. 1, pp. 18-36.
• Warburton, J. (1997), 'Older people as a "rich resource" for volunteer organisations: an overview of the issues', Australian Journal of Volunteering, vol. 2, no. 1, pp. 17-25.
• Weber, M. (1947), The Theory of Social and Economic Organisation, T. Parsons (ed.) Free Press, New York.
• Winter, I. (1994), The Radical Home Owner: Housing Tenure and Social Change, Gordon and Breach, Basel.
• Wolfe, A. (1989), Whose Keeper: Social Science and Moral Obligation, University of California Press, Berkely.
• Woolcock, M. (1998), 'Social capital and economic development: toward a theoretical synthesis and policy framework', Theory and Society, vol. 27, pp. 151-208.